*3Mr. Chief Justice Vinson

delivered the opinion of the Court.

Indictments have been returned in the Southern District of California charging the twelve petitioners with conspiring to violate the Smith Act, 18 U. S. C. (Supp. IV) §§ 371, 2385. Upon their arrest, bail was fixed for each petitioner in the widely varying amounts of $2,500, $7,500, $75,000 and $100,000. On motion of petitioner Schneiderman following arrest in the Southern District of New York, his bail was reduced to $50,000 before his removal to California. On motion of the Government to increase bail in the case of- other petitioners, and after several intermediate procedural steps not material to the issues presented here, bail was fixed in the District Court for the Southern District of California in the uniform amount of $50,000 for each petitioner.

Petitioners moved to reduce bail on the ground that bail as fixed was excessive under the Eighth Amendment.1 In support of their motion, petitioners submitted statements as to their financial resources, family relationships, health, prior criminal records, and other information. The only evidence offered by the Government was a certified record showing that four persons previously convicted under the Smith Act in the Southern District of New York had forfeited bail. No evidence was produced relating those four persons to the petitioners in this case. At a hearing on the motion, petitioners were examined by the District Judge and cross-examined by an attorney for the Government. Petitioners’ factual statements stand uncontroverted.

After their motion to reduce bail was denied, petitioners filed applications for habeas corpus in the same *4District Court. Upon consideration of the record on the motion to reduce bail, the writs were denied. The Court of Appeals for the Ninth Circuit affirmed. 192 F. 2d 56. Prior to filing their petition for certiorari in this Court, petitioners filed with Mr. Justice Douglas an application for bail and an alternative application for habeas corpus seeking interim relief. Both applications were referred to the Court and the matter was set down for argument on specific questions covering the issues raised by this case.

Relief in this type of case must be speedy if it is to be effective. The petition for certiorari and the full record are now before the Court and, since the questions presented by the petition have been fully briefed and argued, we consider it appropriate to dispose of the petition for certiorari at this time. Accordingly, the petition for certiorari is granted for review of questions important to the administration of criminal justice.2

First. From the passage of the Judiciary Act of 1789, 1 Stat. 73, 91, to the present Federal Rules of Criminal Procedure, Rule 46 (a) (1), federal law has unequivocally provided that a person arrested for a non-capital offense shall be admitted to bail. This traditional right to freedom before conviction permits the unhampered preparation of a defense, and serves to prevent the infliction of punishment prior to conviction. See Hudson v. Parker, 156 U. S. 277, 285 (1895). Unless this right to bail before trial is preserved, the presumption of innocence, secured only after centuries of struggle, would lose its meaning.

The right to Release before trial is conditioned upon the accused’s giving .adequate assurance that he will stand trial and submit to sentence if found guilty. Ex *5 parte Milburn, 9 Pet. 704, 710 (1835). Like the ancient practice of securing the oaths of responsible persons to stand as sureties , for the accused, the modern practice of requiring a bail bond or the deposit of a sum of money subject to forfeiture serves as additional assurance of the presence of an accused. Bail set at a figure higher than an amount reasonably calculated to fulfill this purpose is “excessive” under the Eighth Amendment. See United States v. Motlow, 10 F. 2d 657 (1926, opinion by Mr. Justice Butler as Circuit Justice of the Seventh Circuit).

Since the function of bail is limited, the fixing of bail for any individual defendant must be based upon standards relevant to the purpose of assuring the presence of that defendant. The traditional standards as expressed in the Federal Rules of Criminal Procedure3 are to be applied in each case to each defendant. In this case petitioners are charged with offenses under the Smith Act and, if found guilty, their convictions are subject to review with the scrupuloiis care demanded by our Constitution. Dennis v. United States, 341 U. S. 494, 516 (1951). Upon final judgment of conviction, petitioners face imprisonment of not more than five years and a fine of not more than $10,000. It is not denied that bail for each petitioner has been fixed in a sum much higher than that usually imposed for offenses with like penalties and yet there has been no factual showing to justify such action in this case. The Government asks the courts to depart from the norm by assuming, without the introduction of evidence, that each petitioner is a pawn in *6a conspiracy and will, in obedience to a superior, flee the jurisdiction. To infer from the fact of indictment alone a need for bail in an unusually high amount is an arbitrary act. Such conduct would inject into our own system of government the very principles of totalitarianism which Congress was seeking to guard against in passing the statute under which petitioners' have been indicted.

. If bail in an amount greater than that usually fixed for serious charges of. crimes is required in the case of any of the petitioners, that is a matter to which evidence should be directed in a. hearing so that the constitutional rights of each petitioner may be preserved. In the absence of such a showing, we are of the opinion that the fixing of bail before trial in these cases cannot be squared with the statutory and constitutional standards for admission to bail.

Second. The proper procédure for challenging bail as unlawfully fixed is by motion for reduction of bail and appeal to the Court of Appeals from an order denying such motion. Petitioners’ motion to reduce bail did not merely invoke the discretion of the District Court setting bail within a zone of reasonableness, but challenged the bail as violating statutory and constitutional standards. As there is no discretion to refuse to reduce excessive bail, the order denying the motion to reduce bail is appealable as a “final decision” of the District Court under 28 U. S. C. (Supp. IV) § 1291. Cohen v. Beneficial Loan Corp., 337 U. S. 541, 545-547 (1949). In this case, however,-petitioners did not take an appeal from the order of the District Court denying their motion for reduction of bail. Instead, they presented their claims under the Eighth Amendment in applications for writs of habeas corpus. While habeas corpus is an appropriate remedy for one held in custody in violation of the Constitution, 28 U. S. C. (Supp. IV) § 2241 (c) (3), the District Court, should withhold relief in this colláteral *7habeas corpus action where an adequate remedy available in the criminal proceeding has not been exhausted. Ex parte Royall, 117 U. S. 241 (1886); Johnson v. Hoy, 227 U. S. 245 (1913).

The Court concludes that bail has not been fixed by proper methods in this case and that petitioners’ remedy is by motion to reduce bail, with right of appeal to the Court of Appeals. Accordingly, the judgment of the Court of Appeals is vacated and the case is remanded to the District Court with directions to vacate its order denying petitioners’ applications for writs of habeas corpus and to dismiss the applications without prejudice. Petitioners may move for reduction of bail in the criminal proceeding so that a hearing may be held for the purpose of fixing reasonable bail for each petitioner.

It is so ordered.

Mr. Justice Minton took no part in the consideration or decision of this case.

By Mr. Justice Jackson, whom Mr. Justice Frankfurter joins.

I think the principles governing allowance of bail have been misunderstood or too casually applied in these cases and that they should be returned to the Circuit Justice or the District Courts for reconsideration in the light of standards which it is our function to determine. We have heard the parties on only four specific questions relating to bail before conviction — two involving considerations of law and of fact which should determine the amount of bail, and two relating to the procedure for correcting any departure therefrom. I consider first the principles which govern release of accused persons upon bail pending their trial.

The practice of admission to bail, as it has evolved in Anglo-American law, is not a device for keeping persons *8in jail upon mere accusation until it is found convenient to give them a trial. On the contrary, the spirit of the procedure is to enable them to stay out of jail until a trial has found them guilty. Without this conditional privilege, even those wrongly accused are punished by a period of imprisonment while awaiting trial and are handicapped in consulting counsel, searching for evidence and witnesses, and preparing a defense. To open a way of escape from this handicap and possible injustice, Congress commands allowance of bail for one under charge of any offense not punishable by death, Fed. Rules Crim. Proc., 46 (a) (1) providing: “A person arrested for an offense not punishable by death shall be admitted to bail . . .” before conviction.

Admission to bail always involves a risk that the accused will take flight. T*hat is a calculated risk which the law takes as the price of our system of justice. We know that Congress anticipated that bail would enable some escapes, because it proyided a procedure for dealing with them. Fed. Rules Crim. Proc., 46 (f).

In allowance of bail, the duty of the judge is to. reduce the risk by fixing an amount reasonably calculated to hold the accused available for trial and its consequence. Fed. Rules Crim. Proc., 46 (c). But the judge is not free to make the sky the limit, because the Eighth Amendment to the Constitution says: “Excessive bail shall not be required . . . .”

Congress has reduced this generality in providing more precise standards, stating that “. . . the amount thereof shall be such as in the judgment of the commissioner or court or judge or justice will insure the presence of the defendant, having regard to the nature and circumstances of the offense charged, the weight of the evidence against him, the financial ability of the defendant to give bail and the character of the defendant.” Fed. Rules Crim. Proc., 46 (c).

*9These statutory standards are not challenged as unconstitutional, rather the amounts of bail established for these petitioners are alleged to exceed these standards. We submitted no constitutional questions to argument by the parties, and it is our duty to avoid constitutional issues if possible. For me, the record is inadequate to say what amounts would be reasonable in any particular one of these cases and I regard it as not the function of this Court to do so. Furthermore, the whole Court agrees that the remedy pursued in the circumstances of this case is inappropriate to test the question and bring it here. But I do think there is a fair showing that these congressionally enacted standards have not been correctly applied.

It is complained that the District Court fixed a uniform blanket bail chiefly by consideration of the nature of the accusation and did not take into account the difference in circumstances between different defendants. If this occurred, it is a clear violation of Rule 46 (c). Each defendant stands before the bar of justice as an individual. Even on a conspiracy charge defendants do not lose their separateness or identity. While it might be possible that these defendants are identical in financial ability, character and relation to the charge — elements Congress has directed to be regarded in fixing bail — I think it violates the law of probabilities. Each accused is entitled, to any benefits due to his good record, and misdeeds or a bad record should prejudice only those who are guilty of them. The question when application for bail is made relates to each one’s trustworthiness to appear for trial and what security will supply reasonable assurance of his appearance.

Complaint further is made that the courts below have been unduly influenced by recommendations of very high bail made by the grand jury. It is not the function of the grand jury to fix bail, and its volunteered advice is not *10governing. Since the grand jury is a secret body, ordinarily hearing no evidence but the prosecution’s, attended by no counsel except the prosecuting attorneys, it is obvious that it is not in a position to make an impartial recommendation. Its suggestion may indicate that those who have heard the evidence for the prosecution regard it as strongly indicative that the accused may be guilty of the crime charged. It could not mean more than that without hearing the defense, and it adds nothing to the inference from the fact of indictment. Such recommendations are better left unmade, and if made should be given no weight.

But the protest charges, and the defect in the proceedings below appears to be, that, provoked by the flight of certain Communists after conviction, the Government demands and public opinion supports a use of the bail power to keep Communist defendants in jail before conviction. Thus, the, amount is said to have been fixed not as a reasonable assurance of their presence at the trial, but also as an assurance they would remain in jail. There seems reason to believe that this may have been the spirit to which the courts below have yielded, and it is contrary to the whole policy and philosophy of bail. This is not to say that every defendant is entitled to such bail as he can provide, but he is entitled to an opportunity to make it in a reasonable amount. I think the whole matter should be reconsidered by the appropriate judges in the traditional spirit of bail procedure.

The other questions we have heard argued relate to the remedy appropriate when the standards for amount of bail are misapplied. Of course, procedural rights so vital cannot be without means of vindication. In view of the nature of the writ of habeas corpus, we should be reluctant to say that under no circumstances would it be appropriate. But that writ will best serve its purpose and be best protected from discrediting abuse if it *11is reserved for cases in which no other procedure will present the issues to the courts. Its use as a substitute for appeals or as an optional alternative to other remedies is not to be encouraged. Habeas corpus is not, in the absence of extraordinary circumstances, the procedure to test reasonableness of bail.

We think that, properly limited and administered, the motion to reduce bail will afford a practical, simple, adequate and expeditious procedure. In view of prevailing confusions and conflicts in practice, this Court should define and limit the procedure with considerable precision, in the absence of which we may flood the courts with motions and appeals in bail cases.

The-first fixing of bail, whether by a commissioner under Rule 5 (b), or upon removal under Rule 40 (a), Fed. Rules Crim. Proc., or by the court upon arraignment after indictment, 18 U. S. C. § 3141, is a serious exercise of judicial discretion. But often it must be done in haste — the defendant may be taken by surprise, counsel has just been engaged, or for other reasons the bail is fixed, without that full inquiry and consideration which the matter deserves. Some procedure for reconsideration is a practical necessity, and the court’s power over revocation or reduction is a continuing power which either party may invoke as changing circumstances may require. It is highly important that such preliminary matters as bail be disposed of with as much finality as possible in the District Court where the case is to be tried. It is close to the scene of the offense, most accessible to defendant, has opportunity to see and hear the defendant and the witnesses personally, and is likely to be best informed for sound exercise of discretion. Rarely will the original determination be disturbed, if carefully made, but if the accused moves to reduce or the Government to revoke bail, a more careful deliberation may then be made on the relevant evidence presented by the parties, *12and if the defendant or the Government is aggrieved by a denial of the motion an appeal may be taken on the record as it then stands.

It is my conclusion that an order denying reduction of bail is to be regarded as a final decision which may be appealed to the Court of Appeals. But this is not because every claim of excessive bail raises a constitutional question. It is because we may properly hold appeal to be a statutory right. While only a sentence constitutes a final judgment in a criminal case, Berman v. United States, 302 U. S. 211, 212, it is a final decision that Congress has made reviewable. 28 U. S. C. § 1291. While a final judgment always is a final decision, there are instances in which a final decision is not a final judgment. The purpose of the finality requirement is to avoid piecemeal disposition of the basic controversy in a single case “where the result of review will be ‘to halt in the orderly progress of a cause and consider incidentally a question which has happened to cross the path of such litigation....'" Cobbledick v. United States, 309 U. S. 323, 326. But an order fixing bail can be reviewed without halting the main trial — its issues are entirely independent of the issues to be tried — and unless it can be reviewed before sentence, it never can be reviewed at all. The relation of an order fixing bail to final judgment in a criminal case is analogous to an order determining the right to security in a civil proceeding, Cohen v. Beneficial Industrial Loan Corp., 337 U. S. 541, or other interlocutory orders reviewable under 28 U. S. C. § 1292. I would hold, therefore, that such orders are appealable.

I cannot agree, however, that an order determining what amount of bail is reasonable under the standards prescribed does not call for an exercise of discretion. The Court of Appeals is not required to reexamine every order complained of. They represent exercises of discretion, upon questions, usually, of fact. Trivial differences or *13frivolous objections should be dismissed. The Appellate Court should only reverse for clear abuse of discretion or other mistake of law. And it ought to be noted that this Court will not exercise its certiorari power in individual cases except where they are typical of a problem so important and general as to deserve the attention of the* supervisory power.

If we would follow this course of reasoning, I think in actual experience it would protect every right of the accused expeditiously and cheaply. At- the same time, it would not open the floodgates to a multitude of trivial disputes abusive of the motion procedure.

Having found that the habeas corpus proceeding was properly dismissed by the District Court, in which its judgment was affirmed by the Court of Appeals, we should to that extent affirm. Having thus decided that the procedure taken in this case is not the proper one to bring the question of excessiveness of bail before the courts, there is a measure of inconsistency and departure from usual practice in our discussion of matters not before us. Certainly it would be inappropriate to say now that any particular amount as to any particular defendant is either reasonable or excessive.- That concrete amount, in the light of each defendant’s testimony and that of the Government, should be fixed by the appropriate judge or Justice upon evidence relevant to the standards prescribed. It is not appropriate for the Court as a whole to fix bail where the powér has been given 'to individual judges and Justices to do so. But there is little in our books to help guide federal judges in bail practice, and the extraordinary and recurring nature of this particular problem seems to warrant a discussion of the merits in which we would not ordinarily engage.

It remains to answer our own question as to whether the power to grant bail is in the Court or in the Circuit *14Justice. There is considerable confusion as to the source and extent of that power.

Fed. Rules Crim. Proc., 46 (a) (1), with respect to noncapital cases does not state who has power to grant bail before conviction — it simply directs that in such case bail “shall”- be granted. For an answer to the “who” question it is necessary to turn to the Criminal Code.

18 U. S. C. A. § 3141, entitled “Power of courts and magistrates,” provides:

“Bail may be taken by any court, judge or magistrate authorized to arrest and commit offenders, but .in capital cases bail may be taken only by a court ■of the United States having original or appellate jurisdiction in criminal cases or by a justice or judge thereof.”

The power to arrest and commit offenders is contained in 18 U. S. C. A. § 3041, which states that:

“For any offense against the United States, the offender may, by any justice or judge of the United States, ... be arrested and imprisoned, or bailed, as the case may be, for trial before such court of the United States as by law has cognizance of- the offense.” (Italics added.)

The fact that this section specifically grants the power of arrest to “any justice ... of the United States” supports the conclusion that Justices of this Court have the power of arrest, and, having that power under this section, they therefore also have power to grant bail under § 3141.

The Reviser’s Notes to § 3141 disclose that it is the product of Rev. Stat. §§ 1015 and 1016, which were embodied verbatim in 18 U. S. C. (1940 ed.) §§ 596 and 597. The Reviser also states that, “Sections 596 and 597 of Title 18, U. S. C., 1940 ed., except as superseded by rule 46(a) (1) of the Federal Rules of Criminal Procedure are consolidated and rewritten in this section without *15 change of meaning. 80th Congress House Report No. 304.” (Italics added.) Since no change of-meaning was intended, the context of the old sections becomes pertinent.

Rev. Stat. § 1015 reads:

“Bail shall be admitted upon all arrests in criminal cases where the offense is not punishable by death; and in such cases it may be taken by any of the persons authorized by the preceding section to arrest and imprison offenders.”

“The preceding section,” § 1014, is the predecessor of 18 U. S. C. A. § 3041, and reads the same as that section, namely:

“For any crime or offense against the United States, the offender may, by any justice or judge of the United States, ... be arrested and imprisoned, or bailed, as the case may be, for trial'before such court of the United States as by law has cognizance of the offense. . . .” (Italicized words are those omitted in 18 U. S. C. A. § 3041.)

Going on in the Revised Statutes, § 1016 states that:

“Bail may be admitted upon all arrests in criminal cases where the punishment may be death; but in such cases it shall be taken only by the Supreme Court or a-circuit court, or by a justice of the Supreme Court, a circuit judge, or a judge of a district court, who shall exercise their discretion therein, having regard to the nature and circumstance of the offense, and of the evidence, and to the usages of law.”

The evident tenor of §§ 1015 and 1016, taken together with § 1014, is that a Justice of this Court is one of many who can grant bail in a noncapital case but is one of a restricted class who can grant bail in' a capital case. *16Section 1016 appears to narrow the class included in § 1015.

To correlate the Revised Statutes with the present statutory scheme:

1. Rule 46 (a) (1), reading as follows, is taken from Rev. Stat. §§ 1015 and 1016 insofar as the latter govern who shall be admitted to bail and the considerations to be given the admission to bail of a capital case defendant.

Rule 46 (a) (1), “Bail before conviction”:

“A person arrested for an offense not punishable by death shall be admitted to bail. Á person arrested for an offense punishable by death may be admitted to bail by any court or judge authorized by law to do so in the exercise of discretion, giving due weight to the evidence and to the nature and circumstances of the offense.”

2. 18 U. S. C. A. § 3041, governing power of arrest, is taken directly from Rev. Stat. § 1014.

3. 18 U. S. C. A. § 3141, setting out who may grant bail, is taken from Rev. Stat. §§ 1015 and 1016 insofar as the latter are apropos of that subject.

It thus appears that the scheme of the Revised Statutes has been taken over bodily into the present Code and Rules. The only change I perceive is that, under the Revised Statutes, there was no clear statutory authority for a court to grant bail in a noncapital case. Rev. Stat. § 1015 (and § 1014) applicable to such case speak only of individuals. 18 U. S. C. A. § 3141 confers the power on “any court, judge or magistrate authorized to arrest and commit offénders.” The only reasonable construction of the latter is the obvious literal one, that is, that courts as well as the individuals empowered to arrest and com*17mit offenders by 18 U. S. C. A. § 3041 are authorized to grant bail. This is substantiated by the language of Fed. Rules Crim. Proc., 46 (c), “Amount [of bail]”:

“If the defendant is admitted to bail, the amount thereof shall be such as in the judgment of the commissioner or court or judge or justice will insure the ■ presence of the defendant . . . .” (Italics added.)

That is the one difference between the Revised Statutes' scheme and the present — the power to grant bail in non-capital cases now clearly is vested in the courts as well as in individual judges and justices. .

With the premise provided by the Revisor that the power to grant bail before conviction is the same now as under the Revised Statutes, the one exception being the extension to the courts just noted, the conclusion follows that bail can be granted by any court of the United States, including this Court, or by any judge of the United States, including the Justices of this Court.

The next problem is how Rule 45 of the Rules of this Court is to be assimilated, with the foregoing. Only the first and fourth subsections of the Rule have any present pertinence. They read as follows:

“1. Pending review of a decision refusing a writ of habeas corpus, the custody of the prisoner shall not be disturbed.

“4. The initial order respecting the custody or enlargement of the prisoner pending review, as also any recognizance taken, shall be deemed to cover not only the review in the intermediate appellate court but also the further possible review in this court; and only where special reasons therefor are shown to this court will it disturb that order, or make any independent order in that regard.”

*18The apparent conflict between the two subsections disappears when subsection 4 is viewed as a reservation of power in this Court only, not in an individual Justice of this Court, to issue an order in exceptional cases disturbing the custody of the prisoner. No other court and no individual judge or justice can disturb the custody, of the prisoner. See Carlson v. Landon, 341 U. S. 918.

The next problem is the bearing, if any, of Fed. Rules Crim. Proc., 46 (a) (2), covering the right to bail “Upon Review.” It reads:

“Bail may be allowed pending appeal or certiorari only if it appears that the case involves a substantial question which should be determined by the appellate court. Bail may be allowed by the trial judge or by the appellate court or by any judge thereof or by the circuit justice. . ...”

Insofar as it might be applicable to petitioners’ case, since they were seeking a review when .they filed their petition for bail, it would not seem that it has any efficacy. They have not yet been tried for the offense for which they have been indicted, so that the much wider powers of bail conferred by the statutes governing bail before conviction are applicable. Rule 46 (a) (2) is only intended to apply where a review of a conviction on the merits is sought.

Turning back to the case at hand, and treating the application to Mr. Justice Douglas for bail as one for bail pending review of a denial of habeas corpus, I think it clear that he does not have power to grant bail, but the full Court does have' that power. However, since the Court sustains the denial of habeas corpus, treating the application for bail strictly as one pending review of the denial of habeas corpus, the problems it raises are actually moot. If the application to Mr. Justice Douglas be treated as one made for fixing bail in the original case, it is my opinion that he has power to entertain it.

Chief Justice Rehnquist

delivered the opinion of the Court.

The Bail Reform Act of 1984 (Act) allows a federal court to detain an arrestee pending trial if the Government demonstrates by clear and convincing evidence after an adversary hearing that no release conditions “will reasonably assure . . . the safety of any other person and the community.” The United States Court of Appeals for the Second Circuit struck down this provision of the Act as facially unconstitutional, because, in that court’s words, this type of pretrial detention violates “substantive due process.” We granted certiorari because of a conflict among the Courts of Appeals regarding the validity of the Act.1 479 U. S. 929 (1986). We hold that, as against the facial attack mounted by these respondents, the Act fully comports with constitutional requirements. We therefore reverse.

*742r — 4

Responding to “the alarming problem of crimes committed by persons on release,” S. Rep. No. 98-225, p. 3 (1983), Congress formulated the Bail Reform Act of 1984, 18 U. S. C. §3141 et seq. (1982 ed., Supp. III), as the solution to a bail crisis in the federal courts. The Act represents the National Legislature’s considered response to numerous perceived deficiencies in the federal bail process. By providing for sweeping changes in both the way federal courts consider bail applications and the circumstances under which bail is granted, Congress hoped to “give the courts adequate authority to make release decisions that give appropriate recognition to the danger a person may pose to others if released.” S. Rep. No. 98-225, at 3.

To this end, § 3141(a) of the Act requires a judicial officer to determine whether an arrestee shall be detained. Section 3142(e) provides that “[i]f, after a hearing pursuant to the provisions of subsection (f), the judicial officer finds that no condition or combination of conditions will reasonably assure the appearance of the person as required and the safety of any other person and the community, he shall order the detention of the person prior to trial.” Section 3142(f) provides the arrestee with a number of procedural safeguards. He may request the presence of counsel at the detention hearing, he may testify and present witnesses in his behalf, as well as proffer evidence, and he may cross-examine other witnesses appearing at the hearing. If the judicial officer finds that no conditions of pretrial release can reasonably assure the safety of other persons and the community, he must state his findings of fact in writing, § 3142(i), and support his conclusion with “clear and convincing evidence,” § 3142(f).

The judicial officer is not given unbridled discretion in making the detention determination. Congress has specified the considerations relevant to that decision. These factors include the nature and seriousness of the charges, the substan-tiality of the Government’s evidence against the arrestee, the *743arrestee’s background and characteristics, and the nature and seriousness of the danger posed by the suspect’s release. § 3142(g). Should a judicial officer order detention, the detainee is entitled to expedited appellate review of the detention order. §§ 3145(b), (c).

Respondents Anthony Salerno and Vincent Cafaro were arrested on March 21,1986, after being charged in a 29-count indictment alleging various Racketeer Influenced and Corrupt Organizations Act (RICO) violations, mail and wire fraud offenses, extortion, and various criminal gambling violations. The RICO counts alleged 35 acts of racketeering activity, including fraud, extortion, gambling, and conspiracy to commit murder. At respondents’ arraignment, the Government moved to have Salerno and Cafaro detained pursuant to § 3142(e), on the ground that no condition of release would assure the safety of the community or any person. The District Court held a hearing at which the Government made a detailed proffer of evidence. The Government’s case showed that Salerno was the “boss” of the Genovese crime family of La Cosa Nostra and that Cafaro was a “captain” in the Genovese family. According to the Government’s proffer, based in large part on conversations intercepted by a court-ordered wiretap, the two respondents had participated in wide-ranging conspiracies to aid their illegitimate enterprises through violent means. The Government also offered the testimony of two of its trial witnesses, who would assert that Salerno personally participated in two murder conspiracies. Salerno opposed the motion for detention, challenging the credibility of the Government’s witnesses. He offered the testimony of several character witnesses as well as a letter from his doctor stating that he was suffering from a serious medical condition. Cafaro presented no evidence at the hearing, but instead characterized the wiretap conversations as merely “tough talk.”

The District Court granted the Government’s detention motion, concluding that the Government had established by *744clear and convincing evidence that no condition or combination of conditions of release would ensure the safety of the community or any person:

“The activities of a criminal organization such as the Genovese Family do not cease with the arrest of its principals and their release on even the most stringent of bail conditions. The illegal businesses, in place for many years, require constant attention and protection, or they will fail. Under these circumstances, this court recognizes a strong incentive on the part of its leadership to continue business as usual. When business as usual involves threats, beatings, and murder, the present danger such people pose in the community is self-evident.” 631 F. Supp. 1364, 1375 (SDNY 1986).2

Respondents appealed, contending that to the extent that the Bail Reform Act permits pretrial detention on the ground that the arrestee is likely to commit future crimes, it is unconstitutional on its face. Over a dissent, the United States Court of Appeals for the Second Circuit agreed. 794 F. 2d 64 (1986). Although the court agreed that pretrial detention could be imposed if the defendants were likely to intimidate witnesses or otherwise jeopardize the trial process, it found “§ 3142(e)’s authorization of pretrial detention [on the ground of future dangerousness] repugnant to the concept of substantive due process, which we believe prohibits the total deprivation of liberty simply as a means of preventing future crimes.” Id., at 71-72. The court concluded that the Government could not, consistent with due process, detain persons who had not been accused of any crime merely because they were thought to present a danger to the community. Id., at 72, quoting United States v. Melendez-Carrion, 790 F. *7452d 984, 1000-1001 (CA2 1986) (opinion of Newman, J.). It reasoned that our criminal law system holds persons accountable for past actions, not anticipated future actions. Although a court could detain an arrestee who threatened to flee before trial, such detention would be permissible because it would serve the basic objective of a criminal system— bringing the accused to trial. The court distinguished our decision in Gerstein v. Pugh, 420 U. S. 103 (1975), in which we upheld police detention pursuant to arrest. The court construed Gerstein as limiting such detention to the “ ‘administrative steps incident to arrest.’ ” 794 F. 2d, at 74, quoting Gerstein, supra, at 114. The Court of Appeals also found our decision in Schall v. Martin, 467 U. S. 253 (1984), upholding postarrest, pretrial detention of juveniles, inapposite because juveniles have a lesser interest in liberty than do adults. The dissenting judge concluded that on its face, the Bail Reform Act adequately balanced the Federal Government’s compelling interests in public safety against the detainee’s liberty interests.

II

A facial challenge to a legislative Act is, of course, the most difficult challenge to mount successfully, since the challenger must establish that no set of circumstances exists under which the Act would be valid. The fact that the Bail Reform Act might operate unconstitutionally under some conceivable set of circumstances is insufficient to render it wholly invalid, since we have not recognized an “overbreadth” doctrine outside the limited context of the First Amendment. Schall v. Martin, supra, at 269, n. 18. We think respondents have failed to shoulder their heavy burden to demonstrate that the Act is “facially” unconstitutional.3

*746Respondents present two grounds for invalidating the Bail Reform Act’s provisions permitting pretrial detention on the basis of future dangerousness. First, they rely upon the Court of Appeals’ conclusion that the Act exceeds the limitations placed upon the Federal Government by the Due Process Clause of the Fifth Amendment. Second, they contend that the Act contravenes the Eighth Amendment’s proscription, against excessive bail. We treat these contentions in turn.

A

The Due Process Clause of the Fifth Amendment provides that “No person shall... be deprived of life, liberty, or property, without due process of law . . . .” This Court has held that the Due Process Clause protects individuals against two types of government action. So-called “substantive due process” prevents the government from engaging in conduct that “shocks the conscience,” Rochin v. California, 342 U. S. 165, 172 (1952), or interferes with rights “implicit in the concept of ordered liberty,” Palko v. Connecticut, 302 U. S. 319, 325-326 (1937). When government action depriving a person of life, liberty, or property survives substantive due process scrutiny, it must still be implemented in a fair manner. Mathews v. Eldridge, 424 U. S. 319, 335 (1976). This requirement has traditionally been referred to as “procedural” due process.

Respondents first argue that the Act violates substantive due process because the pretrial detention it authorizes constitutes impermissible punishment before trial. See Bell v. Wolfish, 441 U. S. 520, 535, and n. 16 (1979). The Government, however, has never argued that pretrial detention could be upheld if it were “punishment.” The Court of Appeals assumed that pretrial detention under the Bail Reform Act is regulatory, not penal, and we agree that it is.

As an initial matter, the mere fact that a person is detained does not inexorably lead to the conclusion that the government has imposed punishment. Bell v. Wolfish, supra, at *747537. To determine whether a restriction on liberty constitutes impermissible punishment or permissible regulation, we first look to legislative intent. Schall v. Martin, 467 U. S., at 269. Unless Congress expressly intended to impose punitive restrictions, the punitive/regulatory distinction turns on “‘whether an alternative purpose to which [the restriction] may rationally be connected is assignable for it, and whether it appears excessive in relation to the alternative purpose assigned [to it].’” Ibid., quoting Kennedy v. Mendoza-Martinez, 372 U. S. 144, 168-169 (1963).

We conclude that the detention imposed by the Act falls on the regulatory side of the dichotomy. The legislative history of the Bail Reform Act clearly indicates that Congress did not formulate the pretrial detention provisions as punishment for dangerous individuals. See S. Rep. No. 98-225, at 8. Congress instead perceived pretrial detention as a potential solution to a pressing societal problem. Id., at 4-7. There is no doubt that preventing danger to the community is a legitimate regulatory goal. Schall v. Martin, supra.

Nor are the incidents of pretrial detention excessive in relation to the regulatory goal Congress sought to achieve. The Bail Reform Act carefully limits the circumstances under which detention may be sought to the most serious of crimes. See 18 U. S. C. § 3142(f) (detention hearings available if case involves crimes of violence, offenses for which the sentence is life imprisonment or death, serious drug offenses, or certain repeat offenders). The arrestee is entitled to a prompt detention hearing, ibid., and the maximum length of pretrial detention is limited by the stringent time limitations of the Speedy Trial Act.4 See 18 U. S. C. §3161 et seq. (1982 ed. and Supp. III). Moreover, as in Schall v. Martin, the conditions of confinement envisioned by the Act “appear to reflect the regulatory purposes relied upon by the” Government. *748467 U. S., at 270. As in Schall, the statute at issue here requires that detainees be housed in a “facility separate, to the extent practicable, from persons awaiting or serving sentences or being held in custody pending appeal.” 18 U. S. C. § 3142(i)(2). We conclude, therefore, that the pretrial detention contemplated by the Bail Reform Act is regulatory in nature, and does not constitute punishment before trial in violation of the Due Process Clause.

The Court of Appeals nevertheless concluded that “the Due Process Clause prohibits pretrial detention on the ground of danger to the community as a regulatory measure, without regard to the duration of the detention.” 794 F. 2d, at 71. Respondents characterize the Due Process Clause as erecting an impenetrable “wall” in this area that “no governmental interest — rational, important, compelling or otherwise — may surmount.” Brief for Respondents 16.

We do not think the Clause lays down any such categorical imperative. We have repeatedly held that the Government’s regulatory interest in community safety can, in appropriate circumstances, outweigh an individual’s liberty interest. For example, in times of war or insurrection, when society’s interest is at its peak, the Government may detain individuals whom the Government believes to be dangerous. See Ludecke v. Watkins, 335 U. S. 160 (1948) (approving un-reviewable executive power to detain enemy aliens in time of war); Moyer v. Peabody, 212 U. S. 78, 84-85 (1909) (rejecting due process claim of individual jailed without probable cause by Governor in time of insurrection). Even outside the exigencies of war, we have found that sufficiently compelling governmental interests can justify detention of dangerous persons. Thus, we have found no absolute constitutional barrier to detention of potentially dangerous resident aliens pending deportation proceedings. Carlson v. Landon, 342 U. S. 524, 537-542 (1952); Wong Wing v. United States, 163 U. S. 228 (1896). We have also held that the government may detain mentally unstable individuals who present a dan*749ger to the public, Addington v. Texas, 441 U. S. 418 (1979), and dangerous defendants who become incompetent to stand trial, Jackson v. Indiana, 406 U. S. 715, 731-739 (1972); Greenwood v. United States, 350 U. S. 366 (1956). We have approved of postarrest regulatory detention of juveniles when they present a continuing danger to the community. Schall v. Martin, supra. Even competent adults may face substantial liberty restrictions as a result of the operation of our criminal justice system. If the police suspect an individual of a crime, they may arrest and hold him until a neutral magistrate determines whether probable cause exists. Gerstein v. Pugh, 420 U. S. 103 (1975). Finally, respondents concede and the Court of Appeals noted that an arrestee may be incarcerated until trial if he presents a risk of flight, see Bell v. Wolfish, 441 U. S., at 534, or a danger to witnesses.

Respondents characterize all of these cases as exceptions to the “general rule” of substantive due process that the government may not detain a person prior to a judgment of guilt in a criminal trial. Such a “general rule” may freely be conceded, but we think that these cases show a sufficient number of exceptions to the rule that the congressional action challenged here can hardly be characterized as totally novel. Given the well-established authority of the government, in special circumstances, to restrain individuals’ liberty prior to or even without criminal trial and conviction, we think that the present statute providing for pretrial detention on the basis of dangerousness must be evaluated in precisely the same manner that we evaluated the laws in the cases discussed above.

The government’s interest in preventing crime by arrest-ees is both legitimate and compelling. De Veau v. Braisted, 363 U. S. 144, 155 (1960). In Schall, supra, we recognized the strength of the State’s interest in preventing juvenile crime. This general concern with crime prevention is no less compelling when the suspects are adults. Indeed, “[t]he *750harm suffered by the victim of a crime is not dependent upon the age of the perpetrator.” Schall v. Martin, supra, at 264-265. The Bail Reform Act of 1984 responds to an even more particularized governmental interest than the interest we sustained in Schall. The statute we upheld in Schall permitted pretrial detention of any juvenile arrested on any charge after a showing that the individual might commit some undefined further crimes. The Bail Reform Act, in contrast, narrowly focuses on a particularly acute problem in which the Government interests are overwhelming. The Act operates only on individuals who have been arrested for a specific category of extremely serious offenses. 18 U. S. C. § 3142(f). Congress specifically found that these individuals are far more likely to be responsible for dangerous acts in the community after arrest. See S. Rep. No. 98-225, at 6-7. Nor is the Act by any means a scattershot attempt to incapacitate those who are merely suspected of these serious crimes. The Government must first of all demonstrate probable cause to believe that the charged crime has been committed by the arrestee, but that is not enough. In a full-blown adversary hearing, the Government must convince a neutral decisionmaker by clear and convincing evidence that no conditions of release can reasonably assure the safety of the community or any person. 18 U. S. C. § 3142(f). While the Government’s general interest in preventing crime is compelling, even this interest is heightened when the Government musters convincing proof that the arrestee, already indicted or held to answer for a serious crime, presents a demonstrable danger to the community. Under these narrow circumstances, society’s interest in crime prevention is at its greatest.

On the other side of the scale, of course, is the individual’s strong interest in liberty. We do not minimize the importance and fundamental nature of this right. But, as our cases hold, this right may, in circumstances where the government’s interest is sufficiently weighty, be subordinated *751to the greater needs of society. We think that Congress’ careful delineation of the circumstances under which detention will be permitted satisfies this standard. When the Government proves by clear and convincing evidence that an arrestee presents an identified and articulable threat to an individual or the community, we believe that, consistent with the Due Process Clause, a court may disable the arrestee from executing that threat. Under these circumstances, we cannot categorically state that pretrial detention “offends some principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental.” Snyder v. Massachusetts, 291 U. S. 97, 105 (1934).

Finally, we may dispose briefly of respondents’ facial challenge to the procedures of the Bail Reform Act. To sustain them against such a challenge, we need only find them “adequate to authorize the pretrial detention of at least some [persons] charged with crimes,” Schall, supra, at 264, whether or not they might be insufficient in some particular circumstances. We think they pass that test. As we stated in Schall, “there is nothing inherently unattainable about a prediction of future criminal conduct.” 467 U. S., at 278; see Jurek v. Texas, 428 U. S. 262, 274 (1976) (joint opinion of Stewart, Powell, and Stevens, JJ.); id., at 279 (White, J., concurring in judgment).

Under the Bail Reform Act, the procedures by which a judicial officer evaluates the likelihood of future dangerousness are specifically designed to further the accuracy of that determination. Detainees have a right to counsel at the detention hearing. 18 U. S. C. § 3142(f). They may testify in their own behalf, present information by proffer or otherwise, and cross-examine witnesses who appear at the hearing. Ibid. The judicial officer charged with the responsibility of determining the appropriateness of detention is guided by statutorily enumerated factors, which include the nature and the circumstances of the charges, the weight of the evidence, the history and characteristics of the putative of*752fender, and the danger to the community. § 3142(g). The Government must prove its case by clear and convincing evidence. § 3142(f). Finally, the judicial officer must include written findings of fact and a written statement of reasons for a decision to detain. § 3142(i). The Act’s review provisions, § 3145(c), provide for immediate appellate review of the detention decision.

We think these extensive safeguards suffice to repel a facial challenge. The protections are more exacting than those we found sufficient in the juvenile context, see Schall, supra, at 275-281, and they far exceed what we found necessary to effect limited postarrest detention in Gerstein v. Pugh, 420 U. S. 103 (1975). Given the legitimate and compelling regulatory purpose of the Act and the procedural protections it offers, we conclude that the Act is not facially invalid under the Due Process Clause of the Fifth Amendment.

B

Respondents also contend that the Bail Reform Act violates the Excessive Bail Clause of the Eighth Amendment. The Court of Appeals did not address this issue because it found that the Act violates the Due Process Clause. We think that the Act survives a challenge founded upon the Eighth Amendment.

The Eighth Amendment addresses pretrial release by providing merely that “[ejxcessive bail shall not be required.” This Clause, of course, says nothing about whether bail shall be available at all. Respondents nevertheless contend that this Clause grants them a right to bail calculated solely upon considerations of flight. They rely on Stack v. Boyle, 342 U. S. 1, 5 (1951), in which the Court stated that “[bjail set at a figure higher than an amount reasonably calculated [to ensure the defendant’s presence at trial] is ‘excessive’ under the Eighth Amendment.” In respondents’ view, since the Bail Reform Act allows a court essentially to set bail at an infinite amount for reasons not related to the risk of flight, it *753violates the Excessive Bail Clause. Respondents concede that the right to bail they have discovered in the Eighth Amendment is not absolute. A court may, for example, refuse bail in capital cases. And, as the Court of Appeals noted and respondents admit, a court may refuse bail when the defendant presents a threat to the judicial process by intimidating witnesses. Brief for Respondents 21-22. Respondents characterize these exceptions as consistent with what they claim to be the sole purpose of bail — to ensure the integrity of the judicial process.

While we agree that a primary function of bail is to safeguard the courts’ role in adjudicating the guilt or innocence of defendants, we reject the proposition that the Eighth Amendment categorically prohibits the government from pursuing other admittedly compelling interests through regulation of pretrial release. The above-quoted dictum in Stack v. Boyle is far too slender a reed on which to rest this argument. The Court in Stack had no occasion to consider whether the Excessive Bail Clause requires courts to admit all defendants to bail, because the statute before the Court in that case in fact allowed the defendants to be bailed. Thus, the Court had to determine only whether bail, admittedly available in that case, was excessive if set at a sum greater than that necessary to ensure the arrestees’ presence at trial.

The holding of Stack is illuminated by the Court’s holding just four months later in Carlson v. Landon, 342 U. S. 524 (1952). In that case, remarkably similar to the present action, the detainees had been arrested and held without bail pending a determination of deportability. The Attorney General refused to release the individuals, “on the ground that there was reasonable cause to believe that [their] release would be prejudicial to the public interest and would endanger the welfare and safety of the United States.” Id., at 529 (emphasis added). The detainees brought the same challenge that respondents bring to us today: the Eighth Amend*754ment required them to be admitted to bail. The Court squarely rejected this proposition:

“The bail clause was lifted with slight changes from the English Bill of Rights Act. In England that clause has never been thought to accord a right to bail in all cases, but merely to provide that bail shall not be excessive in those cases where it is proper to grant bail. When this clause was carried over into our Bill of Rights, nothing was said that indicated any different concept. The Eighth Amendment has not prevented Congress from defining the classes of cases in which bail shall be allowed in this country. Thus, in criminal cases bail is not compulsory where the punishment may be death. Indeed, the very language of the Amendment fails to say all arrests must be bailable.” Id., at 545-546 (footnotes omitted).

Carlson v. Landon was a civil case, and we need not decide today whether the Excessive Bail Clause speaks at all to Congress’ power to define the classes of criminal arrestees who shall be admitted to bail. For even if we were to conclude that the Eighth Amendment imposes some substantive limitations on the National Legislature’s powers in this area, we would still hold that the Bail Reform Act is valid. Nothing in the text of the Bail Clause limits permissible Government considerations solely to questions of flight. The only arguable substantive limitation of the Bail Clause is that the Government’s proposed conditions of release or detention not be “excessive” in light of the perceived evil. Of course, to determine whether the Government’s response is excessive, we must compare that response against the interest the Government seeks to protect by means of that response. Thus, when the Government has admitted that its only interest is in preventing flight, bail must be set by a court at a sum designed to ensure that goal, and no more. Stack v. Boyle, supra. We believe that when Congress has mandated detention on the basis of a compelling interest other than pre*755vention of flight, as it has here, the Eighth Amendment does not require release on bail.

I — I I — I HH

In our society liberty is the norm, and detention prior to trial or without trial is the carefully limited exception. We hold that the provisions for pretrial detention in the Bail Reform Act of 1984 fall within that carefully limited exception. The Act authorizes the detention prior to trial of arrestees charged with serious felonies who are found after an adversary hearing to pose a threat to the safety of individuals or to the community which no condition of release can dispel. The numerous procedural safeguards detailed above must attend this adversary hearing. We are unwilling to say that this congressional determination, based as it is upon that primary concern of every government — a concern for the safety and indeed the lives of its citizens — on its face violates either the Due Process Clause of the Fifth Amendment or the Excessive Bail Clause of the Eighth Amendment.

The judgment of the Court of Appeals is therefore

Reversed.

Justice Marshall,

with whom Justice Brennan joins, dissenting.

This case brings before the Court for the first time a statute in which Congress declares that a person innocent of any crime may be jailed indefinitely, pending the trial of allegations which are legally presumed to be untrue, if the Government shows to the satisfaction of a judge that the accused is likely to commit crimes, unrelated to the pending charges, at any time in the future. Such statutes, consistent with the usages of tyranny and the excesses of what bitter experience teaches us to call the police state, have long been thought incompatible with the fundamental human rights protected by our Constitution. Today a majority of this Court holds otherwise. Its decision disregards basic principles of justice *756established centuries ago and enshrined beyond the reach of governmental interference in the Bill of Rights.

A few preliminary words are necessary with respect to the majority’s treatment of the facts in this case. The two paragraphs which the majority devotes to the procedural posture are essentially correct, but they omit certain matters which are of substantial legal relevance.

The Solicitor General’s petition for certiorari was filed on July 21, 1986. On October 9, 1986, respondent Salerno filed a response to the petition. No response or appearance of counsel was filed on behalf of respondent Cafaro. The petition for certiorari was granted on November 3, 1986.

On November 19, 1986, respondent Salerno was convicted after a jury trial on charges unrelated to those alleged in the indictment in this case. On January 13, 1987, Salerno was sentenced on those charges to 100 years’ imprisonment. As of that date, the Government no longer required a pretrial detention order for the purpose of keeping Salerno incarcerated; it could simply take him into custody on the judgment and commitment order. The present case thus became moot as to respondent Salerno.1

*757The situation with respect to respondent Cafaro is still more disturbing. In early October 1986, before the Solicitor General’s petition for certiorari was granted, respondent Cafaro became a cooperating witness, assisting the Government’s investigation “by working in a covert capacity.”2 The information that Cafaro was cooperating with the Government was not revealed to his codefendants, including respondent Salerno. On October 9, 1986, respondent Cafaro was released, ostensibly “temporarily for medical care and treatment,” with the Government’s consent. Docket, SS 86 Cr. 245-2, p. 6 (MJL) (SDNY) (Lowe, J.).3 This release was conditioned upon execution of a personal recognizance bond in the sum of $1 million, under the general pretrial *758release provisions of 18 U. S. C. §3141 (1982 ed., Supp. III). In short, respondent Cafaro became an informant and the Government agreed to his release on bail in order that he might better serve the Government’s purposes. As to Cafaro, this case was no longer justiciable even before cer-tiorari was granted, but the information bearing upon the essential issue of the Court’s jurisdiction was not made available to us.

The Government thus invites the Court to address the facial constitutionality of the pretrial detention statute in a case involving two respondents, one of whom has been sentenced to a century of jail time in another case and released pending appeal with the Government’s consent, while the other was released on bail in this case, with the Government’s consent, because he had become an informant. These facts raise, at the very least, a substantial question as to the Court’s jurisdiction, for it is far from clear that there is now an actual controversy between these parties. As we have recently said, “Article III of the Constitution requires that there be a live case or controversy at the time that a federal court decides the case; it is not enough that there may have been a live case or controversy when the case was decided by the court whose judgment we are reviewing.” Burke v. Barnes, 479 U. S. 361, 363 (1987); see Sosna v. Iowa, 419 U. S. 393, 402 (1975); Golden v. Zwickler, 394 U. S. 103, 108 (1969). Only by flatly ignoring these matters is the majority able to maintain the pretense that it has jurisdiction to decide the question which it is in such a hurry to reach.

h-1 1 — 1

The majority approaches respondents’ challenge to the Act by dividing the discussion into two sections, one concerned with the substantive guarantees implicit in the Due Process Clause, and the other concerned with the protection afforded by the Excessive Bail Clause of the Eighth Amendment. This is a sterile formalism, which divides a unitary argument *759into two independent parts and then professes to demonstrate that the parts are individually inadequate.

On the due process side of this false dichotomy appears an argument concerning the distinction between regulatory and punitive legislation. The majority concludes that the Act is a regulatory rather than a punitive measure. The ease with which the conclusion is reached suggests the worthlessness of the achievement. The major premise is that “[ujnless Congress expressly intended to impose punitive restrictions, the punitive/regulatory distinction turns on ‘“whether an alternative purpose to which [the restriction] may rationally be connected is assignable for it, and whether it appears excessive in relation to the alternative purpose assigned [to it].”’” Ante, at 747 (citations omitted). The majority finds that “Congress did not formulate the pretrial detention provisions as punishment for dangerous individuals,” but instead was pursuing the “legitimate regulatory goal” of “preventing danger to the community.” Ibid.4 Concluding that pretrial detention is not an excessive solution to the problem of preventing danger to the community, the majority thus finds that no substantive element of the guarantee of due process invalidates the statute.

*760This argument does not demonstrate the conclusion it purports to justify. Let us apply the majority’s reasoning to a similar, hypothetical case. After investigation, Congress determines (not unrealistically) that a large proportion of violent crime is perpetrated by persons who are unemployed. It also determines, equally reasonably, that much violent crime is committed at night. From amongst the panoply of “potential solutions,” Congress chooses a statute which permits, after judicial proceedings, the imposition of a dusk-to-dawn curfew on anyone who is unemployed. Since this is not a measure enacted for the purpose of punishing the unemployed, and since the majority finds that preventing danger to the community is a legitimate regulatory goal, the curfew statute would, according to the majority’s analysis, be a mere “regulatory” detention statute, entirely compatible with the substantive components of the Due Process Clause.

The absurdity of this conclusion arises, of course, from the majority’s cramped concept of substantive due process. The majority proceeds as though the only substantive right protected by the Due Process Clause is a right to be free from punishment before conviction. The majority’s technique for infringing this right is simple: merely redefine any measure which is claimed to be punishment as “regulation,” and, magically, the Constitution no longer prohibits its imposition. Because, as I discuss in Part III, infra, the Due Process Clause protects other substantive rights which are infringed by this legislation, the majority’s argument is merely an exercise in obfuscation.

The logic of the majority’s Eighth Amendment analysis is equally unsatisfactory. The Eighth Amendment, as the majority notes, states that “[ejxcessive bail shall not be required.” The majority then declares, as if it were undeniable, that: “[t]his Clause, of course, says nothing about whether bail shall be available at all.” Ante, at 752. If excessive bail is imposed the defendant stays in jail. The same result is achieved if bail is denied altogether. Whether the *761magistrate sets bail at $1 billion or refuses to set bail at all, the consequences are indistinguishable. It would be mere sophistry to suggest that the Eighth Amendment protects against the former decision, and not the latter. Indeed, such a result would lead to the conclusion that there was no need for Congress to pass a preventive detention measure of any kind; every federal magistrate and district judge could simply refuse, despite the absence of any evidence of risk of flight or danger to the community, to set bail. This would be entirely constitutional, since, according to the majority, the Eighth Amendment “says nothing about whether bail shall be available at all.”

But perhaps, the majority says, this manifest absurdity can be avoided. Perhaps the Bail Clause is addressed only to the Judiciary. “[W]e need not decide today,” the majority says, “whether the Excessive Bail Clause speaks at all to Congress’ power to define the classes of criminal arrestees who shall be admitted to bail.” Ante, at 754. The majority is correct that this question need not be decided today; it was decided long ago. Federal and state statutes which purport to accomplish what the Eighth Amendment forbids, such as imposing cruel and unusual punishments, may not stand. See, e. g., Trop v. Dulles, 356 U. S. 86 (1958); Furman v. Georgia, 408 U. S. 238 (1972). The text of the Amendment, which provides simply that “[ejxcessive bail shall not be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted,” provides absolutely no support for the majority’s speculation that both courts and Congress are forbidden to inflict cruel and unusual punishments, while only the courts are forbidden to require excessive bail.5

*762The majority’s attempts to deny the relevance of the Bail Clause to this case are unavailing, but the majority is nonetheless correct that the prohibition of excessive bail means that in order “to determine whether the Government’s response is excessive, we must compare that response against the interest the Government seeks to protect by means of that response.” Ante, at 754. The majority concedes, as it must, that “when the Government has admitted that its only interest is in preventing flight, bail must be set by a court at a sum designed to ensure that goal, and no more.” Ibid. But, the majority says, “when Congress has mandated detention on the basis of a compelling interest other than prevention of flight, as it has here, the Eighth Amendment does not require release on bail.” Ante, at 754-755. This conclusion follows only if the “compelling” interest upon which Congress acted is an interest which the Constitution permits Congress to further through the denial of bail. The majority does not ask, as a result of its disingenuous division of the analysis, if there are any substantive limits contained in both the Eighth Amendment and the Due Process Clause which render this system of preventive detention unconstitutional. The majority does not ask because the answer is apparent and, to the majority, inconvenient.

Ill

The essence of this case may be found, ironically enough, in a provision of the Act to which the majority does not refer. Title 18 U. S. C. §3142(j) (1982 ed., Supp. Ill) provides that “[njothing in this section shall be construed as modifying or limiting the presumption of innocence.” But the very pith *763and purpose of this statute is an abhorrent limitation of the presumption of innocence. The majority’s untenable conclusion that the present Act is constitutional arises from a specious denial of the role of the Bail Clause and the Due Process Clause in protecting the invaluable guarantee afforded by the presumption of innocence.

“The principle that there is a presumption of innocence in favor of the accused is the undoubted law, axiomatic and elementary, and its enforcement lies at the foundation of the administration of our criminal law.” Coffin v. United States, 156 U. S. 432, 453 (1895). Our society’s belief, reinforced over the centuries, that all are innocent until the state has proved them to be guilty, like the companion principle that guilt must be proved beyond a reasonable doubt, is “implicit in the concept of ordered liberty,” Palko v. Connecticut, 302 U. S. 319, 325 (1937), and is established beyond legislative contravention in the Due Process Clause. See Estelle v. Williams, 425 U. S. 501, 503 (1976); In re Winship, 397 U. S. 358, 364 (1970). See also Taylor v. Kentucky, 436 U. S. 478, 483 (1978); Kentucky v. Whorton, 441 U. S. 786, 790 (1979) (Stewart, J., dissenting).

The statute now before us declares that persons who have been indicted may be detained if a judicial officer finds clear and convincing evidence that they pose a danger to individuals or to the community. The statute does not authorize the Government to imprison anyone it has evidence is dangerous; indictment is necessary. But let us suppose that a defendant is indicted and the Government shows by clear and convincing evidence that he is dangerous and should be detained pending a trial, at which trial the defendant is acquitted. May the Government continue to hold the defendant in detention based upon its showing that he is dangerous? The answer cannot be yes, for that would allow the Government to imprison someone for uncommitted crimes based upon “proof” not beyond a reasonable doubt. The result must therefore be that once the indictment has failed, detention *764cannot continue. But our fundamental principles of justice declare that the defendant is as innocent on the day before his trial as he is on the morning after his acquittal. Under this statute an untried indictment somehow acts to permit a detention, based on other charges, which after an acquittal would be unconstitutional. The conclusion is inescapable that the indictment has been turned into evidence, if not that the defendant is guilty of the crime charged, then that left to his own devices he will soon be guilty of something else. “ ‘If it suffices to accuse, what will become of the innocent?’” Coffin v. United States, supra, at 455 (quoting Ammianus Marcellinus, Rerum Gestarum Libri Qui Supersunt, L. XVIII, c. 1, A. D. 359).

To be sure, an indictment is not without legal consequences. It establishes that there is probable cause to believe that an offense was committed, and that the defendant committed it. Upon probable cause a warrant for the defendant’s arrest may issue; a period of administrative detention may occur before the evidence of probable cause is presented to a neutral magistrate. See Gerstein v. Pugh, 420 U. S. 103 (1975). Once a defendant has been committed for trial he may be detained in custody if the magistrate finds that no conditions of release will prevent him from becoming a fugitive. But in this connection the charging instrument is evidence of nothing more than the fact that there will be a trial, and

“release before trial is conditioned upon the accused’s giving adequate assurance that he will stand trial and submit to sentence if found guilty. Like the ancient practice of securing the oaths of responsible persons to stand as sureties for the accused, the modern practice of requiring a bail bond or the deposit of a sum of money subject to forfeiture serves as additional assurance of the *765presence of an accused.” Stack v. Boyle, 342 U. S. 1, 4-5 (1951) (citation omitted).6

The finding of probable cause conveys power to try, and the power to try imports of necessity the power to assure that the processes of justice will not be evaded or obstructed.7 “Pretrial detention to prevent future crimes against society at large, however, is not justified by any concern for holding a trial on the charges for which a defendant has been arrested.” 794 F. 2d 64, 73 (CA2 1986) (quoting United States v. Melendez-Carrion, 790 F. 2d 984, 1002 (CA2 1986) (opinion of Newman, J.)). The detention purportedly authorized by this statute bears no relation to the Government’s power to try charges supported by a finding of probable cause, and thus the interests it serves are outside the scope of interests which may be considered in weighing the excessiveness of bail under the Eighth Amendment.

*766It is not a novel proposition that the Bail Clause plays a vital role in protecting the presumption of innocence. Reviewing the application for bail pending appeal by members of the American Communist Party convicted under the Smith Act, 18 U. S. C. § 2385, Justice Jackson wrote:

“Grave public danger is said to result from what [the defendants] may be expected to do, in addition to what they have done since their conviction. If I assume that defendants are disposed to commit every opportune disloyal act helpful to Communist countries, it is still difficult to reconcile with traditional American law the jailing of persons by the courts because of anticipated but as yet uncommitted crimes. Imprisonment to protect society from predicted but unconsummated offenses is . . . unprecedented in this country and . . . fraught with danger of excesses and injustice....” Williamson v. United States, 95 L. Ed. 1379, 1382 (1950) (opinion in chambers) (footnote omitted).

As Chief Justice Vinson wrote for the Court in Stack v. Boyle, supra: “Unless th[e] right to bail before trial is preserved, the presumption of innocence, secured only after centuries of struggle, would lose its meaning.” 342 U. S., at 4.

> I — i

There is a connection between the peculiar facts of this case and the evident constitutional defects in the statute which the Court upholds today. Respondent Cafaro was originally incarcerated for an indeterminate period at the request of the Government, which believed (or professed to believe) that his release imminently threatened the safety of the community. That threat apparently vanished, from the Government’s point of view, when Cafaro agreed to act as a covert agent of the Government. There could be no more eloquent demonstration of the coercive power of authority to imprison upon prediction, or of the dangers which the almost *767inevitable abuses pose to the cherished liberties of a free society.

“It is a fair summary of history to say that the safeguards of liberty have frequently been forged in controversies involving not very nice people.” United States v. Rabinowitz, 339 U. S. 56, 69 (1950) (Frankfurter, J., dissenting). Honoring the presumption of innocence is often difficult; sometimes we must pay substantial social costs as a result of our commitment to the values we espouse. But at the end of the day the presumption of innocence protects the innocent; the shortcuts we take with those whom we believe to be guilty injure only those wrongfully accused and, ultimately, ourselves.

Throughout the world today there are men, women, and children interned indefinitely, awaiting trials which may never come or which may be a mockery of the word, because their governments believe them to be “dangerous.” Our Constitution, whose construction began two centuries ago, can shelter us forever from the evils of such unchecked power. Over 200 years it has slowly, through our efforts, grown more durable, more expansive, and more just. But it cannot protect us if we lack the courage, and the self-restraint, to protect ourselves. Today a majority of the Court applies itself to an ominous exercise in demolition. Theirs is truly a decision which will go forth without authority, and come back without respect.

I dissent.

Justice Stevens,

dissenting.

There may be times when the Government’s interest in protecting the safety of the community will justify the brief detention of a person who has not committed any crime, see ante, at 748-749, see also United States v. Greene, 497 F. 2d 1068, 1088-1089 (CA7 1974) (Stevens, J., dissenting).1 To *768use Judge Feinberg’s example, it is indeed difficult to accept the proposition that the Government is without power to detain a person when it is a virtual certainty that he or she would otherwise kill a group of innocent people in the immediate future. United States v. Salerno, 794 F. 2d 64, 77 (CA2 1986) (dissenting opinion). Similarly, I am unwilling to decide today that the police may never impose a limited curfew during a time of crisis. These questions are obviously not presented in this case, but they lurk in the background and preclude me from answering the question that is presented in as broad a manner as Justice. Marshall has. Nonetheless, I firmly agree with Justice Marshall that the provision of the Bail Reform Act allowing pretrial detention on the basis of future dangerousness is unconstitutional. Whatever the answers are to the questions I have mentioned, it is clear to me that a pending indictment may not be given any weight in evaluating an individual’s risk to the community or the need for immediate detention.

If the evidence of imminent danger is strong enough to warrant emergency detention, it should support that preventive measure regardless of whether the person has been charged, convicted, or acquitted of some other offense. In this case, for example, it is unrealistic to assume that the danger to the community that was present when respondents were at large did not justify their detention before they were indicted, but did require that measure the moment that the grand jury found probable cause to believe they had committed crimes in the past.2 It is equally unrealistic to assume that the danger will vanish if a jury happens to acquit them. *769Justice Marshall has demonstrated that the fact of indictment cannot, consistent with the presumption of innocence and the Eighth Amendment’s Excessive Bail Clause, be used to create a special class, the members of which are, alone, eligible for detention because of future dangerousness.

Several factors combine to give me an uneasy feeling about the case the Court decides today. The facts set forth in Part I of Justice Marshall’s opinion strongly support the possibility that the Government is much more interested in litigating a “test case” than in resolving an actual controversy concerning respondents’ threat to the safety of the community. Since Salerno has been convicted and sentenced on other crimes, there is no need to employ novel pretrial detention procedures against him. Cafaro’s case is even more curious because he is apparently at large and was content to have his case argued by Salerno’s lawyer even though his interests would appear to conflict with Salerno’s. But if the merits must be reached, there is no answer to the arguments made in Parts II and III of Justice Marshall’s dissent. His conclusion, and not the Court’s, is faithful to the “fundamental principles as they have been understood by the traditions of our people and our law.” Lochner v. New York, 198 U. S. 45, 76 (1905) (Holmes, J., dissenting). Accordingly, I respectfully dissent.

Justice GINSBURG delivered the opinion of the Court.

*686Tyson Timbs pleaded guilty in Indiana state court to dealing in a controlled substance and conspiracy to commit theft. The trial court sentenced him to one year of home detention and five years of probation, which included a court-supervised addiction-treatment program. The sentence also required Timbs to pay fees and costs totaling $ 1,203. At the time of Timbs's arrest, the police seized his vehicle, a Land Rover SUV Timbs had purchased for about $ 42,000. Timbs paid for the vehicle with money he received from an insurance policy when his father died.

The State engaged a private law firm to bring a civil suit for forfeiture of Timbs's Land Rover, charging that the vehicle had been used to transport heroin. After Timbs's guilty plea in the criminal case, the trial court held a hearing on the forfeiture demand. Although finding that Timbs's vehicle had been used to facilitate violation of a criminal statute, the court denied the requested forfeiture, observing that Timbs had recently purchased the vehicle for $ 42,000, more than four times the maximum $ 10,000 monetary fine assessable against him for his drug conviction. Forfeiture of the Land Rover, the court determined, would be grossly disproportionate to the gravity of Timbs's offense, hence unconstitutional under the Eighth Amendment's Excessive Fines Clause. The Court of Appeals of Indiana affirmed that determination, but the Indiana Supreme Court reversed. 84 N.E.3d 1179 (2017). The Indiana Supreme Court did not decide whether the forfeiture would be excessive. Instead, it held that the Excessive Fines Clause constrains only federal action and is inapplicable to state impositions. We granted certiorari. 585 U.S. ----, 138 S.Ct. 2650, 201 L.Ed.2d 1049 (2018).

The question presented: Is the Eighth Amendment's Excessive Fines Clause an "incorporated" protection applicable to the States under the Fourteenth Amendment's Due Process Clause? Like the Eighth Amendment's proscriptions of "cruel and unusual punishment" and "[e]xcessive bail," the protection against excessive fines guards against abuses of government's punitive or criminal-law-enforcement authority. This safeguard, we hold, is "fundamental to our scheme of ordered liberty," with "dee[p] root[s] in *687[our] history and tradition." McDonald v. Chicago , 561 U.S. 742, 767, 130 S.Ct. 3020, 177 L.Ed.2d 894 (2010) (internal quotation marks omitted; emphasis deleted). The Excessive Fines Clause is therefore incorporated by the Due Process Clause of the Fourteenth Amendment.

I

A

When ratified in 1791, the Bill of Rights applied only to the Federal Government. Barron ex rel. Tiernan v. Mayor of Baltimore , 7 Pet. 243, 8 L.Ed. 672 (1833). "The constitutional Amendments adopted in the aftermath of the Civil War," however, "fundamentally altered our country's federal system." McDonald , 561 U.S., at 754, 130 S.Ct. 3020. With only "a handful" of exceptions, this Court has held that the Fourteenth Amendment's Due Process Clause incorporates the protections contained in the Bill of Rights, rendering them applicable to the States. Id. , at 764-765, and nn. 12-13, 130 S.Ct. 3020. A Bill of Rights protection is incorporated, we have explained, if it is "fundamental to our scheme of ordered liberty," or "deeply rooted in this Nation's history and tradition." Id., at 767, 130 S.Ct. 3020 (internal quotation marks omitted; emphasis deleted).

Incorporated Bill of Rights guarantees are "enforced against the States under the Fourteenth Amendment according to the same standards that protect those personal rights against federal encroachment." Id., at 765, 130 S.Ct. 3020 (internal quotation marks omitted). Thus, if a Bill of Rights protection is incorporated, there is no daylight between the federal and state conduct it prohibits or requires.1

B

Under the Eighth Amendment, "[e]xcessive bail shall not be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted." Taken together, these Clauses place "parallel limitations" on "the power of those entrusted with the criminal-law function of government." Browning-Ferris Industries of Vt., Inc. v. Kelco Disposal, Inc. , 492 U.S. 257, 263, 109 S.Ct. 2909, 106 L.Ed.2d 219 (1989) (quoting Ingraham v. Wright , 430 U.S. 651, 664, 97 S.Ct. 1401, 51 L.Ed.2d 711 (1977) ). Directly at issue here is the phrase "nor excessive fines imposed," which "limits the government's power to extract payments, whether in cash or in kind, 'as punishment for some offense.' " United States v. Bajakajian , 524 U.S. 321, 327-328, 118 S.Ct. 2028, 141 L.Ed.2d 314 (1998) (quoting Austin v. United States , 509 U.S. 602, 609-610, 113 S.Ct. 2801, 125 L.Ed.2d 488 (1993) ). The Fourteenth Amendment, we hold, incorporates this protection.

The Excessive Fines Clause traces its venerable lineage back to at least 1215, when Magna Carta guaranteed that "[a] Free-man shall not be amerced for a small fault, but after the manner of the fault; and for a great fault after the greatness thereof, saving to him his contenement ...." § 20, 9 Hen. III, ch. 14, in 1 Eng.

*688Stat. at Large 5 (1225).2 As relevant here, Magna Carta required that economic sanctions "be proportioned to the wrong" and "not be so large as to deprive [an offender] of his livelihood." Browning-Ferris , 492 U.S., at 271, 109 S.Ct. 2909. See also 4 W. Blackstone, Commentaries on the Laws of England 372 (1769) ("[N]o man shall have a larger amercement imposed upon him, than his circumstances or personal estate will bear ...."). But cf. Bajakajian , 524 U.S., at 340, n. 15, 118 S.Ct. 2028 (taking no position on the question whether a person's income and wealth are relevant considerations in judging the excessiveness of a fine).

Despite Magna Carta, imposition of excessive fines persisted. The 17th century Stuart kings, in particular, were criticized for using large fines to raise revenue, harass their political foes, and indefinitely detain those unable to pay. E.g. , The Grand Remonstrance ¶¶17, 34 (1641), in The Constitutional Documents of the Puritan Revolution 1625-1660, pp. 210, 212 (S. Gardiner ed., 3d ed. rev. 1906); Browning-Ferris , 492 U.S., at 267, 109 S.Ct. 2909. When James II was overthrown in the Glorious Revolution, the attendant English Bill of Rights reaffirmed Magna Carta's guarantee by providing that "excessive Bail ought not to be required, nor excessive Fines imposed; nor cruel and unusual Punishments inflicted." 1 Wm. & Mary, ch. 2, § 10, in 3 Eng. Stat. at Large 441 (1689).

Across the Atlantic, this familiar language was adopted almost verbatim, first in the Virginia Declaration of Rights, then in the Eighth Amendment, which states: "Excessive bail shall not be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted."

Adoption of the Excessive Fines Clause was in tune not only with English law; the Clause resonated as well with similar colonial-era provisions. See, e.g. , Pa. Frame of Govt., Laws Agreed Upon in England, Art. XVIII (1682), in 5 Federal and State Constitutions 3061 (F. Thorpe ed. 1909) ("[A]ll fines shall be moderate, and saving men's contenements, merchandize, or wainage."). In 1787, the constitutions of eight States-accounting for 70% of the U.S. population-forbade excessive fines. Calabresi, Agudo, & Dore, State Bills of Rights in 1787 and 1791, 85 S. Cal. L. Rev. 1451, 1517 (2012).

An even broader consensus obtained in 1868 upon ratification of the Fourteenth Amendment. By then, the constitutions of 35 of the 37 States-accounting for over 90% of the U.S. population-expressly prohibited excessive fines. Calabresi & Agudo, Individual Rights Under State Constitutions When the Fourteenth Amendment Was Ratified in 1868, 87 Texas L. Rev. 7, 82 (2008).

Notwithstanding the States' apparent agreement that the right guaranteed by the Excessive Fines Clause was fundamental, abuses continued. Following the Civil War, Southern States enacted Black Codes to subjugate newly freed slaves and maintain the prewar racial hierarchy. Among these laws' provisions were draconian fines for violating broad proscriptions on "vagrancy" and other dubious offenses. See, e.g. , Mississippi Vagrant Law, Laws of Miss. § 2 (1865), in 1 W. Fleming, Documentary *689History of Reconstruction 283-285 (1950). When newly freed slaves were unable to pay imposed fines, States often demanded involuntary labor instead. E.g. , id. § 5; see Finkelman, John Bingham and the Background to the Fourteenth Amendment, 36 Akron L. Rev 671, 681-685 (2003) (describing Black Codes' use of fines and other methods to "replicate, as much as possible, a system of involuntary servitude"). Congressional debates over the Civil Rights Act of 1866, the joint resolution that became the Fourteenth Amendment, and similar measures repeatedly mentioned the use of fines to coerce involuntary labor. See, e.g. , Cong. Globe, 39th Cong., 1st Sess., 443 (1866); id., at 1123-1124.

Today, acknowledgment of the right's fundamental nature remains widespread. As Indiana itself reports, all 50 States have a constitutional provision prohibiting the imposition of excessive fines either directly or by requiring proportionality. Brief in Opposition 8-9. Indeed, Indiana explains that its own Supreme Court has held that the Indiana Constitution should be interpreted to impose the same restrictions as the Eighth Amendment. Id. , at 9 (citing Norris v. State , 271 Ind. 568, 576, 394 N.E.2d 144, 150 (1979) ).

For good reason, the protection against excessive fines has been a constant shield throughout Anglo-American history: Exorbitant tolls undermine other constitutional liberties. Excessive fines can be used, for example, to retaliate against or chill the speech of political enemies, as the Stuarts' critics learned several centuries ago. See Browning-Ferris , 492 U.S., at 267, 109 S.Ct. 2909. Even absent a political motive, fines may be employed "in a measure out of accord with the penal goals of retribution and deterrence," for "fines are a source of revenue," while other forms of punishment "cost a State money." Harmelin v. Michigan , 501 U.S. 957, 979, n. 9, 111 S.Ct. 2680, 115 L.Ed.2d 836 (1991) (opinion of Scalia, J.) ("it makes sense to scrutinize governmental action more closely when the State stands to benefit"). This concern is scarcely hypothetical. See Brief for American Civil Liberties Union et al. as Amici Curiae 7 ("Perhaps because they are politically easier to impose than generally applicable taxes, state and local governments nationwide increasingly depend heavily on fines and fees as a source of general revenue.").

In short, the historical and logical case for concluding that the Fourteenth Amendment incorporates the Excessive Fines Clause is overwhelming. Protection against excessive punitive economic sanctions secured by the Clause is, to repeat, both "fundamental to our scheme of ordered liberty" and "deeply rooted in this Nation's history and tradition." McDonald , 561 U.S., at 767, 130 S.Ct. 3020 (internal quotation marks omitted; emphasis deleted).

II

The State of Indiana does not meaningfully challenge the case for incorporating the Excessive Fines Clause as a general matter. Instead, the State argues that the Clause does not apply to its use of civil in rem forfeitures because, the State says, the Clause's specific application to such forfeitures is neither fundamental nor deeply rooted.

In Austin v. United States , 509 U.S. 602, 113 S.Ct. 2801, 125 L.Ed.2d 488 (1993), however, this Court held that civil in rem forfeitures fall within the Clause's protection when they are at least partially punitive. Austin arose in the federal context. But when a Bill of Rights protection is incorporated, the protection applies "identically to both the Federal Government and the States."

*690McDonald , 561 U.S., at 766, n. 14, 130 S.Ct. 3020. Accordingly, to prevail, Indiana must persuade us either to overrule our decision in Austin or to hold that, in light of Austin , the Excessive Fines Clause is not incorporated because the Clause's application to civil in rem forfeitures is neither fundamental nor deeply rooted. The first argument is not properly before us, and the second misapprehends the nature of our incorporation inquiry.

A

In the Indiana Supreme Court, the State argued that forfeiture of Timbs's SUV would not be excessive. See Brief in Opposition 5. It never argued, however, that civil in rem forfeitures were categorically beyond the reach of the Excessive Fines Clause. The Indiana Supreme Court, for its part, held that the Clause did not apply to the States at all, and it nowhere addressed the Clause's application to civil in rem forfeitures. See 84 N.E.3d 1179. Accordingly, Timbs sought our review of the question "[w]hether the Eighth Amendment's Excessive Fines Clause is incorporated against the States under the Fourteenth Amendment." Pet. for Cert. i. In opposing review, Indiana attempted to reformulate the question to ask "[w]hether the Eighth Amendment's Excessive Fines Clause restricts States' use of civil asset forfeitures." Brief in Opposition i. And on the merits, Indiana has argued not only that the Clause is not incorporated, but also that Austin was wrongly decided. Respondents' "right, in their brief in opposition, to restate the questions presented," however, "does not give them the power to expand [those] questions." Bray v. Alexandria Women's Health Clinic , 506 U.S. 263, 279, n. 10, 113 S.Ct. 753, 122 L.Ed.2d 34 (1993) (emphasis deleted). That is particularly the case where, as here, a respondent's reformulation would lead us to address a question neither pressed nor passed upon below. Cf. Cutter v. Wilkinson , 544 U.S. 709, 718, n. 7, 125 S.Ct. 2113, 161 L.Ed.2d 1020 (2005) ("[W]e are a court of review, not of first view ...."). We thus decline the State's invitation to reconsider our unanimous judgment in Austin that civil in rem forfeitures are fines for purposes of the Eighth Amendment when they are at least partially punitive.

B

As a fallback, Indiana argues that the Excessive Fines Clause cannot be incorporated if it applies to civil in rem forfeitures. We disagree. In considering whether the Fourteenth Amendment incorporates a protection contained in the Bill of Rights, we ask whether the right guaranteed-not each and every particular application of that right-is fundamental or deeply rooted.

Indiana's suggestion to the contrary is inconsistent with the approach we have taken in cases concerning novel applications of rights already deemed incorporated. For example, in Packingham v. North Carolina , 582 U.S. ----, 137 S.Ct. 1730, 198 L.Ed.2d 273 (2017), we held that a North Carolina statute prohibiting registered sex offenders from accessing certain commonplace social media websites violated the First Amendment right to freedom of speech. In reaching this conclusion, we noted that the First Amendment's Free Speech Clause was "applicable to the States under the Due Process Clause of the Fourteenth Amendment." Id., at ----, 137 S.Ct., at 1733. We did not, however, inquire whether the Free Speech Clause's application specifically to social media websites was fundamental or deeply rooted. See also, e.g. , Riley v. California , 573 U.S. 373, 134 S.Ct. 2473, 189 L.Ed.2d 430 (2014) (holding, without separately considering incorporation, that States' warrantless *691search of digital information stored on cell phones ordinarily violates the Fourth Amendment). Similarly here, regardless of whether application of the Excessive Fines Clause to civil in rem forfeitures is itself fundamental or deeply rooted, our conclusion that the Clause is incorporated remains unchanged.

* * *

For the reasons stated, the judgment of the Indiana Supreme Court is vacated, and the case is remanded for further proceedings not inconsistent with this opinion.

It is so ordered.

Justice GORSUCH, concurring.

The majority faithfully applies our precedent and, based on a wealth of historical evidence, concludes that the Fourteenth Amendment incorporates the Eighth Amendment's Excessive Fines Clause against the States. I agree with that conclusion. As an original matter, I acknowledge, the appropriate vehicle for incorporation may well be the Fourteenth Amendment's Privileges or Immunities Clause, rather than, as this Court has long assumed, the Due Process Clause. See, e.g. , post , at 691 - 692 (THOMAS, J., concurring in judgment); McDonald v. Chicago , 561 U.S. 742, 805-858, 130 S.Ct. 3020, 177 L.Ed.2d 894 (2010) (THOMAS, J., concurring in part and concurring in judgment) (documenting evidence that the "privileges or immunities of citizens of the United States" include, at minimum, the individual rights enumerated in the Bill of Rights); Wildenthal, Nationalizing the Bill of Rights: Revisiting the Original Understanding of the Fourteenth Amendment in 1866-67, 68 Ohio St. L.J. 1509 (2007) ; A. Amar, The Bill of Rights: Creation and Reconstruction 163-214 (1998); M. Curtis, No State Shall Abridge: The Fourteenth Amendment and the Bill of Rights (1986). But nothing in this case turns on that question, and, regardless of the precise vehicle, there can be no serious doubt that the Fourteenth Amendment requires the States to respect the freedom from excessive fines enshrined in the Eighth Amendment.

Justice THOMAS, concurring in the judgment.

I agree with the Court that the Fourteenth Amendment makes the Eighth Amendment's prohibition on excessive fines fully applicable to the States. But I cannot agree with the route the Court takes to reach this conclusion. Instead of reading the Fourteenth Amendment's Due Process Clause to encompass a substantive right that has nothing to do with "process," I would hold that the right to be free from excessive fines is one of the "privileges or immunities of citizens of the United States" protected by the Fourteenth Amendment.

I

The Fourteenth Amendment provides that "[n]o State shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States." "On its face, this appears to grant ... United States citizens a certain collection of rights-i.e. , privileges or immunities-attributable to that status." McDonald v. Chicago , 561 U.S. 742, 808, 130 S.Ct. 3020, 177 L.Ed.2d 894 (2010) (THOMAS, J., concurring in part and concurring in judgment). But as I have previously explained, this Court "marginaliz[ed]" the Privileges or Immunities Clause in the late 19th century by defining the collection of rights covered by the Clause "quite narrowly." Id. , at 808-809, 130 S.Ct. 3020. Litigants seeking federal protection of substantive rights against the States thus needed "an alternative fount of such rights," and this Court "found one in a *692most curious place," id. , at 809, 130 S.Ct. 3020 -the Fourteenth Amendment's Due Process Clause, which prohibits "any State" from "depriv[ing] any person of life, liberty, or property, without due process of law."

Because this Clause speaks only to "process," the Court has "long struggled to define" what substantive rights it protects. McDonald , supra , at 810, 130 S.Ct. 3020 (opinion of THOMAS, J.). The Court ordinarily says, as it does today, that the Clause protects rights that are "fundamental." Ante , at 686 - 687, 687 - 688, 689 - 690, 690 - 691. Sometimes that means rights that are " 'deeply rooted in this Nation's history and tradition.' " Ante , at 687 - 688, 690 - 691 (quoting McDonald , supra , at 767, 130 S.Ct. 3020 (majority opinion)). Other times, when that formulation proves too restrictive, the Court defines the universe of "fundamental" rights so broadly as to border on meaningless. See, e.g. , Obergefell v. Hodges , 576 U.S. ----, ---- - ----, 135 S.Ct. 2584, 2593, 192 L.Ed.2d 609 (2015) ("rights that allow persons, within a lawful realm, to define and express their identity"); Planned Parenthood of Southeastern Pa. v. Casey , 505 U.S. 833, 851, 112 S.Ct. 2791, 120 L.Ed.2d 674 (1992) ("At the heart of liberty is the right to define one's own concept of existence, of meaning, of the universe, and of the mystery of human life"). Because the oxymoronic "substantive" "due process" doctrine has no basis in the Constitution, it is unsurprising that the Court has been unable to adhere to any "guiding principle to distinguish 'fundamental' rights that warrant protection from nonfundamental rights that do not." McDonald , supra , at 811, 130 S.Ct. 3020 (opinion of THOMAS, J.). And because the Court's substantive due process precedents allow the Court to fashion fundamental rights without any textual constraints, it is equally unsurprising that among these precedents are some of the Court's most notoriously incorrect decisions. E.g. , Roe v. Wade , 410 U.S. 113, 93 S.Ct. 705, 35 L.Ed.2d 147 (1973) ; Dred Scott v. Sandford , 19 How. 393, 450, 15 L.Ed. 691 (1857).

The present case illustrates the incongruity of the Court's due process approach to incorporating fundamental rights against the States. Petitioner argues that the forfeiture of his vehicle is an excessive punishment. He does not argue that the Indiana courts failed to " 'proceed according to the "law of the land"-that is, according to written constitutional and statutory provisions,' " or that the State failed to provide "some baseline procedures." Nelson v. Colorado , 581 U.S. ----, ----, n. 1, 137 S.Ct. 1249, 1264, n. 1, 197 L.Ed.2d 611 (2017) (THOMAS, J., dissenting). His claim has nothing to do with any "process" "due" him. I therefore decline to apply the "legal fiction" of substantive due process. McDonald , 561 U.S., at 811, 130 S.Ct. 3020 (opinion of THOMAS, J.).

II

When the Fourteenth Amendment was ratified, "the terms 'privileges' and 'immunities' had an established meaning as synonyms for 'rights.' " Id. , at 813, 130 S.Ct. 3020. Those "rights" were the "inalienable rights" of citizens that had been "long recognized," and "the ratifying public understood the Privileges or Immunities Clause to protect constitutionally enumerated rights" against interference by the States. Id. , at 822, 837, 130 S.Ct. 3020. Many of these rights had been adopted from English law into colonial charters, then state constitutions and bills of rights, and finally the Constitution. "Consistent with their English heritage, the founding generation generally did not consider many of the rights identified in [the Bill of Rights] as new entitlements, but as inalienable rights of all men, given legal *693effect by their codification in the Constitution's text." Id. , at 818, 130 S.Ct. 3020.

The question here is whether the Eighth Amendment's prohibition on excessive fines was considered such a right. The historical record overwhelmingly demonstrates that it was.

A

The Excessive Fines Clause "was taken verbatim from the English Bill of Rights of 1689," United States v. Bajakajian , 524 U.S. 321, 335, 118 S.Ct. 2028, 141 L.Ed.2d 314 (1998), which itself formalized a longstanding English prohibition on disproportionate fines. The Charter of Liberties of Henry I, issued in 1101, stated that "[i]f any of my barons or men shall have committed an offence he shall not give security to the extent of forfeiture of his money, as he did in the time of my father, or of my brother, but according to the measure of the offence so shall he pay ...." Sources of English Legal and Constitutional History ¶8, p. 50 (M. Evans & R. Jack eds. 1984) (emphasis added). Expanding this principle, Magna Carta required that "amercements (the medieval predecessors of fines) should be proportioned to the offense and that they should not deprive a wrongdoer of his livelihood," Bajakajian , supra , at 335, 118 S.Ct. 2028 :

"A free man shall be amerced for a small fault only according to the measure thereof, and for a great crime according to its magnitude, saving his position; and in like manner, a merchant saving his trade, and a villein saving his tillage, if they should fall under Our mercy." Magna Carta, ch. 20 (1215), in A. Howard, Magna Carta: Text & Commentary 42 (rev. ed. 1998).

Similar clauses levying amercements "only in proportion to the measure of the offense" applied to earls, barons, and clergymen. Chs. 21-22, ibid. One historian posits that, due to the prevalence of amercements and their use in increasing the English treasury, "[v]ery likely there was no clause in Magna Carta more grateful to the mass of the people than that about amercements." Pleas of the Crown for the County of Gloucester xxxiv (F. Maitland ed. 1884). The principle was reiterated in the First Statute of Westminster, which provided that no man should "be amerced, without reasonable cause, and according to the quantity of his Trespass." 3 Edw. I, ch. 6 (1275). The English courts have long enforced this principle. In one early case, for example, the King commanded the bailiff "to take a moderate amercement proper to the magnitude and manner of th[e] offense, according to the tenour of the Great Charter of the Liberties of England," and the bailiff was sued for extorting "a heavier ransom." Le Gras v. Bailiff of Bishop of Winchester , Y.B. Mich. 10 Edw. II, pl. 4 (1316), reprinted in 52 Selden Society 3, 5 (1934); see also Richard Godfrey's Case , 11 Co. Rep. 42a, 44a, 77 Eng. Rep. 1199, 1202 (1615) (excessive fines are "against law").

During the reign of the Stuarts in the period leading up to the Glorious Revolution of 1688-1689, fines were a flashpoint "in the constitutional and political struggles between the king and his parliamentary critics." L. Schwoerer, The Declaration of Rights, 1689, p. 91 (1981) (Schwoerer). From 1629 to 1640, Charles I attempted to govern without convening Parliament, but "in the absence of parliamentary grants," he needed other ways of raising revenue. 4 H. Walter, A History of England 135 (1834); see 1 T. Macaulay, History of England 85 (1899). He thus turned "to exactions, some odious and obsolete, some of very questionable legality, and others clearly against law." 1 H. Hallam, Constitutional History of England: From the Accession of Henry VII to the Death of *694George II 462 (1827) (Hallam); see 4 Walter, supra , at 135.

The Court of Star Chamber, for instance, "imposed heavy fines on the king's enemies," Schwoerer 91, in disregard "of the provision of the Great Charter, that no man shall be amerced even to the full extent of his means...." 2 Hallam 46-47. "[T]he strong interest of th[is] court in these fines ... had a tendency to aggravate the punishment...." 1 id. , at 490. "The statute abolishing" the Star Chamber in 1641 "specifically prohibited any court thereafter from ... levying ... excessive fines." Schwoerer 91.

"But towards the end of Charles II's reign" in the 1670s and early 1680s, courts again "imposed ruinous fines on the critics of the crown." Ibid. In 1680, a committee of the House of Commons "examined the transcripts of all the fines imposed in King's Bench since 1677" and found that "the Court of King's Bench, in the Imposition of Fines on Offenders of late Years, hath acted arbitrarily, illegally, and partially; favouring Papists and Persons popishly affected; and excessively oppressing his Majesty's Protestant Subjects." Ibid. ; 9 Journals of the House of Commons 692 (Dec. 23, 1680). The House of Commons determined that the actions of the judges of the King's Bench, particularly the actions of Chief Justice William Scroggs, had been so contrary to law that it prepared articles of impeachment against him. The articles alleged that Scroggs had "most notoriously departed from all Rules of Justice and Equality, in the Imposition of Fines upon Persons convicted of Misdemeanors" without "any Regard to the Nature of the Offences, or the Ability of the Persons." Id. , at 698.

Yet "[o]ver the next few years fines became even more excessive and partisan." Schwoerer 91. The King's Bench, presided over by the infamous Chief Justice Jeffreys, fined Anglican cleric Titus Oates 2,000 marks (among other punishments) for perjury. Id. , at 93. For speaking against the Duke of York, the sheriff of London was fined £ 100,000 in 1682, which corresponds to well over $ 10 million in present-day dollars1 -"an amount, which, as it extended to the ruin of the criminal, was directly contrary to the spirit of [English] law." The History of England Under the House of Stuart, pt. 2, p. 801 (1840). The King's Bench fined Sir Samuel Barnadiston £ 10,000 for allegedly seditious letters, a fine that was overturned by the House of Lords as "exorbitant and excessive." 14 Journals of the House of Lords 210 (May 14, 1689). Several members of the committees that would draft the Declaration of Rights-which included the prohibition on excessive fines that was enacted into the English Bill of Rights of 1689-had themselves "suffered heavy fines." Schwoerer 91-92. And in 1684, judges in the case of John Hampden held that Magna Carta did not limit "fines for great offences" against the King, and imposed a £ 40,000 fine. Trial of Hampden , 9 State Trials 1054, 1125 (K. B. 1684); 1 J. Stephen, A History of the Criminal Law of England 490 (1883).

"Freedom from excessive fines" was considered "indisputably an ancient right of the subject," and the Declaration of Rights' indictment against James II "charged that during his reign judges had imposed excessive fines, thereby subverting the laws and liberties of the kingdom." Schwoerer 90. Article 10 of the Declaration declared "[t]hat excessive Bayle ought not *695to be required nor excessive fynes imposed nor cruel and unusuall Punishments inflicted." Id. , at 297.

Shortly after the English Bill of Rights was enacted, Parliament addressed several excessive fines imposed before the Glorious Revolution. For example, the House of Lords overturned a £ 30,000 fine against the Earl of Devonshire as "excessive and exorbitant, against Magna Charta, the common right of the subject, and against the law of the land." Case of Earl of Devonshire , 11 State Trials 1354, 1372 (K. B. 1687). Although the House of Lords refused to reverse the judgments against Titus Oates, a minority argued that his punishments were "contrary to Law and ancient Practice" and violated the prohibition on "excessive Fines." Harmelin v. Michigan , 501 U.S. 957, 971, 111 S.Ct. 2680, 115 L.Ed.2d 836 (1991) ; Trial of Oates , 10 State Trials 1080, 1325 (K. B. 1685). The House of Commons passed a bill to overturn Oates's conviction, and eventually, after a request from Parliament, the King pardoned Oates. Id. , at 1329-1330.

Writing a few years before our Constitution was adopted, Blackstone-"whose works constituted the preeminent authority on English law for the founding generation," Alden v. Maine , 527 U.S. 706, 715, 119 S.Ct. 2240, 144 L.Ed.2d 636 (1999) -explained that the prohibition on excessive fines contained in the English Bill of Rights "had a retrospect to some unprecedented proceedings in the court of king's bench." 4 W. Blackstone, Commentaries 372 (1769). Blackstone confirmed that this prohibition was "only declaratory ... of the old constitutional law of the land," which had long "regulated" the "discretion" of the courts in imposing fines. Ibid.

In sum, at the time of the founding, the prohibition on excessive fines was a longstanding right of Englishmen.

B

"As English subjects, the colonists considered themselves to be vested with the same fundamental rights as other Englishmen," McDonald , 561 U.S., at 816, 130 S.Ct. 3020 (opinion of THOMAS, J.), including the prohibition on excessive fines. E.g. , J. Dummer, A Defence of the New-England Charters 16-17 (1721) ("The Subjects Abroad claim the Privilege of Magna Charta , which says that no Man shall be fin'd above the Nature of his Offence, and whatever his Miscarriage be, a Salvo Contenemento suo is to be observ'd by the Judge"). Thus, the text of the Eighth Amendment was " 'based directly on ... the Virginia Declaration of Rights,' which 'adopted verbatim the language of the English Bill of Rights.' " Browning-Ferris Industries of Vt., Inc. v. Kelco Disposal, Inc. , 492 U.S. 257, 266, 109 S.Ct. 2909, 106 L.Ed.2d 219 (1989) (quoting Solem v. Helm , 463 U.S. 277, 285, n. 10, 103 S.Ct. 3001, 77 L.Ed.2d 637 (1983) ); see Jones v. Commonwealth , 5 Va. 555, 557 (1799) (opinion of Carrington, J.) (explaining that the clause in the Virginia Declaration of Rights embodied the traditional legal understanding that any "fine or amercement ought to be according to the degree of the fault and the estate of the defendant").

When the States were considering whether to ratify the Constitution, advocates for a separate bill of rights emphasized the need for an explicit prohibition on excessive fines mirroring the English prohibition. In colonial times, fines were "the drudge-horse of criminal justice," "probably the most common form of punishment." L. Friedman, Crime and Punishment in American History 38 (1993). To some, this fact made a constitutional prohibition on excessive fines all the more important. As the well-known Anti-Federalist Brutus argued in an essay, a prohibition *696on excessive fines was essential to "the security of liberty" and was "as necessary under the general government as under that of the individual states; for the power of the former is as complete to the purpose of requiring bail, imposing fines, inflicting punishments, ... and seizing ... property ... as the other." Brutus II (Nov. 1, 1787), in The Complete Bill of Rights 621 (N. Cogan ed. 1997). Similarly, during Virginia's ratifying convention, Patrick Henry pointed to Virginia's own prohibition on excessive fines and said that it would "depart from the genius of your country" for the Federal Constitution to omit a similar prohibition. Debate on Virginia Convention (June 14, 1788), in 3 Debates on the Federal Constitution 447 (J. Elliot 2d ed. 1854). Henry continued: "[W]hen we come to punishments, no latitude ought to be left, nor dependence put on the virtue of representatives" to "define punishments without this control." Ibid.

Governor Edmund Randolph responded to Henry, arguing that Virginia's charter was "nothing more than an investiture, in the hands of the Virginia citizens, of those rights which belonged to British subjects." Id. , at 466. According to Randolph, "the exclusion of excessive bail and fines ... would follow of itself without a bill of rights," for such fines would never be imposed absent "corruption in the House of Representatives, Senate, and President," or judges acting "contrary to justice." Id. , at 467-468.

For all the debate about whether an explicit prohibition on excessive fines was necessary in the Federal Constitution, all agreed that the prohibition on excessive fines was a well-established and fundamental right of citizenship. When the Excessive Fines Clause was eventually considered by Congress, it received hardly any discussion before "it was agreed to by a considerable majority." 1 Annals of Cong. 754 (1789). And when the Bill of Rights was ratified, most of the States had a prohibition on excessive fines in their constitutions.2

Early commentary on the Clause confirms the widespread agreement about the fundamental nature of the prohibition on excessive fines. Justice Story, writing a few decades before the ratification of the Fourteenth Amendment, explained that the Eighth Amendment was "adopted, as an admonition to all departments of the national government, to warn them against such violent proceedings, as had taken place in England in the arbitrary reigns of some of the Stuarts," when "[e]normous fines and amercements were ... sometimes imposed." 3 J. Story, Commentaries on the Constitution of the United States § 1896, pp. 750-751 (1833). Story included the prohibition on excessive fines as a right, along with the "right to bear arms" and others protected by the Bill of Rights, that "operates, as a qualification upon powers, actually granted by the people to the government"; without such a "restrict[ion]," the government's "exercise or *697abuse" of its power could be "dangerous to the people." Id. , § 1858, at 718-719.

Chancellor Kent likewise described the Eighth Amendment as part of the "right of personal security ... guarded by provisions which have been transcribed into the constitutions in this country from magna carta , and other fundamental acts of the English Parliament." 2 J. Kent, Commentaries on American Law 9 (1827). He understood the Eighth Amendment to "guard against abuse and oppression," and emphasized that "the constitutions of almost every state in the Unio[n] contain the same declarations in substance, and nearly in the same language." Ibid. Accordingly, "they must be regarded as fundamental doctrines in every state, for all the colonies were parties to the national declaration of rights in 1774, in which the ... rights and liberties of English subjects were peremptorily claimed as their undoubted inheritance and birthright."Ibid. ; accord, W. Rawle, A View of the Constitution of the United States of America 125 (1825) (describing the prohibition on excessive fines as "founded on the plainest principles of justice").

C

The prohibition on excessive fines remained fundamental at the time of the Fourteenth Amendment. In 1868, 35 of 37 state constitutions "expressly prohibited excessive fines." Ante , at 688. Nonetheless, as the Court notes, abuses of fines continued, especially through the Black Codes adopted in several States. Ante , at 688 - 689. The "centerpiece" of the Codes was their "attempt to stabilize the black work force and limit its economic options apart from plantation labor." E. Foner, Reconstruction: America's Unfinished Revolution 1863-1877, p. 199 (1988). Under the Codes, "the state would enforce labor agreements and plantation discipline, punish those who refused to contract, and prevent whites from competing among themselves for black workers." Ibid. The Codes also included " 'antienticement' measures punishing anyone offering higher wages to an employee already under contract." Id. , at 200.

The 39th Congress focused on these abuses during its debates over the Fourteenth Amendment, the Civil Rights Act of 1866, and the Freedmen's Bureau Act. During those well-publicized debates, Members of Congress consistently highlighted and lamented the "severe penalties" inflicted by the Black Codes and similar measures, Cong. Globe, 39th Cong., 1st Sess., 474 (1866) (Sen. Trumbull), suggesting that the prohibition on excessive fines was understood to be a basic right of citizenship.

For example, under Mississippi law, adult "freedmen, free negroes and mulattoes" "without lawful employment" faced $ 50 in fines and 10 days' imprisonment for vagrancy. Reports of Assistant Commissioners of Freedmen, and Synopsis of Laws on Persons of Color in Late Slave States, S. Exec. Doc. No. 6, 39th Cong., 2d Sess., § 2, p. 192 (1867). Those convicted had five days to pay or they would be arrested and leased to "any person who will, for the shortest period of service, pay said fine and forfeiture and all costs." § 5, ibid. Members of Congress criticized such laws "for selling [black] men into slavery in punishment of crimes of the slightest magnitude." Cong. Globe, 39th Cong., 1st Sess., 1123 (1866) (Rep. Cook); see id. , at 1124 ("It is idle to say these men will be protected by the States").

Similar examples abound. One congressman noted that Alabama's "aristocratic and anti-republican laws, almost reenacting slavery, among other harsh inflictions impose ... a fine of fifty dollars and six months' imprisonment on any servant or *698laborer (white or black) who loiters away his time or is stubborn or refractory." Id. , at 1621 (Rep. Myers). He also noted that Florida punished vagrants with "a fine not exceeding $ 500 and imprison[ment] for a term not exceeding twelve months, or by being sold for a term not exceeding twelve months, at the discretion of the court." Ibid. At the time, such fines would have been ruinous for laborers. Cf. id. , at 443 (Sen. Howe) ("A thousand dollars! That sells a negro for his life").

These and other examples of excessive fines from the historical record informed the Nation's consideration of the Fourteenth Amendment. Even those opposed to civil-rights legislation understood the Privileges or Immunities Clause to guarantee those "fundamental principles" "fixed" by the Constitution, including "immunity from ... excessive fines." 2 Cong. Rec. 384-385 (1874) (Rep. Mills); see also id. , at App. 241 (Sen. Norwood). And every post-1855 state constitution banned excessive fines. S. Calabresi & S. Agudo, Individual Rights Under State Constitutions When the Fourteenth Amendment Was Ratified in 1868, 87 Texas L. Rev. 7, 82 (2008). The attention given to abusive fines at the time of the Fourteenth Amendment, along with the ubiquity of state excessive-fines provisions, demonstrates that the public continued to understand the prohibition on excessive fines to be a fundamental right of American citizenship.

* * *

The right against excessive fines traces its lineage back in English law nearly a millennium, and from the founding of our country, it has been consistently recognized as a core right worthy of constitutional protection. As a constitutionally enumerated right understood to be a privilege of American citizenship, the Eighth Amendment's prohibition on excessive fines applies in full to the States.

Pee Curiam.

Petitioner in No. 69-5003 was convicted of murder in Georgia and was sentenced to death pursuant to Ga. Code Ann. §26-1005 (Supp. 1971) (effective prior to July 1, 1969). 225 Ga. 253, 167 S. E. 2d 628 (1969). Petitioner in No. 69-5030 was convicted of rape in Georgia and was sentenced to death pursuant to Ga. Code Ann. § 26-1302 (Supp. 1971) (effective prior to July 1, 1969). 225 Ga. 790, 171 S. E. 2d 501 (1969). Petitioner in No. 69-5031 was convicted of rape in Texas and was sentenced to death pursuant to Tex. Penal Code, Art. 1189 (1961). 447 S. W. 2d 932 (Ct. Crim. App. 1969). Certiorari was granted limited to the following question: “Does the imposition and carrying out of the death penalty in [these cases] constitute cruel and unusual punishment in violation of the Eighth and Fourteenth Amendments?” 403 U. S. 952 (1971). The Court holds that the imposi*240tion and carrying out of the death penalty in these cases constitute cruel and unusual punishment in violation of the Eighth and Fourteenth Amendments. The judgment in each case is therefore reversed insofar as it leaves undisturbed the death sentence imposed, and the cases are remanded for further proceedings.

So ordered.

Me. Justice Douglas, Mr. Justice Brennan, Mr. Justice Stewart, Mr. Justice White, and Mr. Justice Marshall have filed separate opinions in support of the judgments. The Chief Justice, Mr. Justice Black-mun, Mr. Justice Powell, and Mr. Justice Rehnquist have filed separate dissenting opinions.

Mr. Justice Douglas,

concurring.

In these three cases the death penalty was imposed, one of them for murder, and two for rape. In each the determination of whether the penalty should be death or a lighter punishment was left by the State to the discretion of the judge or of the jury. In each of the three cases the trial was to a jury. They are here on petitions for certiorari which we granted limited to the question whether the imposition and execution of the death penalty constitute “cruel and unusual punishment” within the meaning of the Eighth Amendment as applied to the States by the Fourteenth.1 I vote to vacate each judgment, believing that the exaction of the death penalty does violate the Eighth and Fourteenth Amendments.

*241That the requirements of due process ban cruel and unusual punishment is now settled. Louisiana ex rel. Francis v. Resweber, 329 U. S. 459, 463, and 473-474 (Burton, J., dissenting); Robinson v. California, 370 U. S. 660, 667. It is also settled that the proscription of cruel and unusual punishments forbids the judicial imposition of them as well as their imposition by the legislature. Weems v. United States, 217 U. S. 349, 378-382.

Congressman Bingham, in proposing the Fourteenth Amendment, maintained that “the privileges or immunities of citizens of the United States” as protected by the Fourteenth Amendment included protection against “cruel and unusual punishments:”

“[M]any instances of State injustice and oppression have already occurred in the State legislation of this Union, of flagrant violations of the guarantied privileges of citizens of the United States, for which the national Government furnished and could furnish by law no remedy whatever. Contrary to the express letter of your Constitution, ‘cruel and unusual punishments’ have been inflicted under State laws within this Union upon citizens, not only for crimes committed, but for sacred duty done, for which and against which the Government of the United States had provided no remedy and could provide none.” Cong. Globe, 39th Cong., 1st Sess., 2542.

Whether the privileges and immunities route is followed, or the due process route, the result is the same.

It has been assumed in our decisions that punishment by death is not cruel, unless the manner of execution can be said to be inhuman and barbarous. In re Kemmler, 136 U. S. 436, 447. It is also said in our opinions *242that the proscription of cruel and unusual punishments “is not fastened to the obsolete but may acquire meaning as public opinion becomes enlightened by a humane justice.” Weems v. United States, supra, at 378. A like statement was made in Trop v. Dulles, 356 U. S. 86, 101, that the Eighth Amendment “must draw its meaning from the evolving standards of decency that mark the progress of a maturing society.”

The generality of a law inflicting capital punishment is one thing. What may be said of the validity of a law on the books and what may be done with the law in its application do, or may, lead to quite different conclusions.

It would seem to be incontestable that the death penalty inflicted on one defendant is “unusual” if it discriminates against him by reason of his race, religion, wealth, social position, or class, or if it is imposed under a procedure that gives room for the play of such prejudices.

There is evidence that the provision of the English Bill of Rights of 1689, from which the language of the Eighth Amendment was taken, was concerned primarily with selective or irregular application of harsh penalties and that its aim was to forbid arbitrary and discriminatory penalties of a severe nature: 2

“Following the Norman conquest of England in 1066, the old system of penalties, which ensured equality between crime and punishment, suddenly disappeared. By the time systematic judicial records were kept, its demise was almost complete. With the exception of certain grave crimes for which the punishment was death or outlawry, the arbitrary fine was replaced by a discretionary *243amercement. Although amercement’s discretionary character allowed the circumstances of each case to be taken into account and the level of cash penalties to be decreased or increased accordingly, the amercement presented an opportunity for excessive or oppressive fines.

“The problem of excessive amercements became so prevalent that three chapters of the Magna Carta were devoted to their regulation. Maitland said of Chapter 14 that ‘very likely there was no clause in the Magna Carta more grateful to the mass of the people.’ Chapter 14 clearly stipulated as fundamental law a prohibition of excessiveness in punishments:

“ ‘A free man shall not be amerced for a trivial offence, except in accordance with the degree of the offence; and for a serious offence he shall be amerced according to its gravity, saving his livelihood; and a merchant likewise, saving his merchandise; in the same way a villein shall be amerced saving his wainage; if they fall into our mercy. And none of the aforesaid amercements shall be imposed except by the testimony of reputable men of the neighborhood.’ ”

The English Bill of Rights, enacted December 16,1689, stated that “excessive bail ought not to be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted.” 3 These were the words chosen for our Eighth Amendment. A like provision had been in Virginia’s Constitution of 17764 and in the constitutions *244of seven other States.5 The Northwest Ordinance, enacted under the Articles of Confederation, included a prohibition of cruel and unusual punishments.6 But the debates of the First Congress on the Bill of Rights throw little light on its intended meaning. All that appears is the following: 7

“Mr. Smith, of South Carolina, objected to the words ‘nor cruel and unusual punishments;’ the import of them being too indefinite.

“Mr. Livermore: The clause seems to express a great deal of humanity, on which account I have no objection to it; but as it seems to have no meaning in it, I do not think it necessary. What is meant by the terms excessive bail? Who are to be the judges? What is understood by excessive fines? It lies with the court to determine. No cruel and unusual punishment is to be inflicted; it is sometimes necessary to hang a man, villains often deserve whipping, and perhaps having their ears cut off; but are we in future to be prevented from inflicting these punishments because they are cruel? If a more lenient mode of correcting vice and deterring others from the commission of it could be invented, it would be very prudent in the Legislature to adopt it; but until we have some security that this will be done, we ought not to be restrained from making necessary laws by any declaration of this kind.”

The words “cruel and unusual” certainly include pen*245alties that are barbaric. But the words, at least when read in light of the English proscription against selective and irregular use of penalties, suggest that it is “cruel and unusual” to apply the death penalty — or any other penalty — selectively to minorities whose numbers are few, who are outcasts of society, and who are unpopular, but whom society is willing to see suffer though it would not countenance general application of the same penalty across the board.8 Judge Tuttle, indeed, made abundantly clear in Novak v. Beto, 453 F. 2d 661, 673-679 (CA5) (concurring in part and dissenting in part), that solitary confinement may at times be “cruel and unusual” punishment. Cf. Ex parte Medley, 134 U. S. 160; Brooks v. Florida, 389 U. S. 413.

The Court in McGautha v. California, 402 U. S. 183, 198, noted that in this country there was almost from the beginning a “rebellion against the common-law rule imposing a mandatory death sentence on all convicted *246murderers.” The first attempted remedy was to restrict the death penalty to defined offenses such as “premeditated” murder.9 Ibid. But juries “took the *247law into their own hands” and refused to convict on the capital offense. Id., at 199.

“In order to meet the problem of jury nullification, legislatures did not try, as before, to refine further the definition of capital homicides. Instead they adopted the method of forthrightly granting juries the discretion which they had been exercising in fact.” Ibid.

The Court concluded: “In light of history, experience, and the present limitations of human knowledge, we find it quite impossible to say that committing to the untrammeled discretion of the jury the power to pronounce life or death in capital cases is offensive to anything in the Constitution.” Id., at 207.

The Court refused to find constitutional dimensions in the argument that those who exercise their discretion to send a person to death should be given standards by which that discretion should be exercised. Id., at 207-208.

A recent witness at the Hearings before Subcommittee No. 3 of the House Committee on the Judiciary, 92d Cong., 2d Sess., Ernest van den Haag, testifying on H. R. 8414 et al.,10 stated:

“Any penalty, a fine, imprisonment or the death penalty could be unfairly or unjustly applied. The *248vice in this case is not in the penalty but in the process by which it is inflicted. It is unfair to inflict unequal penalties on equally guilty parties, or on any innocent parties, regardless of what the penalty is.” Id., at 116-117. (Emphasis supplied.)

But those who advance that argument overlook McGautha, supra.

We are now imprisoned in the McGautha holding. Indeed the seeds of the present cases are in McGautha. Juries (or judges, as the case may be) have practically untrammeled discretion to let an accused live or insist that he die.11

*249Mr. Justice Field, dissenting in O’Neil v. Vermont, 144 U. S. 323, 340, said, “The State may, indeed, make the drinking of one drop of liquor an offence to be punished by imprisonment, but it would be an unheard-of cruelty if it should count the drops in a single glass and make thereby a thousand offences, and thus extend the punishment for drinking the single glass of liquor to an imprisonment of almost indefinite duration.” What the legislature may not do for all classes uniformly and systematically, a judge or jury may not do for a class that prejudice sets apart from the community.

There is increasing recognition of the fact that the basic theme of equal protection is implicit in “cruel and unusual” punishments. “A penalty . . . should be considered 'unusually’ imposed if it is administered arbitrarily or discriminatorily.” 12 The same authors add that “[t]he extreme rarity with which applicable death penalty provisions are put to use raises a strong inference of arbitrariness.” 13 The President’s Commission on Law Enforcement and Administration of Justice recently concluded: 14

“Finally there is evidence that the imposition of the death sentence and the exercise of dispensing power by the courts and the executive follow discriminatory patterns. The death sentence is disproportionately imposed and carried out on the *250poor, the Negro, and the members of unpopular groups.”

A study of capital cases in Texas from 1924 to 1968 reached the following conclusions:15

“Application of the death penalty is unequal: most of those executed were poor, young, and ignorant.

*251“Seventy-five of the 460 cases involved co-defendants, who, under Texas law, were given separate trials. In several instances where a white and a Negro were co-defendants, the white was sentenced to life imprisonment or a term of years, and the Negro was given the death penalty.

“Another ethnic disparity is found in the type of sentence imposed for rape. The Negro convicted of rape is far more likely to get the death penalty than a term sentence, whereas whites and Latins are far more likely to get a term sentence than the death penalty.”

Warden Lewis E. Lawes of Sing Sing said:16

“Not only does capital punishment fail in its justification, but no punishment could be invented with so many inherent defects. It is an unequal punishment in the way it is applied to the rich and to the poor. The defendant of wealth and position never goes to the electric chair or to the gallows. Juries do not intentionally favour the rich, the law is theoretically impartial, but the defendant with ample means is able to have his case presented with every favourable aspect, while the poor defendant often has a lawyer assigned by the court. Sometimes such assignment is considered part of political patronage; usually the lawyer assigned has had no experience whatever in a capital case.”

Former Attorney General Ramsey Clark has said, “It is the poor, the sick, the ignorant, the powerless and the hated who are executed.” 17 One searches our chron*252icles in vain for the execution of any member of the affluent strata of this society. The Leopolds and Loebs are given prison terms, not sentenced to death.

Jackson, a black, convicted of the rape of a white woman, was 21 years old. A court-appointed psychiatrist said that Jackson was of average education and average intelligence, that he was not an imbecile, or schizophrenic, or psychotic, that his traits were the product of environmental influences, and that he was competent to stand trial. Jackson had entered the house after the husband left for work. He held scissors against the neck of the wife, demanding money. She could find none and a struggle ensued for the scissors, a battle which she lost; and she was then raped, Jackson keeping the scissors pressed against her neck. While there did not appear to be any long-term traumatic impact on the victim, she was bruised and abrased in the struggle but was not hospitalized. Jackson was a convict who had escaped from a work gang in the area, a result of a three-year sentence for auto theft. He was at large for three days and during that time had committed several other offenses — burglary, auto theft, and assault and battery.

Furman, a black, killed a householder while seeking to enter the home at night. Furman shot the deceased through a closed door. He was 26 years old and had finished the sixth grade in school. Pending trial, he was committed to the Georgia Central State Hospital for a psychiatric examination on his plea of insanity tendered by court-appointed counsel. The superintendent reported that a unanimous staff diagnostic conference had concluded “that this patient should retain his present diagnosis of Mental Deficiency, Mild to Moderate, with Psychotic Episodes associated with Convulsive Disorder.” The physicians agreed that “at present the patient is not psychotic, but he is not capable of cooperating with his counsel in the preparation of his *253defense”; and the staff believed “that he is in need of further psychiatric hospitalization and treatment.”

Later, the superintendent reported that the staff diagnosis was Mental Deficiency, Mild to Moderate, with Psychotic Episodes associated with Convulsive Disorder. He concluded, however, that Furman was “not psychotic at present, knows right from wrong and is able to cooperate with his counsel in preparing his defense.”

Branch, a black, entered the rural home of a 65-year-old widow, a white, while she slept and raped her, holding his arm against her throat. Thereupon he demanded money and for 30 minutes or more the widow searched for money, finding little. As he left, Jackson said if the widow told anyone what happened, he would return and kill her. The record is barren of any medical or psychiatric evidence showing injury to her as a result of Branch’s attack.

He had previously been convicted of felony theft and found to be a borderline mental deficient and well below the average IQ of Texas prison inmates. He had the equivalent of five and a half years of grade school education. He had a “dull intelligence” and was in the lowest fourth percentile of his class.

We cannot say from facts disclosed in these records that these defendants were sentenced to death because they were black. Yet our task is not restricted to an effort to divine what motives impelled these death penalties. Rather, we deal with a system of law and of justice that leaves to the uncontrolled discretion of judges or juries the determination whether defendants committing these crimes should die or be imprisoned. Under these laws no standards govern the selection of the penalty. People live or die, dependent on the whim of one man or of 12.

Irving Brant has given a detailed account of the Bloody Assizes, the reign of terror that occupied the *254closing years of the rule of Charles II and the opening years of the regime of James II (the Lord Chief Justice was George Jeffreys):

“Nobody knows how many hundreds of men, innocent or of unproved guilt, Jeffreys sent to their deaths in the pseudo trials that followed Monmouth’s feeble and stupid attempt to seize the throne. When the ordeal ended, scores had been executed and 1,260 were awaiting the hangman in three counties. To be absent from home during the uprising was evidence of guilt. Mere death was considered much too mild for the villagers and farmers rounded up in these raids. The directions to a high sheriff were to provide an ax, a cleaver, 'a furnace or cauldron to boil their heads and quarters, and soil to boil therewith, half a bushel to each traitor, and tar to tar them with, and a sufficient number of spears and poles to fix their heads and quarters’ along the highways. One could have crossed a good part of northern England by their guidance.

“The story of The Bloody Assizes, widely known to Americans, helped to place constitutional limitations on the crime of treason and to produce a bar against cruel and unusual punishments. But in the polemics that led to the various guarantees of freedom, it had no place compared with the tremendous thrust of the trial and execution of Sidney. The hundreds of judicial murders committed by Jeffreys and his fellow judges were totally inconceivable in a free American republic, but any American could imagine himself in Sidney’s place— executed for putting on paper, in his closet, words that later on came to express the basic principles of republican government. Unless barred by fundamental law, the legal rulings that permitted this *255result could easily be employed against any person whose political opinions challenged the party in power.” The Bill of Rights 154-155 (1965).

Those who wrote the Eighth Amendment knew what price their forebears had paid for a system based, not on equal justice, but on discrimination. In those days the target was not the blacks or the poor, but the dissenters, those who opposed absolutism in government, who struggled for a parliamentary regime, and who opposed governments’ recurring efforts to foist a particular religion on the people. Id., at 155-163. But the tool of capital punishment was used with vengeance against the opposition and those unpopular with the regime. One cannot read this history without realizing that the desire for equality was reflected in the ban against “cruel and unusual punishments” contained in the Eighth Amendment.

In a Nation committed to equal protection of the laws there is no permissible “caste” aspect18 of law enforcement. Yet we know that the discretion of judges and juries in imposing the death penalty enables the penalty to be selectively applied, feeding prejudices against the accused if he is poor and despised, and lacking political clout, or if he is a member of a suspect or unpopular minority, and saving those who by social position may be in a more protected position. In ancient Hindu law a Brahman was exempt from capital punishment,19 and under that law, “[g] enerally, in the law books, punishment increased in severity as social status diminished.” 20 We have, I fear, taken in practice the same position, partially as a result of making the death pen*256alty discretionary and partially as a result of the ability of the rich to purchase the services of the most respected and most resourceful legal talent in the Nation.

The high service rendered by the “cruel and unusual” punishment clause of the Eighth Amendment is to require legislatures to write penal laws that are evenhanded, nonselective, and nonarbitrary, and to require judges to see to it that general laws are not applied sparsely, selectively, and spottily to unpopular groups.

A law that stated that anyone making more than $50,000 would be exempt from the death penalty would plainly fall, as would a law that in terms said that blacks, those who never went beyond the fifth grade in school, those who made less than $3,000 a year, or those who were unpopular or unstable should be the only people executed. A law which in the overall view reaches that result in practice 21 has no more sanctity than a law which in terms provides the same.

Thus, these discretionary statutes are unconstitutional *257in their operation. They are pregnant with discrimination and discrimination is an ingredient not compatible with the idea of equal protection of the laws that is implicit in the ban on “cruel and unusual” punishments.

Any law which is nondiscriminatory on its face may be applied in such a way as to violate the Equal Protection Clause of the Fourteenth Amendment. Yick Wo v. Hopkins, 118 U. S. 356. Such conceivably might be the fate of a mandatory death penalty, where equal or lesser sentences were imposed on the elite, a harsher one on the minorities or members of the lower castes. Whether a mandatory death penalty would otherwise be constitutional is a question I do not reach.

I concur in the judgments of the Court.

Mr. Justice Brennan,

concurring.

The question presented in these cases is whether death is today a punishment for crime that is “cruel and unusual” and consequently, by virtue of the Eighth and Fourteenth Amendments, beyond the power of the State to inflict.1

*258Almost a century ago, this Court observed that “[d]ifficulty would attend the effort to define with exactness the extent of the constitutional provision which provides that cruel and unusual punishments shall not be inflicted.” Wilkerson v. Utah, 99 U. S. 130, 135-136 (1879). Less than 15 years ago, it was again noted that “[t]he exact scope of the constitutional phrase 'cruel and unusual’ has not been detailed by this Court.” Trop v. Dulles, 356 U. S. 86, 99 (1958). Those statements remain true today. The Cruel and Unusual Punishments Clause, like the other great clauses of the Constitution, is not susceptible of precise definition. Yet we know that the values and ideals it embodies are basic to our scheme of government. And we know also that the Clause imposes upon this Court the duty, when the issue is properly presented, to determine the constitutional validity of a challenged punishment, whatever that punishment may be. In these cases, “[t]hat issue confronts us, and the task of resolving it is inescapably ours.” Id., at 103.

I

We have very little evidence of the Framers’ intent in including the Cruel and Unusual Punishments Clause among those restraints upon the new Government enumerated in the Bill of Rights. The absence of such a restraint from the body of the Constitution was alluded to, so far as we now know, in the debates of only two of the state ratifying conventions. In the Massachusetts convention, Mr. Holmes protested:

“What gives an additional glare of horror to these gloomy circumstances is the consideration, that Congress have to ascertain, point out, and deter*259mine, what kind of punishments shall be inflicted on persons convicted of crimes. They are nowhere restrained from inventing the most cruel and unheard-of punishments, and annexing them to crimes; and there is no constitutional check on them, but that racks and gibbets may be amongst the most mild instruments of their discipline.” 2 J. Elliot’s Debates 111 (2d ed. 1876).

Holmes’ fear that Congress would have unlimited power to prescribe punishments for crimes was echoed by Patrick Henry at the Virginia convention:

“. . . Congress, from their general powers, may fully go into business of human legislation. They may legislate, in criminal cases, from treason to the lowest offence — petty larceny. They may define crimes and prescribe punishments. In the definition of crimes, I trust they will be directed by what wise representatives ought to be governed by. But when we come to punishments, no latitude ought to be left, nor dependence put on the virtue of representatives. What says our [Virginia] bill of rights?— 'that excessive bail ought not to be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted.’ Are you not, therefore, now calling on those gentlemen who are to compose Congress, to ... define punishments without this control? Will they find sentiments there similar to this bill of rights? You let them loose; you do more — you depart from the genius of your country. . . .

“In this business of legislation, your members of Congress will loose the restriction of not imposing excessive fines, demanding excessive bail, and inflicting cruel and unusual punishments. These are prohibited by your [Virginia] declaration of rights. What has distinguished our ancestors?— *260That they would not admit of tortures, or cruel and barbarous punishment.” 3 id., at 447.2

These two statements shed some light on what the Framers meant by “cruel and unusual punishments.” Holmes referred to “the most cruel and unheard-of punishments,” Henry to “tortures, or cruel and barbarous punishment.” It does not follow, however, that the Framers were exclusively concerned with prohibiting torturous punishments. Holmes and Henry were objecting to the absence of a Bill of Rights, and they cited to support their objections the unrestrained legislative power to prescribe punishments for crimes. Certainly we may suppose that they invoked the specter of the most drastic punishments a legislature might devise.

In addition, it is quite clear that Holmes and Henry focused wholly upon the necessity to restrain the legislative power. Because they recognized “that Congress have to ascertain, point out, and determine, what kinds of punishments shall be inflicted on persons convicted of crimes,” they insisted that Congress must be limited in its power to punish. Accordingly, they *261called for a “constitutional check” that would ensure that “when we come to punishments, no latitude ought to be left, nor dependence put on the virtue of representatives.” 3

The only further evidence of the Framers’ intent appears from the debates in the First Congress on the adoption of the Bill of Rights.4 As the Court noted in Weems v. United States, 217 U. S. 349, 368 (1910), *262the Cruel and Unusual Punishments Clause “received very little debate.” The extent of the discussion, by two opponents of the Clause in the House of Representatives, was this:

“Mr. Smith, of South Carolina, objected to the words 'nor cruel and unusual punishments;’ the import of them being too indefinite.

“Mr. Livermore. — The [Eighth Amendment] seems to express a great deal of humanity, on which account I have no objection to it; but as it seems to have no meaning in it, I do not think it necessary. ... No cruel and unusual punishment is to be inflicted; it is sometimes necessary to hang a man, villains often deserve whipping, and perhaps having their ears cut off; but are we in future to be prevented from inflicting these punishments because they are cruel? If a more lenient mode of correcting vice and deterring others from the commission of it could be invented, it would be very prudent in the Legislature to adopt it; but until we have some security that this will be done, we ought not to be restrained from making necessary laws by any declaration of this kind.

“The question was put on the [Eighth Amendment], and it was agreed to by a considerable majority.” 1 Annals of Cong. 754 (1789).5

Livermore thus agreed with Holmes and Henry that the Cruel and Unusual Punishments Clause imposed a limitation upon the legislative power to prescribe pun*263ishments. However, in contrast to Holmes and Henry, who were supporting the Clause, Livermore, opposing it, did not refer to punishments that were considered barbarous and torturous. Instead, he objected that the Clause might someday prevent the legislature from inflicting what were then quite common and, in his view, “necessary” punishments — death, whipping, and earcrop-ping.6 The only inference to be drawn from Liver-more’s statement is that the “considerable majority” was prepared to run that risk. No member of the House rose to reply that the Clause was intended merely to prohibit torture.

Several conclusions thus emerge from the history of the adoption of the Clause. We know that the Framers’ concern was directed specifically at the exercise of legislative power. They included in the Bill of Rights a prohibition upon “cruel and unusual punishments” precisely because the legislature would otherwise have had the unfettered power to prescribe punishments for crimes. Yet we cannot now know exactly what the Framers thought “cruel and unusual punishments” were. Certainly they intended to ban torturous punishments, but the available evidence does not support the further conclusion that only torturous punishments were to be outlawed. As Livermore’s comments demonstrate, the Framers were well aware that the reach of the Clause was not limited to the proscription of unspeakable atrocities. Nor did they intend simply to forbid punishments considered “cruel and unusual” at the time. The “import” of the Clause is, indeed, “indefinite,” and for good reason. A constitutional provision “is enacted, it is true, from an experience of evils, but its general lan*264guage should not, therefore, be necessarily confined to the form that evil had theretofore taken. Time works changes, brings into existence new conditions and purposes. Therefore a principle to be vital must be capable of wider application than the mischief which gave it birth.” Weems v. United States, 217 U. S., at 373.

It was almost 80 years before this Court had occasion to refer to the Clause. See Pervear v. The Commonwealth, 5 Wall. 475, 479-480 (1867). These early cases, as the Court pointed out in Weems v. United States, supra, at 369, did not undertake to provide “an exhaustive definition” of “cruel and unusual punishments.” Most of them proceeded primarily by “looking backwards for examples by which to fix the meaning of the clause,” id., at 377, concluding simply that a punishment would be “cruel and unusual” if it were similar to punishments considered “cruel and unusual” at the time the Bill of Rights was adopted.7 In Wilkerson v. Utah, 99 U. S., at 136, for instance, the Court found it “safe to affirm that punishments of torture . . . and all others in the same line of unnecessary cruelty, are forbidden.” The “punishments of torture,” which the Court labeled “atrocities,” were cases where the criminal “was embowelled alive, beheaded, and quartered,” and cases “of public dissection . . . and burning alive.” Id., at 135. Similarly, in In re Kemm-*265ler, 136 U. S. 436, 446 (1890), the Court declared that “if the punishment prescribed for an offence against the laws of the State were manifestly cruel and unusual, as burning at the stake, crucifixion, breaking on the wheel, or the like, it would be the duty of the courts to adjudge such penalties to be within the constitutional prohibition.” The Court then observed, commenting upon the passage just quoted from Wilkerson v. Utah, supra, and applying the “manifestly cruel and unusual” test, that “[pjunishments are cruel when they involve torture or a lingering death; but the punishment of death is not cruel, within the meaning of that word as used in the Constitution. It implies there something inhuman and barbarous, something more than the mere extinguishment of life.” 136 U. S., at 447.

Had this “historical” interpretation of the Cruel and Unusual Punishments Clause prevailed, the Clause would have been effectively read out of the Bill of Rights. As the Court noted in Weems v. United States, supra, at 371, this interpretation led Story to conclude “that the provision ‘would seem to be wholly unnecessary in a free government, since it is scarcely possible that any department of such a government should authorize or justify such atrocious conduct.’ ” And Cooley in his book, Constitutional Limitations, said the Court, “apparently in a struggle between the effect to be given to ancient examples and the inconsequence of a dread of them in these enlightened times, . . . hesitate [d] to advance definite views.” Id., at 375. The result of a judicial application of this interpretation was not surprising. A state court, for example, upheld the constitutionality of the whipping post: “In comparison with the ‘barbarities of quartering, hanging in chains, castration, etc.,’ it was easily reduced to insignificance.” Id., at 377.

*266But this Court in Weems decisively repudiated the “historical” interpretation of the Clause. The Court, returning to the intention of the Framers, “rel[ied] on the conditions which existed when the Constitution was adopted.” And the Framers knew “that government by the people instituted by the Constitution would not imitate the conduct of arbitrary monarchs. The abuse of power might, indeed, be- apprehended, but not that it would be manifested in provisions or practices which would shock the sensibilities of men.” Id., at 375. The Clause, then, guards against “[t]he abuse of power”; contrary to the implications in Wilkerson v. Utah, supra, and In re Kemmler, supra, the prohibition of the Clause is not “confine[d] ... to such penalties and punishment as were inflicted by the Stuarts.” 217 U. S., at 372. Although opponents of the Bill of Rights “felt sure that the spirit of liberty could be trusted, and that its ideals would be represented, not debased, by legislation,” ibid., the Framers disagreed:

“[Patrick] Henry and those who believed as he did would take no chances. Their predominant political impulse was distrust of power, and they insisted on constitutional limitations against its abuse. But surely they intended more than to register a fear of the forms of abuse that went out of practice with the Stuarts. Surely, their [jealousy] of power had a saner justification than that. They were men of action, practical and sagacious, not beset with vain imagining, and it must have come to them that there could be exercises of cruelty by laws other than those which inflicted bodily pain or mutilation. With power in a legislature great, if not unlimited, to give criminal character to the actions of men, with power unlimited to fix terms of imprisonment with what accompaniments they *267might, what more potent instrument of cruelty could be put into the hands of power? And it was believed that power might be tempted to cruelty. This was the motive of the clause, and if we are to attribute an intelligent providence to its advocates we cannot think that it was intended to prohibit only practices like the [Stuarts’,] or to prevent only an exact repetition of history. We cannot think that the possibility of a coercive cruelty being exercised through other forms of punishment was overlooked.” Id., at 372-373.

The Court in Weems thus recognized that this “restraint upon legislatures” possesses an “expansive and vital character” that is “ ‘essential ... to the rule of law and the maintenance of individual freedom.’ ” Id., at 376-377. Accordingly, the responsibility lies with the courts to make certain that the prohibition of the Clause is enforced.8 Referring to cases in which “prominence [was] given to the power of the legislature to define crimes and their punishment,” the Court said:

“We concede the power in most of its exercises. We disclaim the right to assert a judgment *268against that of the legislature of the expediency of the laws or the right to oppose the judicial power to the legislative power to define crimes and fix their punishment, unless that power encounters in its exercise a constitutional prohibition. In such case not our discretion but our legal duty, strictly defined and imperative in its direction, is invoked.” Id., at 378.9

In short, this Court finally adopted the Framers’ view of the Clause as a “constitutional check” to ensure that “when we come to punishments, no latitude ought to be left, nor dependence put on the virtue of representatives.” That, indeed, is the only view consonant with our constitutional form of government. If the judicial conclusion that a punishment is “cruel and unusual” “depend [ed] upon virtually unanimous condemnation of the penalty at issue,” then, “[l]ike no other constitutional provision, [the Clause’s] only function would be to legitimize advances already made by the other departments and opinions already the conventional wisdom.” We know that the Framers did not envision “so narrow a role for this basic guaranty of human rights.” Goldberg & Dershowitz, Declaring the Death Penalty Unconstitutional, 83 Harv. L. Rev. 1773, 1782 (1970). The right to be free of cruel and unusual punishments, like the other guarantees of the Bill of Rights, “may not be submitted to vote; [it] depend [s] on the outcome of no elections.” “The very purpose of a Bill of Rights was to withdraw certain subjects from the vicissitudes of political controversy, to place them beyond the reach of majorities and officials and to establish them as legal principles to be applied *269by the courts." Board of Education v. Barnette, 319 U. S. 624, 638 (1943).

Judicial enforcement of the Clause, then, cannot be evaded by invoking the obvious truth that legislatures have the power to prescribe punishments for crimes. That is precisely the reason the Clause appears in the Bill of Rights. The difficulty arises, rather, in formulating the “legal principles to be applied by the courts" when a legislatively prescribed punishment is challenged as “cruel and unusual." In formulating those constitutional principles, we must avoid the insertion of “judicial conception [s] of . . . wisdom or propriety," Weems v. United States, 217 U. S., at 379, yet we must not, in the guise of “judicial restraint,” abdicate our fundamental responsibility to enforce the Bill of Rights. Were we to do so, the “constitution would indeed be as easy of application as it would be deficient in efficacy and power. Its general principles would have little value and be converted by precedent into impotent and lifeless formulas. Rights declared in words might be lost in reality.” Id., at 373. The Cruel and Unusual Punishments Clause would become, in short, “little more than good advice.” Trop v. Dulles, 366 U. S., at 104.

II

Ours would indeed be a simple task were we required merely to measure a challenged punishment against those that history has long condemned. That narrow and unwarranted view of the Clause, however, was left behind with the 19th century. Our task today is more complex. We know “that the words of the [Clause] are not precise, and that their scope is not static." We know, therefore, that the Clause “must draw its meaning from the evolving standards of decency that mark the prog*270ress of a maturing society.” Id., at 100-101.10 That knowledge, of course, is but the beginning of the inquiry.

In Trop v. Dulles, supra, at 99, it was said that “[t]he question is whether [a] penalty subjects the individual to a fate forbidden by the principle of civilized treatment guaranteed by the [Clause].” It was also said that a challenged punishment must be examined “in light of the basic prohibition against inhuman treatment” embodied in the Clause. Id., at 100 n. 32. It was said, finally, that:

“The basic concept underlying the [Clause] is nothing less than the dignity of man. While the State has the power to punish, the [Clause] stands to assure that this power be exercised within the limits of civilized standards.” Id., at 100.

At bottom, then, the Cruel and Unusual Punishments Clause prohibits the infliction of uncivilized and inhuman punishments. The State, even as it punishes, must treat its members with respect for their intrinsic worth as human beings. A punishment is “cruel and unusual,” therefore, if it does not comport with human dignity.

This formulation, of course, does not of itself yield principles for assessing the constitutional validity of particular punishments. Nevertheless, even though “[t]his Court has had little occasion to give precise content to the [Clause],” ibid., there are principles recognized in our cases and inherent in the Clause sufficient to permit a judicial determination whether a challenged punishment comports with human dignity.

*271The primary principle is that a punishment must not be so severe as to be degrading to the dignity of human beings. Pain, certainly, may be a factor in the judgment. The infliction of an extremely severe punishment will often entail physical suffering. See Weems v. United States, 217 U. S., at 366.11 Yet the Framers also knew “that there could be exercises of cruelty by laws other than those which inflicted bodily pain or mutilation.” Id., at 372. Even though “[t]here may be involved no physical mistreatment, no primitive torture,” Trop v. Dulles, supra, at 101, severe mental pain may be inherent in the infliction of a particular punishment. See Weems v. United States, supra, at 366.12 That, indeed, was one of the conclusions underlying the holding of the plurality in Trop v. Dulles that the punishment of expatriation violates the Clause.13 And the *272physical and mental suffering inherent in the punishment of cadena temporal, see nn. 11-12, supra, was an obvious basis for the Court’s decision in Weems v. United States that the punishment was “cruel and unusual.” 14

More than the presence of pain, however, is comprehended in the judgment that the extreme severity of a punishment makes it degrading to the dignity of human beings. The barbaric punishments condemned by history, “punishments which inflict torture, such as the rack, the thumbscrew, the iron boot, the stretching of limbs and the like,” are, of course, “attended with acute pain and suffering.” O’Neil v. Vermont, 144 U. S. 323, 339 (1892) (Field, J., dissenting). When we consider why they have been condemned, however, we realize that the pain involved is not the only reason. The true significance of these punishments is that they treat *273members of the human race as nonhumans, as objects to be toyed with and discarded. They are thus inconsistent with the fundamental premise of the Clause that even the vilest criminal remains a human being possessed of common human dignity.

The infliction of an extremely severe punishment, then, like the one before the Court in Weems v. United States, from which “[n]o circumstance of degradation [was] omitted,” 217 U. S., at 366, may reflect the attitude that the person punished is not entitled to recognition as a fellow human being. That attitude may be apparent apart from the severity of the punishment itself. In Louisiana ex rel. Francis v. Resweber, 329 U. S. 459, 464 (1947), for example, the unsuccessful electrocution, although it caused "mental anguish and physical pain,” was the result of “an unforeseeable accident.” Had the failure been intentional, however, the punishment would have been, like torture, so degrading and indecent as to amount to a refusal to accord the criminal human status. Indeed, a punishment may be degrading to human dignity solely because it is a punishment. A State may not punish a person for being “mentally ill, or a leper, or . . . afflicted with a venereal disease,” or for being addicted to narcotics. Robinson v. California, 370 U. S. 660, 666 (1962). To inflict punishment for having a disease is to treat the individual as a diseased thing rather than as a sick human being. That the punishment is not severe, “in the abstract,” is irrelevant; “[e]ven one day in prison would be a cruel and unusual punishment for the 'crime’ of having a common cold.” Id., at 667. Finally, of course, a punishment may be degrading simply by reason of its enormity. A prime example is expatriation, a “punishment more primitive than torture,” Trop v. Dulles, 356 U. S., at 101, for it necessarily involves a *274denial by society of the individual's existence as a member of the human community.15

In determining whether a punishment comports with human dignity, we are aided also by a second principle inherent in the Clause — that the State must not arbitrarily inflict a severe punishment. This principle derives from the notion that the State does not respect human dignity when, without reason, it inflicts upon some people a severe punishment that it does not inflict upon others. Indeed, the very words “cruel and unusual punishments” imply condemnation of the arbitrary infliction of severe punishments. And, as we now know, the English history of the Clause10 reveals a particular concern with the establishment of a safeguard against arbitrary punishments. See Granucci, “Nor Cruel and Unusual Punishments Inflicted:” The Original Meaning, 57 Calif. L. Rev. 839, 857-860 (1969).17

*275This principle has been recognized in our cases.18 In Wilkerson v. Utah, 99 U. S., at 133 — 134, the Court reviewed various treatises on military law in order to demonstrate that under “the custom of war” shooting was a common method of inflicting the punishment of death. On that basis, the Court concluded:

“Cruel and unusual punishments are forbidden by the Constitution, but the authorities referred to [treatises on military law] are quite sufficient to show that the punishment of shooting as a mode of executing the death penalty for the crime of murder in the first degree is not included in that *276category, within the meaning of the [Clause]. Soldiers convicted of desertion or other capital military offenses are in the great majority of cases sentenced to be shot, and the ceremony for such occasions is given in great fulness by the writers upon the subject of courts-martial.” Id., at 13A-135.

The Court thus upheld death by shooting, so far as appears, solely on the ground that it was a common method of execution.19

As Wilkerson v. Utah suggests, when a severe punishment is inflicted "in the great majority of cases” in which it is legally available, there is little likelihood that the State is inflicting it arbitrarily. If, however, the infliction of a severe punishment is “something different from that which is generally done” in such cases, Trop v. Dulles, 356 U. S., at 101 n. 32,20 there is a sub*277stantial likelihood that the State, contrary to the requirements of regularity and fairness embodied in the Clause, is inflicting the punishment arbitrarily. This principle is especially important today. There is scant danger, given the political processes “in an enlightened democracy such as ours,” id., at 100, that extremely severe punishments will be widely applied. The more significant function of the Clause, therefore, is to protect against the danger of their arbitrary infliction.

A third principle inherent in the Clause is that a severe punishment must not be unacceptable to contemporary society. Rejection by society, of course, is a strong indication that a severe punishment does not comport with human dignity. In applying this principle, however, we must make certain that the judicial determination is as objective as possible.21 *278Thus, for example, Weems v. United States, 217 U. S., at 380, and Trop v. Dulles, 356 U. S., at 102-103, suggest that one factor that may be considered is the existence of the punishment in jurisdictions other than those before the Court. Wilkerson v. Utah, supra, suggests that another factor to be considered is the historic usage of the punishment.22 Trop v. Dulles, supra, at 99, combined present acceptance with past usage by observing that “the death penalty has been employed throughout our history, and, in a day when it is still widely accepted, it cannot be said to violate the constitutional concept of cruelty.” In Robinson v. California, 370 U. S., at 666, which involved the infliction of punishment for narcotics addiction, the Court went a step further, concluding simply that “in the light of contemporary human knowledge, a law which made a criminal offense of such a disease would doubtless be universally thought to be an infliction of cruel and unusual punishment.”

The question under this principle, then, is whether there are objective indicators from which a court can conclude that contemporary society considers a severe punishment unacceptable. Accordingly, the judicial *279task is to review the history of a challenged punishment and to examine society’s present practices with respect to its use. Legislative authorization, of course, does not establish acceptance. The acceptability of a severe punishment is measured, not by its availability, for it might become so offensive to society as never to be inflicted, but by its use.

The final principle inherent in the Clause is that a severe punishment must not be excessive. A punishment is excessive under this principle if it is unnecessary: The infliction of a severe punishment by the State cannot comport with human dignity when it is nothing more than the pointless infliction of suffering. If there is a significantly less severe punishment adequate to achieve the purposes for which the punishment is inflicted, cf. Robinson v. California, supra, at 666; id., at 677 (Douglas, J., concurring); Trop v. Dulles, supra, at 114 (Brennan, J., concurring), the punishment inflicted is unnecessary and therefore excessive.

This principle first appeared in our cases in Mr. Justice Field’s dissent in O’Neil v. Vermont, 144 U. S., at 337.23 He there took the position that:

“[The Clause] is directed, not only against punishments of the character mentioned [torturous punishments], but against all punishments which by *280their excessive length or severity are greatly dispro-portioned to the offences charged. The whole inhibition is against that which is excessive either in the bail required, or fine imposed, or punishment inflicted.” Id., at 339-340.

Although the determination that a severe punishment is excessive may be grounded in a judgment that it is disproportionate to the crime,24 the more significant basis is that the punishment serves no penal purpose more effectively than a less severe punishment. This view of the principle was explicitly recognized by the Court in Weems v. United States, supra. There the Court, reviewing a severe punishment inflicted for the falsification of an official record, found that “the highest punishment possible for a crime which may cause the loss of many thousand [s] of dollars, and to prevent which the duty of the State should be as eager as to prevent the perversion of truth in a public document, is not greater than that which may be imposed for falsifying a single item of a public account.” Id., at 381. Stating that “this contrast shows more than different exercises of legislative judgment,” the Court concluded that the punishment was unnecessarily severe in view of the purposes for which it was imposed. Ibid.25 *281See also Trop v. Dulles, 356 U. S., at 111-112 (Brennan, J., concurring).26

There are, then, four principles by which we may determine whether a particular punishment is “cruel and unusual.” The primary principle, which I believe supplies the essential predicate for the application of the others, is that a punishment must not by its severity be degrading to human dignity. The paradigm violation of this principle would be the infliction of a torturous punishment of the type that the Clause has always prohibited. Yet “ [i] t is unlikely that any State at this moment in history,” Robinson v. California, 370 U. S., at 666, would pass a law providing for the infliction of such a punishment. Indeed, no such punishment has ever been before this Court. The same may be said of the other principles. It is unlikely that this Court will confront a severe punishment that is obviously inflicted in wholly arbitrary fashion; no State would engage in a reign of blind terror. Nor is it likely that this Court will be called upon to review a severe punishment that is clearly and totally rejected throughout society; no legislature would be able even to authorize the infliction of such a punishment. Nor, finally, is it likely that this Court will have to consider a severe punishment that is patently unnecessary; no State today would inflict a severe punishment knowing that there was no reason whatever for doing so. In short, we are unlikely to have occasion to determine that a punishment is fatally offensive under any one principle.

*282Since the Bill of Rights was adopted, this Court has adjudged only three punishments to be within the prohibition of the Clause. See Weems v. United States, 217 U. S. 349 (1910) (12 years in chains at hard and painful labor); Trop v. Dulles, 356 U. S. 86 (1958) (expatriation) ; Robinson v. California, 370 U. S. 660 (1962) (imprisonment for narcotics addiction). Each punishment, of course, was degrading to human dignity, but of none could it be said conclusively that it wasl fatally offensive under one or the other of the principles. Rather, these “cruel and unusual punishments” seriously implicated several of the principles, and it was the application of the principles in combination that supported the judgment. That, indeed, is not surprising. The function of these principles, after all, is simply to provide means by which a court can determine whether a challenged punishment comports with human dignity. They are, therefore, interrelated, and in most cases it will be their convergence that will justify the conclusion that a punishment is “cruel and unusual.” The test, then, will ordinarily be a cumulative one: If a punishment is unusually severe, if there is a strong probability that it is inflicted arbitrarily, if it is substantially rejected by contemporary society, and if there is no reason to believe that it serves any penal purpose more effectively than some less severe punishment, then the continued infliction of that punishment violates the command of the Clause that the State may not inflict inhuman and uncivilized punishments upon those convicted of crimes.

Ill

The punishment challenged in these cases is death. Death, of course, is a “traditional” punishment, Trop v. Dulles, supra, at 100, one that “has been employed throughout our history,” id., at 99, and its constitu*283tional background is accordingly an appropriate subject of inquiry.

There is, first, a textual consideration raised by the Bill of Rights itself. The Fifth Amendment declares that if a particular crime is punishable by death, a person charged with that crime is entitled to certain procedural protections.27 We can thus infer that the Framers recognized the existence of what was then a common punishment. We cannot, however, make the further inference that they intended to exempt this particular punishment from the express prohibition of the Cruel and Unusual Punishments Clause.28 Nor is there any indication in the debates on the Clause that a special exception was to be made for death. If anything, the indication is to the contrary, for Livermore specifically mentioned death as a candidate for future proscription under the Clause. See supra, at 262. Finally, it does not advance analysis to insist that the Framers did not believe that adoption *284of the Bill of Rights would immediately prevent the infliction of the punishment of death; neither did they believe that it would immediately prevent the infliction of other corporal punishments that, although common at the time, see n. 6, supra, are now acknowledged to be impermissible.29

There is also the consideration that this Court has decided three cases involving constitutional challenges to particular methods of inflicting this punishment. In Wilkerson v. Utah, 99 U. S. 130 (1879), and In re Kemmler, 136 U. S. 436 (1890), the Court, expressing in both cases the since-rejected “historical” view of the Clause, see supra, at 264-265, approved death by shooting and death by electrocution. In Wilkerson, the Court concluded that shooting was a common method of execution, see supra, at 275-276;30 in Kemmler, the Court held that the Clause did not apply to the States, 136 U. S., at 447-449.31 *285In Louisiana ex rel. Francis v. Resweber, supra, the Court approved a second attempt at electrocution after the first had failed. It was said that “ [t] he Fourteenth [Amendment] would prohibit by its due process clause execution by a state in a cruel manner,” 329 U. S., at 463, but that the abortive attempt did not make the “subsequent execution any more cruel in the constitutional sense than any other execution,” id., at 464.32 These three decisions thus reveal that the Court, while ruling upon various methods of inflicting death, has assumed in the past that death was a constitutionally permissible punishment.33 Past assumptions, however, are not sufficient to limit the scope of our examination of this punishment today. The constitutionality of death itself under the Cruel and Unusual Punishments Clause is before this Court for the first time; we cannot avoid the question by recalling past cases that never directly considered it.

The question, then, is whether the deliberate infliction of death is today consistent with the command of the Clause that the State may not inflict punishments that do not comport with human dignity. I will analyze the punishment of death in terms of the principles *286set out above and the cumulative test to which they lead: It is a denial of human dignity for the State arbitrarily to subject a person to an unusually severe punishment that society has indicated it does not regard as acceptable, and that cannot be shown to serve any penal purpose more effectively than a significantly less drastic punishment. Under these principles and this test, death is today a “cruel and unusual” punishment.

Death is a unique punishment in the United States. In a society that so strongly affirms the sanctity of life, not surprisingly the common view is that death is the ultimate sanction. This natural human feeling appears all about us. There has been no national debate about punishment, in general or by imprisonment, comparable to the debate about the punishment of death. No other punishment has been so continuously restricted, see infra, at 296-298, nor has any State yet abolished prisons, as some have abolished this punishment. And those States that still inflict death reserve it for the most heinous crimes. Juries, of course, have always treated death cases differently, as have governors exercising their commutation powers. Criminal defendants are of the same view. “As all practicing lawyers know, who have defended persons charged with capital offenses, often the only goal possible is to avoid the death penalty.” Griffin v. Illinois, 351 U. S. 12, 28 (1956) (Burton and Minton, JJ., dissenting). Some legislatures have required particular procedures, such as two-stage trials and automatic appeals, applicable only in death cases. “It is the universal experience in the administration of criminal justice that those charged with capital offenses are granted special considerations.” Ibid. See Williams v. Florida, 399 U. S. 78, 103 (1970) (all States require juries of 12 in death cases). This Court, too, almost *287always treats death cases as a class apart.34 And the unfortunate effect of this punishment upon the functioning of the judicial process is well known; no other punishment has a similar effect.

The only explanation for the uniqueness of death is its extreme severity. Death is today an unusually severe punishment, unusual in its pain, in its finality, and in its enormity. No other existing punishment is comparable to death in terms of physical and mental suffering. Although our information is not conclusive, it appears that there is no method available that guarantees an immediate and painless death.35 Since the discon*288tinuance of flogging as a constitutionally permissible punishment, Jackson v. Bishop, 404 F. 2d 571 (CA8 1968), death remains as the only punishment that may involve the conscious infliction of physical pain. In addition, we know that mental pain is an inseparable part of our practice of punishing criminals by death, for the prospect of pending execution exacts a frightful toll during the inevitable long wait between the imposition of sentence and the actual infliction of death. Cf. Ex parte Medley, 134 U. S. 160, 172 (1890). As the California Supreme Court pointed out, “the process of carrying out a verdict of death is often so degrading and brutalizing to the human spirit as to constitute psychological torture.” People v. Anderson, 6 Cal. 3d 628, 649, 493 P. 2d 880, 894 (1972).36 Indeed, as Mr. Justice Frankfurter noted, “the onset of insanity while awaiting *289execution of a death sentence is not a rare phenomenon.” Solesbee v. Balkcom, 339 U. S. 9, 14 (1950) (dissenting opinion). The “fate of ever-increasing fear and distress” to which the expatriate is subjected, Trop v. Dulles, 356 U. S., at 102, can only exist to a greater degree for a person confined in prison awaiting death.37

The unusual severity of death is manifested most clearly in its finality and enormity. Death, in these respects, is in a class by itself. Expatriation, for example, is a punishment that “destroys for the individual the political existence that was centuries in the development,” that “strips the citizen of his status in the national and international political community,” and that puts “[h]is very existence” in jeopardy. Expatriation thus inherently entails “the total destruction of the individual’s status in organized society.” Id., at 101. “In short, the expatriate has lost the right to have rights.” Id., at 102. Yet, demonstrably, expatriation is not “a fate worse than death.” Id., at 125 (Frankfurter, J., dissenting).38 Although death, like expatriation, destroys the *290individual’s “political existence” and his “status in organized society,” it does more, for, unlike expatriation, death also destroys “[h]is very existence.” There is, too, at least the possibility that the expatriate will in the future regain “the right to have rights.” Death forecloses even that possibility.

Death is truly an awesome punishment. The calculated killing of a human being by the State involves, by its very nature, a denial of the executed person’s humanity. The contrast with the plight of a person punished by imprisonment is evident. An individual in prison does not lose “the right to have rights.” A prisoner retains, for example, the constitutional rights to the free exercise of religion, to be free of cruel and unusual punishments, and to treatment as a “person” for purposes of due process of law and the equal protection of the laws. A prisoner remains a member of the human family. Moreover, he retains the right of access to the courts. His punishment is not irrevocable. Apart from the common charge, grounded upon the recognition of human fallibility, that the punishment of death must inevitably be inflicted upon innocent men, we know that death has been the lot of men whose convictions were unconstitutionally secured in view of later, retroactively applied, holdings of this Court. The punishment itself may have been unconstitutionally inflicted, see Witherspoon v. Illinois, 391 U. S. 510 (1968), yet the finality of death precludes relief. An executed person has indeed “lost the right to have rights.” As one 19th century proponent of punishing criminals by death declared, “When a man is hung, there is an end of our relations with him. His execution is a way of saying, ‘You are not fit for this world, take your chance elsewhere.’ ”39

*291In comparison to all other punishments today, then, the deliberate extinguishment of human life by the State is uniquely degrading to human dignity. I would not hesitate to hold, on that ground alone, that death is today a “cruel and unusual” punishment, were it not that death is a punishment of longstanding usage and acceptance in this country. I therefore turn to the second principle — that the State may not arbitrarily inflict an unusually severe punishment.

The outstanding characteristic of our present practice of punishing criminals by death is the infrequency with which we resort to it. The evidence is conclusive that death is not the ordinary punishment for any crime.

There has been a steady decline in the infliction of this punishment in every decade since the 1930’s, the earliest period for which accurate statistics are available. In the 1930’s, executions averaged 167 per year; in the 1940’s, the average was 128; in the 1950’s, it was 72; and in the years 1960-1962, it was 48. There have been a total of. 46 executions since then, 36 of them in 1963-1964.40 Yet our population and the number of capital crimes committed have increased greatly over the past four decades. The contemporary rarity of the infliction of this punishment is thus the end result of a long-continued decline. That rarity is plainly revealed by an examination of the years 1961-1970, the last 10-year period for which statistics are available. During that time, an average of 106 death sentences *292was imposed each year.41 Not nearly that number, however, could be carried out, for many were precluded by commutations to life or a term of years,42 transfers to mental institutions because of insanity,43 resentences to life or a term of years, grants of new trials and orders for resentencing, dismissals of indictments and reversals of convictions, and deaths by suicide and natural causes.44 On January 1, 1961, the death row population was 219; on December 31, 1970, it was 608; during that span, there were 135 executions.45 Consequently, had the 389 additions to death row also been executed, the annual average would have been 52.46 In short, the country *293might, at most, have executed one criminal each week. In fact, of course, far fewer were executed. Even before the moratorium on executions began in 1967, executions totaled only 42 in 1961 and 47 in 1962, an average of less than one per week; the number dwindled to 21 in 1963, to 15 in 1964, and to seven in 1965; in 1966, there was one execution, and in 1967, there were two.47

When a country of over 200 million people inflicts an unusually severe punishment no more than 50 times a year, the inference is strong that the punishment is hot being regularly and fairly applied. To dispel it would indeed require a clear showing of nonarbitrary infliction.

Although there are no exact figures available, we know that thousands of murders and rapes are committed annually in States where death is an authorized punishment for those crimes. However the rate of infliction is characterized — as “freakishly” or “spectacularly” rare, or simply as rare — it would take the purest sophistry to deny that death is inflicted in only a minute fraction of these cases. How much rarer, after all, could the infliction of death be?

When the punishment of death is inflicted in a trivial number of the cases in which it is legally available, the conclusion is virtually inescapable that it is being inr flicted arbitrarily. Indeed, it smacks of little more than a lottery system. The States claim, however, that this rarity is evidence not of arbitrariness, but of informed selectivity: Death is inflicted, they say, only in “extreme” cases.

Informed selectivity, of course, is a value not to be denigrated. Yet presumably the States could make precisely the same claim if there were 10 executions per *294year, or five, or even if tHere were but one. That there may be as many as 50 per year does not strengthen the claim. When, the rate of infliction is at this low level, it is highly implausible that only the worst criminals or the criminals who commit the worst crimes are selected for this punishment. No one has yet suggested a rational basis that could differentiate in those terms the few who die from the many who go to prison. Crimes and criminals simply do not admit of a distinction that can be drawn so finely as to explain, on that ground, the execution of such a tiny sample of those eligible. Certainly the laws that provide for this punishment do not attempt to draw that distinction; all cases to which the laws apply are necessarily “extreme.” Nor is the distinction credible in fact. If, for example, petitioner Furman or his crime illustrates the “extreme,” then nearly all murderers and their murders are also “extreme.” 48 Furthermore, our procedures in death cases, *295rather than resulting in the selection of “extreme” cases for this punishment, actually sanction an arbitrary selection. For this Court has held that juries may, as they do, make the decision whether to impose a death sentence wholly unguided by standards governing that decision. McGautha v. California, 402 U. S. 183, 196—208 (1971). In other words, our procedures are not constructed to guard against the totally capricious selection of criminals for the punishment of death.

Although it is difficult to imagine what further facts would be necessary in order to prove that death is, as my Brother Stewart puts it, “wantonly and . . . freakishly” inflicted, I need not conclude that arbitrary infliction is patently obvious. I am not considering this punishment by the isolated light of one principle. The probability of arbitrariness is sufficiently substantial that it can be relied upon, in combination with the other principles, in reaching a judgment on the constitutionality of this punishment.

When there is a strong probability that an unusually severe and degrading punishment is being inflicted arbitrarily, we may well expect that society will disapprove of its infliction. I turn, therefore, to the third principle. An examination of the history and present operation of the American practice of punishing criminals by death reveals that this punishment has been almost totally rejected by contemporary society.

I cannot add to my Brother Marshall’s comprehensive treatment of the English and American history of *296this punishment. I emphasize, however, one significant conclusion that emerges from that history. From the beginning of our Nation, the punishment of death has stirred acute public controversy. Although pragmatic arguments for and against the punishment have been frequently advanced, this longstanding and heated controversy cannot be explained solely as the result of differences over the practical wisdom of a particular government policy. At bottom, the battle has been waged on moral grounds. The country has debated whether a society for which the dignity of the individual is the supreme value can, without a fundamental inconsistency, follow the practice of deliberately putting some of its members to death. In the United States, as in other nations of the western world, “the struggle about this punishment has been .one between ancient and deeply rooted beliefs in retribution, atonement or vengeance on the one hand, and, on the other, beliefs in the personal value and dignity of the common man that were born of the democratic movement of the eighteenth century, as well as beliefs in the scientific approach to an understanding of the motive forces of human conduct, which are the result of the growth of the sciences of behavior during the nineteenth and twentieth centuries.” 49 It is this essentially moral conflict that forms the backdrop for the past changes in and the present operation of our system of imposing death as a punishment for crime.

Our practice of punishing criminals by death has changed greatly over the years. One significant change has been in our methods of inflicting death. Although this country never embraced the more violent and repulsive methods employed in England, we did for a long time rely almost exclusively upon the gallows and the firing squad. Since the development of the supposedly *297more humane methods of electrocution late in the 19th century and lethal gas in the 20th, however, hanging and shooting have virtually ceased.50 Our concern for decency and human dignity, moreover, has compelled changes in the circumstances surrounding the execution itself. No longer does our society countenance the spectacle of public executions, once thought desirable as a deterrent to criminal behavior by others. Today we reject public executions as debasing and brutalizing to us all.

Also significant is the drastic decrease in the crimes for which the punishment of death is actually inflicted. While esoteric capital crimes remain on the books, since 1930 murder and rape have accounted for nearly 99% of the total executions, and murder alone for about 87%.51 In addition, the crime of capital murder has itself been limited. As the Court noted in McGautha v. California, 402 U. S., at 198, there was in this country a “rebellion against the common-law rule imposing a mandatory death sentence on all convicted murderers.” Initially, that rebellion resulted in legislative definitions that distinguished between degrees of murder, retaining the mandatory death sentence only for murder in the first degree. Yet “[t]h'is new legislative criterion for isolating crimes appropriately punishable by death soon proved as unsuccessful as the concept of ‘malice aforethought,’ ” ibid., the common-law means of separating murder from manslaughter. Not only was the distinction between degrees of murder confusing and uncertain in practice, but even in clear cases of first-degree murder juries continued to take the law into *298their own hands: if they felt that death was an inappropriate punishment, “they simply refused to convict of the capital offense.” Id., at 199. The phenomenon of jury nullification thus remained to counteract the rigors of mandatory death sentences. Bowing to reality, “legislatures did not try, as before, to refine further the definition of capital homicides. Instead they adopted the method of forthrightly granting juries the discretion which they had been exercising in fact.” Ibid. In consequence, virtually all death sentences today are discre-tionarily imposed. Finally, it is significant that nine States no longer inflict the punishment of death under any circumstances,52 and five others have restricted it to extremely rare crimes.53

*299Thus, although “the death penalty has been employed throughout our history,” Trop v. Dulles, 356 U. S., at 99, in fact the history of this punishment is one of successive restriction. What was once a common punishment has become, in the context of a continuing moral debate, increasingly rare. The evolution of this punishment evidences, not that it is an inevitable part of the American scene, but that it has proved progressively more troublesome to the national conscience. The result of this movement is our current system of administering the punishment, under which death sentences are rarely imposed and death is even more rarely inflicted. It is, of course, “We, the People” who are responsible for the rarity both of the imposition and the carrying out of this punishment. Juries, “express [ing] the conscience of the community on the ultimate question of life or death,” Witherspoon v. Illinois, 391 U. S., at 519, have been able to bring themselves to vote for death in a mere 100 or so cases among the thousands tried each year where the punishment is available. Governors, elected by and acting for us, have regularly commuted a substantial number of those sentences. And it is our society that insists upon due process of law to the end that no person will be unjustly put to death, thus ensuring that many more of those sentences will not be carried out. In sum, we have made death a rare punishment today.

The progressive decline in, and the current rarity of, the infliction of death demonstrate that our society seriously questions the appropriateness of this punishment today. The States point out that many legislatures authorize death as the punishment for certain crimes and that substantial segments of the public, as reflected in opinion polls and referendum votes, continue to support it. Yet the availability of this punishment through statutory authorization, as well as the polls and refer-*300enda, which amount simply to approval of that authorization, simply underscores the extent to which our society has in fact rejected this punishment. When an unusually severe punishment is authorized for wide-scale application but not, because of society’s refusal, inflicted save in a few instances, the inference is compelling that there is a deep-seated reluctance to inflict it. Indeed, the likelihood is great that the punishment is tolerated only because of its disuse. The objective indicator of society’s view of an unusually severe punishment is what society does with it, and today society will inflict death upon only a small sample of the eligible criminals. Rejection could hardly be more complete without becoming absolute. At the very least, I must conclude that contemporary society views this punishment with substantial doubt.

The final principle to be considered is that an unusually severe and degrading punishment may not be excessive in view of the purposes for which it is inflicted. This principle, too, is related to the others. When there is a strong probability that the State is arbitrarily inflicting an unusually severe punishment that is subject to grave societal doubts, it is likely also that the punishment cannot be shown to be serving any penal purpose that could not be served equally well by some less severe punishment.

The States’ primary claim is that death is a necessary punishment because it prevents the commission of capital crimes more effectively than any less severe punishment. The first part of this claim is that the infliction of death is necessary to stop the individuals executed from committing further crimes. The sufficient answer to this is that if a criminal convicted of a capital crime poses a danger to society, effective administration of the State’s pardon and parole laws can delay or deny his release from prison, and techniques of isolation can elim*301inate or minimize the danger while he remains confined.

The more significant argument is that the threat of death prevents the commission of capital crimes because it deters potential criminals who would not be deterred by the threat of imprisonment. The argument is not based upon evidence that the threat of death is a superior deterrent. Indeed, as my Brother Marshall establishes, the available evidence uniformly indicates, although it does not conclusively prove, that the threat of death has no greater deterrent effect than the threat of imprisonment. The States argue, however, that they are entitled to rely upon common human experience, and that experience, they say, supports the conclusion that death must be a more effective deterrent than any less severe punishment. Because people fear death the most, the argument runs, the threat of death must be the greatest deterrent.

It is important to focus upon the precise import of this argument. It is not denied that many, and probably most, capital crimes cannot be deterred by the threat of punishment. Thus the argument can apply only to those who think rationally about the commission of capital crimes. Particularly is that true when the potential criminal, under this argument, must not only consider the risk of punishment, but also distinguish between two possible punishments. The concern, then, is with a particular type of potential criminal, the rational person who will commit a capital crime knowing that the punishment is long-term imprisonment, which may well be for the rest of his life, but will not commit the crime knowing that the punishment is death. On the face of it, the assumption that such persons exist is implausible.

In any event, this argument cannot be appraised in the abstract. We are not presented with the theoretical question whether under any imaginable circumstances the *302threat of death might be a greater deterrent to the commission of capital crimes than the threat of imprisonment. We are concerned with the practice of punishing criminals by death as it exists in the United States today. Proponents of this argument necessarily admit that its validity depends upon the existence of a system in which the punishment of death is invariably and swiftly imposed. Our system, of course, satisfies neither condition. A rational person contemplating a murder or rape is confronted, not with the certainty of a speedy death, but with the slightest possibility that he will be executed in the distant future. The risk of death is remote and improbable; in contrast, the risk of long-term imprisonment is near and great. In short, whatever the speculative validity of the assumption that the threat of death is a superior deterrent, there is no reason to believe that as currently administered the punishment of death is necessary to deter the commission of capital crimes. Whatever might be the case were all or substantially all eligible criminals quickly put to death, unverifiable possibilities are an insufficient basis upon which to conclude that the threat of death today has any greater deterrent efficacy than the threat of imprisonment.54

*303There is, however, another aspect to the argument that the punishment of death is necessary for the protection of society. The infliction of death, the States urge, serves to manifest the community’s outrage at the commission of the crime. It is, they say, a concrete public expression of moral indignation that inculcates respect for the law and helps assure a more peaceful community. Moreover, we are told, not only does the punishment of death exert this widespread moralizing influence upon community values, it also satisfies the popular demand for grievous condemnation of abhorrent crimes and thus prevents disorder, lynching, and attempts by private citizens to take the law into their own hands.

The question, however, is not whether death serves these supposed purposes of punishment, but whether death serves them more effectively than imprisonment. There is no evidence whatever that utilization of imprisonment rather than death encourages private blood feuds and other disorders. Surely if there were such a danger, the execution of a handful of criminals each year would not prevent it. The assertion that death alone is a sufficiently emphatic denunciation for capital crimes suffers from the same defect. If capital crimes require the punishment of death in order to provide moral reinforcement for the basic values of the community, those values can only be undermined when death is so rarely inflicted upon the criminals who commit the crimes. Furthermore, it is certainly doubtful that the infliction of death by the State does in fact strengthen the community’s moral code; if the deliberate extin-guishment of human life has any effect at all, it more likely tends to lower our respect for life and brutalize our values. That, after all, is why we no longer carry out public executions. In any event, this claim simply means that one purpose of punishment is to indicate social disapproval of crime. To serve that purpose our *304laws distribute punishments according to the gravity of crimes and punish more severely the crimes society regards as more serious. That purpose cannot justify any particular punishment as the upper limit of severity.

There is, then, no substantial reason to believe that the punishment of death, as currently administered, is necessary for the protection of society. The only other purpose suggested, one that is independent of protection for society, is retribution. Shortly stated, retribution in this context means that criminals are put to death because they deserve it.

Although it is difficult to believe that any State today wishes to proclaim adherence to “naked vengeance,” Trop v. Dulles, 356 U. S., at 112 (Brennan, J., concurring), the States claim, in reliance upon its statutory authorization, that death is the only fit punishment for capital crimes and that this retributive purpose justifies its infliction. In the past, judged by its statutory authorization, death was considered the only fit punishment for the crime of forgery, for the first federal criminal statute provided a mandatory death penalty for that crime. Act of April 30, 1790, § 14, 1 Stat. 115. Obviously, concepts of justice change; no immutable moral order requires death for murderers and rapists. The claim that death is a just punishment necessarily refers to the existence of certain public beliefs. The claim must be that for capital crimes death alone comports with society’s notion of proper punishment. As administered today, however, the punishment of death cannot be justified as a necessary means of exacting retribution from criminals. When the overwhelming number of criminals who commit capital crimes go to prison, it cannot be concluded that death serves the purpose of retribution more effectively than imprisonment. The asserted public belief that murderers and rapists deserve to die is flatly inconsistent with the execution of a random *305few. As the history of the punishment of death in this country shows, our society wishes to prevent crime; we have no desire to kill criminals simply to get even with them.

In sum, the punishment of death is inconsistent with all four principles: Death is an unusually severe and degrading punishment; there is a strong probability that it is inflicted arbitrarily; its rejection by contemporary society is virtually total; and there is no reason to believe that it serves any penal purpose more effectively than the less severe punishment of imprisonment. The function of these principles is to enable a court to determine whether a punishment comports with human dignity. Death, quite simply, does not.

IV

When this country was founded, memories of the Stuart horrors were fresh and severe corporal punishments were common. Death was not then a unique punishment. The practice of punishing criminals by death, moreover, was widespread and by and large acceptable to society. Indeed, without developed prison systems, there was frequently no workable alternative. Since that time, successive restrictions, imposed against the background of a continuing moral controversy, have drastically curtailed the use of this punishment. Today death is a uniquely and unusually severe punishment. When examined by the principles applicable under the Cruel and Unusual Punishments Clause, death stands condemned as fatally offensive to human dignity. The punishment of death is therefore “cruel and unusual,” and the States may no longer inflict it as a punishment for crimes. Rather than kill an arbitrary handful of criminals each year, the States will confine them in prison. “The State thereby suffers nothing and loses no power. The purpose of punishment is fulfilled, crime *306is repressed by penalties of just, not tormenting, severity, its repetition is prevented, and hope is given for the reformation of the criminal/’ Weems v. United States, 217 U. S., at 381.

1 concur in the judgments of the Court.

Me. Justice Stewaet,

concurring.

The penalty of death differs from all other forms of criminal punishment, not in degree but in kind. It is unique in its total irrevocability. It is unique in its rejection of rehabilitation of the convict as a basic purpose of criminal justice. And it is unique, finally, in its absolute renunciation of all that is embodied in our concept of humanity.

For these and other reasons, at least two of my Brothers have concluded that the infliction of the death penalty is constitutionally impermissible in all circumstances under the Eighth and Fourteenth Amendments. Their case is a strong one. But I find it unnecessary to reach the ultimate question they would decide. See Ashwander v. Tennessee Valley Authority, 297 U. S. 288, 347 (Brandéis, J., concurring).

The opinions of other Justices today have set out in admirable and thorough detail the origins and judicial history of the Eighth Amendment’s guarantee against the infliction of cruel and unusual punishments,1 and the origin and judicial history of capital punishment.2 There *307is thus no need for me to review the historical materials here, and what I have to say can, therefore, be briefly stated.

Legislatures — state and federal — have sometimes specified that the penalty of death shall be the mandatory punishment for every person convicted of engaging in certain designated criminal conduct. Congress, for example, has provided that anyone convicted of acting as a spy for the enemy in time of war shall be put to death.3 The Rhode Island Legislature has ordained the death penalty for a life term prisoner who commits murder.4 Massachusetts has passed a law imposing the death penalty upon anyone convicted of murder in the commission of a forcible rape.5 An Ohio law imposes the mandatory penalty of death upon the assassin of the President of the United States or the Governor of a State.6

If we were reviewing death sentences imposed under these or similar laws, we would be faced with the need to decide whether capital punishment is unconstitutional for all crimes and under all circumstances. We would need to decide whether a legislature — state or federal— could constitutionally determine that certain criminal conduct is so atrocious that society’s interest in deterrence and retribution wholly outweighs any considerations of reform or rehabilitation of the perpetrator, and that, despite the inconclusive empirical evidence,7 only *308the automatic penalty of death will provide maximum deterrence.

On that score I would say only that I cannot agree that retribution is a constitutionally impermissible ingredient in the imposition of punishment. The instinct for retribution is part of the nature of man, and channeling that instinct in the administration of criminal justice serves an important purpose in promoting the stability of a society governed by law. When people begin to believe that organized society is unwilling or unable to impose upon criminal offenders the punishment they “deserve,” then there are sown the seeds of anarchy — of self-help, vigilante justice, and lynch law.

The constitutionality of capital punishment in the abstract is not, however, before us in these cases. For the Georgia and Texas Legislatures have not provided that the death penalty shall be imposed upon all those who are found guilty of forcible rape.8 And the Georgia Legislature has not ordained that death shall be the automatic punishment for murder.9 In a word, neither State *309has made a legislative determination that forcible rape and murder can be deterred only by imposing the penalty of death upon all who perpetrate those offenses. As Mr. Justice White so tellingly puts it, the “legislative will is not frustrated if the penalty is never imposed.” Post, at 311.

Instead, the death sentences now before us are the product of a legal system that brings them, I believe, within the very core of the Eighth Amendment’s guarantee against cruel and unusual punishments, a guarantee applicable against the States through the Fourteenth Amendment. Robinson v. California, 370 U. S. 660. In the first place, it is clear that these sentences are “cruel” in the sense that they excessively go beyond, not in degree but in kind, the punishments that the state legislatures have determined to be necessary. Weems v. United States, 217 U. S. 349. In the second place, it is equally clear that these sentences are “unusual” in the sense that the penalty of death is infrequently imposed for murder, and that its imposition for rape is extraordinarily rare.10 But I do not rest my conclusion upon these two propositions alone.

These death sentences are cruel and unusual in the same way that being struck by lightning is cruel and unusual. For, of all the people convicted of rapes and murders in 1967 and 1968,11 many just as reprehensible as these, the petitioners are among a capriciously *310selected random handful upon whom the sentence of death has in fact been imposed.12 My concurring Brothers have demonstrated that, if any basis can be discerned for the selection of these few to be sentenced to die, it is the constitutionally impermissible basis of race.13 See McLaughlin v. Florida, 379 U. S. 184. But racial discrimination has not been proved,14 and I put it to one side. I simply conclude that the Eighth and Fourteenth Amendments cannot tolerate the infliction of a sentence of death under legal systems that permit this unique penalty to be so wantonly and so freakishly imposed.

For these reasons I concur in the judgments of the Court.

Me. Justice White,

concurring.

The facial constitutionality of statutes requiring the imposition of the death penalty for first-degree murder, for more narrowly defined categories of murder, or for rape would present quite different issues under the Eighth Amendment than are posed by the cases before us. In joining the Court's judgments, therefore, I do not at all *311intimate that the death penalty is unconstitutional per se or that there is no system of capital punishment that would comport with the Eighth Amendment. That question, ably argued by several of my Brethren, is not presented by these cases and need not be decided.

The narrower question to which I address myself concerns the constitutionality of capital punishment statutes under which (1) the legislature authorizes the imposition of the death penalty for murder or rape; (2) the legislature does not itself mandate the penalty in any particular class or kind of case (that is, legislative will is not frustrated if the penalty is never imposed), but delegates to judges or juries the decisions as to those cases, if any, in which the penalty will be utilized; and (3) judges and juries have ordered the death penalty with such infrequency that the odds are now very much against imposition and execution of the penalty with respect to any convicted murderer or rapist. It is in this context that we must consider whether the execution of these petitioners would violate the Eighth Amendment.

I begin with what I consider a near truism: that the death penalty could so seldom be imposed that it would cease to be a credible deterrent or measurably to contribute to any other end of punishment in the criminal justice system. It is perhaps true that no matter how infrequently those convicted of rape or murder are executed, the penalty so imposed is not disproportionate to the crime and those executed may deserve exactly what they received. It would also be clear that executed defendants are finally and completely incapacitated from again committing rape or murder or any other crime. But when imposition of the penalty reaches a certain degree of infrequency, it would be very doubtful that any existing general need for retribution would be measurably satisfied. Nor could it be said with confidence that society’s need for specific deterrence justifies death *312for so few when for so many in like circumstances life imprisonment or shorter prison terms are judged sufficient, or that community values are measurably reinforced by authorizing a penalty so rarely invoked.

Most important, a major goal of the criminal law— to deter others by punishing the convicted criminal— would not be substantially served where the penalty is so seldom invoked that it ceases to be the credible threat essential to influence the conduct of others. For present purposes I accept the morality and utility of punishing one person to influence another. I accept also the effectiveness of punishment generally and need not reject the death penalty as a more effective- deterrent than a lesser punishment. But common sense and experience tell us that seldom-enforced laws become ineffective measures for controlling human conduct and that the death penalty, unless- imposed with sufficient frequency, will make little contribution to deterring those crimes for which it may be exacted.

The imposition and execution of the death penalty are obviously cruel in the dictionary sense. But the penalty has not been considered cruel and unusual punishment in the constitutional sense because it was thought justified by the social ends it was deemed to serve. At the moment that it ceases realistically to further these purposes, however, the emerging question is whether its imposition in such circumstances would violate the Eighth Amendment. It is my view that it would, for its imposition would then be the pointless and needless extinction of life with only marginal contributions to any discernible social or public purposes. A penalty with such negligible returns to the State would be patently excessive and cruel and unusual punishment vio-lative of the Eighth Amendment.

It is also my judgment that this point has been reached with respect to capital punishment as it is presently ad*313ministered under the statutes involved in these cases. Concededly, it is difficult to prove as a general proposition that capital punishment, however administered, more effectively serves the ends of the criminal law than does imprisonment. But however that may be, I cannot avoid the conclusion that as the statutes before us are now administered, the penalty is so infrequently imposed that the threat of execution is too attenuated to be of substantial service to criminal justice.

I need not restate the facts and figures that appear in the opinions of my Brethren. Nor can I “prove” my conclusion from these data. But, like my Brethren, I must arrive at judgment; and I can do no more than state a conclusion based' on 10 years of almost daily exposure to the facts and circumstances of hundreds and hundreds of federal and state criminal cases involving crimes for which death is the authorized penalty. That conclusion, as I have said, is that the death penalty is exacted with great infrequency even for the most atrocious crimes and that there is no meaningful basis for distinguishing the few cases in which it is imposed from the many cases in which it is not. The short of it is that the policy of vesting sentencing authority primarily in juries — a decision largely motivated by the desire to mitigate the harshness of the law and to bring community judgment to bear on the sentence as well as guilt or innocence — has so effectively achieved its aims that capital punishment within the confines of the statutes now before us has for all practical purposes run its course.

Judicial review, by definition, often involves a conflict between judicial and legislative judgment as to what the Constitution means or requires. In this respect, Eighth Amendment cases come to us in no different posture. It seems conceded by all that the Amendment imposes some obligations on the judiciary to judge the *314constitutionality of punishment and that there are punishments that the Amendment would bar whether legislatively approved or not. Inevitably, then, there will be occasions when we will differ with Congressi or state legislatures with respect to the validity of punishment. There will also be cases in which we shall strongly disagree among ourselves. Unfortunately, this is one .of them. But as I see it, this case is no different in kind from many others, although it may have wider impact and provoke sharper disagreement.

In this respect, I add only that past and present legislative judgment with respect to the death penalty loses much of its force when viewed in light of the recurring practice of delegating sentencing authority to the jury and the fact that a jury, in its own discretion and without violating its trust or any statutory policy, may refuse to impose the death penalty no matter what the circumstances of the crime. Legislative “policy” is thus necessarily defined not by what is legislatively authorized but by what juries and judges do in exercising the discretion so regularly conferred upon them. In my judgment what was done in these cases violated the Eighth Amendment.

I concur in the judgments of the Court.

Mr. Justice Marshall,

concurring.

These three cases present the question whether the death penalty is a cruel and unusual punishment prohibited by the Eighth Amendment to the United States Constitution.1

*315In No. 69-5003, Furman was convicted of murder for shooting the father of five children when he discovered that Furman had broken into his home early one morning. Nos. 69-5030 and 69-5031 involve state convictions for forcible rape. Jackson was found guilty of rape during the course of a robbery in the victim’s home. The rape was accomplished as he held the pointed ends of scissors at the victim’s throat. Branch also was convicted of a rape committed in the victim’s home. No weapon was utilized, but physical force and threats of physical force were employed.

The criminal acts with which we are confronted are ugly, vicious, reprehensible acts. Their sheer brutality cannot and should not be minimized. But, we are not called upon to condone the penalized conduct; we are asked only to examine the penalty imposed on each of the petitioners and to determine whether or not it violates the Eighth Amendment. The question then is not whether we condone rape or murder, for surely we do not; it is whether capital punishment is “a punishment no longer consistent with our own self-respect” 2 and, therefore, violative of the Eighth Amendment.

The elasticity of the constitutional provision under consideration presents dangers of too little or too much self-restraint.3 Hence, we must proceed with caution to answer the question presented.4 By first examining the historical derivation of the Eighth Amendment and *316the construction given it in the past by this Court, and then exploring the history and attributes of capital punishment in this country, we can answer the question presented with objectivity and a proper measure of self-restraint.

Candor is critical to such an inquiry. All relevant material must be marshaled and sorted and forthrightly examined. We must not only be precise as to the standards of judgment that we are utilizing, but exacting in examining the relevant material in light of those standards.

Candor compels me to confess that I am not oblivious to the fact that this is truly a matter of life and death. Not only does it involve the lives of these three petitioners, but those of the almost 600 other condemned men and women in this country currently awaiting execution. While this fact cannot affect our ultimate decision, it necessitates that the decision be free from any possibility of error.

I

The Eighth Amendment's ban against cruel and unusual punishments derives from English law. In 1583, John Whitgift, Archbishop of Canterbury, turned the High Commission into a permanent ecclesiastical court, and the Commission began to use torture to extract confessions from persons suspected of various offenses.5 Sir Robert Beale protested that cruel and barbarous torture violated Magna Carta, but his protests were made in vain.6

*317Cruel punishments were not confined to those accused of crimes, but were notoriously applied with even greater relish to those who were convicted. Blackstone described in ghastly detail the myriad of inhumane forms of punishment imposed on persons found guilty of any of a large number of offenses.7 Death, of course, was the usual result.8

The treason trials of 1685 — the “Bloody Assizes”— which followed an abortive rebellion by the Duke of Monmouth, marked the culmination of the parade of horrors, and most historians believe that it was this event that finally spurred the adoption of the English Bill of Rights containing the progenitor of our prohibition against cruel and unusual punishments.9 The conduct of Lord Chief Justice Jeffreys at those trials has been described as an “insane lust for cruelty” which was “stimulated by orders from the King” (James II).10 The assizes received wide publicity from Puritan pamphleteers and doubtless had some influence on the adoption of a cruel and unusual punishments clause. But, *318the legislative history of the English Bill of Rights of 1689 indicates that the assizes may not have been as critical to the adoption of the clause as is widely thought. After William and Mary of Orange crossed the channel to invade England, James II fled. Parliament was summoned into session and a committee was appointed to draft general statements containing “such things as are absolutely necessary to be considered for the better securing of our religion, laws and liberties.”11 An initial draft of the Bill of Rights prohibited “illegal” punishments, but a later draft referred to the infliction by James II of “illegal and cruel” punishments, and declared “cruel and unusual” punishments to be prohibited.12 The use of the word “unusual” in the final draft appears to be inadvertent.

This legislative history has led at least one legal historian to conclude “that the cruel and unusual punishments clause of the Bill of Rights of 1689 was, first, an objection to the imposition of punishments that were unauthorized by statute and outside the jurisdiction of the sentencing court, and second, a reiteration of the English policy against disproportionate penalties,” 13 and not primarily a reaction to the torture of the High Commission, harsh sentences, or the assizes.

*319Whether the English Bill of Rights prohibition against cruel and unusual punishments is properly read as a response to excessive or illegal punishments, as a reaction to barbaric and objectionable modes of punishment, or as both, there is no doubt whatever that in borrowing the language and in including it in the Eighth Amendment, our Founding Fathers intended to outlaw torture and other cruel punishments.14

The precise language used in the Eighth Amendment first appeared in America on June 12, 1776, in Virginia’s “Declaration of Rights,” § 9 of which read: “That excessive bail ought not to be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted.” 15 This language was drawn verbatim from the English Bill of Rights of 1689. Other States adopted similar clauses,16 and there is evidence in the debates of the various state conventions that were *320called upon to ratify the Constitution of great concern for the omission of any prohibition against torture or other cruel punishments.17

The Virginia Convention offers some clues as to what the Founding Fathers had in mind in prohibiting cruel and unusual punishments. At one point George Mason advocated the adoption of a Bill of Rights, and Patrick Henry concurred, stating:

“By this Constitution, some of the best barriers of human rights are thrown away. Is there not an additional reason to have a bill of rights? . . . Congress, from their general powers, may fully go into business of human legislation. They may legislate, in criminal cases, from treason to the lowest offence — petty larceny. They may define crimes and prescribe punishments. In the definition of crimes, I trust they will be directed by what wise representatives ought to be governed by. But, when we come to punishments, no latitude ought to be left, nor dependence put on the virtue of representatives. What says our bill of rights? — ‘that excessive bail ought not to be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted.’ Are you not, therefore, now calling on those gentlemen who are to compose Congress, to prescribe trials and define punishments without this control? Will they find sentiments there similar to this bill of rights? You let them loose; you do more — you depart from the genius of your country. . . .

“In this business of legislation, your members of Congress will loose the restriction of not imposing excessive fines, demanding excessive bail, and in*321flicting cruel and unusual punishments. These are prohibited by your declaration of rights. What has distinguished our ancestors? — That they would not admit of tortures, or cruel and barbarous punishment. But Congress may introduce the practice of the civil law, in preference to that of the common law. They may introduce the practice of France, Spain, and Germany — of torturing, to extort a confession of the crime. They will say that they might as well draw examples from those countries as from Great Britain, and they will tell you that there is such a necessity of strengthening the arm of government, that they must have a criminal equity, and extort confession by torture, in order to punish with still more relentless severity. We are then lost and undone.” 18

Henry’s statement indicates that he wished to insure that “relentless severity” would be prohibited by the Constitution. Other expressions with respect to the proposed Eighth Amendment by Members of the First Congress indicate that they shared Henry’s view of the need for and purpose of the Cruel and Unusual Punishments Clause.19

*322Thus, the history of the clause clearly establishes that it was intended to prohibit cruel punishments. We must now turn to the case law to discover the manner in which courts have given meaning to the term “cruel.”

II

This Court did not squarely face the task of interpreting the cruel and unusual punishments language for the first time until Wilkerson v. Utah, 99 U. S. 130 (1879), although the language received a cursory examination in several prior cases. See, e. g., Pervear v. Commonwealth, 5 Wall. 475 (1867). In Wilkerson, the Court unanimously upheld a sentence of public execution by shooting imposed pursuant to a conviction for premeditated murder. In his opinion for the Court, Mr. Justice Clifford wrote:

“Difficulty would attend the effort to define with exactness the extent of the constitutional provision which provides that cruel and unusual punishments shall not be inflicted; but it is safe to affirm that punishments of torture, . . . and all others in the same line of unnecessary cruelty, are forbidden by that amendment to the Constitution.” 99 U. S., at 135-136.

Thus, the Court found that unnecessary cruelty was no more permissible than torture. To determine whether the punishment under attack Was unnecessarily cruel, the Court examined the history of the Utah Territory and the then-current writings on capital punishment, and compared this Nation’s practices with those of other countries. It is apparent that the Court felt it could not dispose of the question simply by referring to traditional practices; instead, it felt bound to examine developing thought.

Eleven years passed before the Court again faced a challenge to a specific punishment under the Eighth *323Amendment. In the case of In re Kemmler, 136 U. S. 436 (1890), Chief Justice Fuller wrote an opinion for a unanimous Court upholding electrocution as a permissible mode of punishment. While the Court ostensibly held that the Eighth Amendment did not apply' to the States, it is very apparent that the nature of the punishment involved was examined under the Due Process Clause of the Fourteenth Amendment. The Court held that the punishment was not objectionable. Today, Kemmler stands primarily for the proposition that a punishment is not necessarily unconstitutional simply because it is unusual, so long as the legislature has a humane purpose in selecting it.20

Two years later in O’Neil v. Vermont, 144 U. S. 323 (1892), the Court reaffirmed that the Eighth Amendment was not applicable to the States. O’Neil was found guilty on 307 counts of selling liquor in violation of Vermont law. A fine of $6,140 ($20 for each offense) and the costs of prosecution ($497.96) were imposed. O’Neil was committed to prison until the fine and the costs were paid; and the court provided that if they were not paid before a specified date, O’Neil was to be confined in the house of corrections for 19,914 days (approximately 54 years) at hard labor. Three Justices — Field, Harlan, and Brewer — dissented. They maintained not only that the Cruel and Unusual Punishments Clause was applicable to the States, but that in O’Neil’s case it had been violated. Mr. Justice Field wrote:

“That designation [cruel and unusual], it is true, is usually applied to punishments which inflict torture, such as the rack, the thumbscrew, the iron boot, the stretching of limbs and the like, which *324are attended with acute pain and suffering. . . . The inhibition is directed, not only against punishments of the character mentioned, but against all punishments which by their excessive length or severity are greatly disproportioned to the offences charged. The whole inhibition is against that which is excessive . . . Id., at 339-340.

In Howard v. Fleming, 191 U. S. 126 (1903), the Court, in essence, followed the approach advocated by the dissenters in O’Neil. In rejecting the claim that 10-year sentences for conspiracy to defraud were cruel and unusual, the Court (per Mr. Justice Brewer) considered the nature of the crime, the purpose of the law, and the length of the sentence imposed.

The Court used the same approach seven years later in the landmark case of Weems v. United States, 217 U. S. 349 (1910). Weems, an officer of the Bureau of Coast Guard and Transportation of the United States Government of the Philippine Islands, was convicted of falsifying a “public and official document.” He was sentenced to 15 years’ incarceration at hard labor with chains on his ankles, to an unusual loss of his civil rights, and to perpetual surveillance. Called upon to determine whether this was a cruel and unusual punishment, the Court found that it was.21 The Court emphasized that the Constitution was not an “ephemeral” enactment, or one “designed to meet passing occasions.” 22 Recognizing that “[t]ime works changes, [and] brings into existence new conditions and purposes,” 23 the Court commented that “[i]n the application of a constitu*325tion . . . our contemplation cannot be only of what has been but of what may be.” 24

In striking down the penalty imposed on Weems, the Court examined the punishment in relation to the offense, compared the punishment to those inflicted for other crimes and to those imposed in other jurisdictions, and concluded that the punishment was excessive.25 Justices White and Holmes dissented and argued that the cruel and unusual prohibition was meant to prohibit only those things that were objectionable at the time the Constitution was adopted.26

Weems is a landmark case because it represents the first time that the Court invalidated a penalty prescribed by a legislature for a particular offense. The Court made it plain beyond any reasonable doubt that excessive punishments were as objectionable as those that were inherently cruel. Thus, it is apparent that the dissenters’ position in O’Neil had become the opinion of the Court in Weems.

Weems was followed by two cases that added little to our knowledge of the scope of the cruel and unusual language, Badders v. United States, 240 U. S. 391 (1916), and United States ex rel. Milwaukee Social Democratic Publishing Co. v. Burleson, 255 U. S. 407 (1921).27 Then *326came another landmark case, Louisiana ex rel. Francis v. Resweber, 329 U. S. 459 (1947).

Francis had been convicted of murder and sentenced to be electrocuted. The first time the current passed through him, there was a mechanical failure and he did not die. Thereafter, Francis sought to prevent a second electrocution on the ground that it would be a cruel and unusual punishment. Eight members of the Court assumed the applicability of the Eighth Amendment to the States.28 The Court was virtually unanimous in agreeing that “[t]he traditional humanity of modern Anglo-American law forbids the infliction of unnecessary pain,” 29 but split 5-4 on whether Francis would, under the circumstances, be forced to undergo any excessive pain. Five members of the Court treated the case like In re Kemmler and held that the legislature adopted electrocution for a humane purpose, and that its will should not be thwarted because, in its desire to reduce pain and suffering in most cases, it may have inadvertently increased suffering in one particular case.30 *327The four dissenters felt that the case should be remanded for further facts.

As in Weems, the Court was concerned with excessive punishments. Resweber is perhaps most significant because the analysis of cruel and unusual punishment questions first advocated by the dissenters in O’Neil was at last firmly entrenched in the minds of an entire Court.

Trop v. Dulles, 356 U. S. 86 (1958), marked the next major cruel and unusual punishment case in this Court. Trop, a native-born American, was declared to have lost his citizenship by reason of a conviction by court-martial for wartime desertion. Writing for himself and Justices Black, Douglas, and Whittaker, Chief Justice Warren concluded that loss of citizenship amounted to a cruel and unusual punishment that violated the Eighth Amendment.31

Emphasizing the flexibility inherent in the words “cruel and unusual,” the Chief Justice wrote that “{t]he Amendment must draw its meaning from the evolving standards of decency that mark the progress of a maturing society.” 32 His approach to the problem was that utilized by the Court in Weems: he scrutinized the severity of the penalty in relation to the offense, examined the practices of other civilized nations of the world, and concluded that involuntary statelessness was an excessive and, therefore, an unconstitutional punishment. Justice Frankfurter, dissenting, urged that expatriation was not punishment, and that even if it were, it was not excessive. While he criticized the conclusion arrived at by the Chief Justice, his approach to the Eighth Amendment question was identical.

*328Whereas in Trop a majority of the Court failed to agree on whether loss of citizenship was a cruel and unusual punishment, four years later a majority did agree in Robinson v. California, 370 U. S. 660 (1962), that a sentence of 90 days’ imprisonment for violation of a California statute making it a crime to “be addicted to the use of narcotics” was cruel and unusual. Mr. Justice Stewart, writing the opinion of the Court, reiterated what the Court had said in Weems and what Chief Justice Warren wrote in Trop — that the cruel and unusual punishment clause was not a static concept, but one that must be continually re-examined “in the light of contemporary human knowledge.” 33 The fact that the penalty under attack was only 90 days evidences the Court’s willingness to carefully examine the possible excessiveness of punishment in a given case even where what is involved is a penalty that is familiar and widely accepted.34

We distinguished Robinson in Powell v. Texas, 392 U. S. 514 (1968), where we sustained a conviction for drunkenness in a public place and a fine of $20. Four Justices dissented on the ground that Robinson was controlling. The analysis in both cases was the same; only the conclusion as to whether or not the punishment was excessive differed. Powell marked the last time prior to today’s decision that the Court has had occasion to construe the meaning of the term “cruel and unusual” punishment.

Several principles emerge from these prior cases and serve as a beacon to an enlightened decision in the instant cases.

*329III

Perhaps the most important principle in analyzing “cruel and unusual” punishment questions is one that is reiterated again and again in the prior opinions of the Court: i. e., the cruel and unusual language “must draw its meaning from the evolving standards of decency that mark the progress of a maturing society.” 35 Thus, a penalty that was permissible at one time in our Nation’s history is not necessarily permissible today.

The fact, therefore, that the Court, or individual Justices, may have in the past expressed an opinion that the death penalty is constitutional is not now binding on us. A fair reading of Wilkerson v. Utah, supra; In re Kemmler, supra; and Louisiana ex rel. Francis v. Resweber, supra, would certainly indicate an acceptance sub silentio of capital punishment as constitutionally permissible. Several Justices have also expressed their individual opinions that the death penalty is constitutional.36 Yet, some of these same Justices and others have at times expressed concern over capital punishment.37 *330There is no holding directly in point, and the very nature of the Eighth Amendment would dictate that unless a very recent decision existed, stare decisis would bow to changing values, and the question of the constitutionality of capital punishment at a given moment in history would remain open.

Faced with an open question, we must establish our standards for decision. The decisions discussed in the previous section imply that a punishment may be deemed cruel and unusual for any one of four distinct reasons.

First, there are certain punishments that inherently involve so much physical pain and suffering that civilized people cannot tolerate them — e. g., use of the rack, the thumbscrew, or other modes of torture. See O’Neil v. Vermont, 144 U. S., at 339 (Field, J., dissenting). Regardless of public sentiment with respect to imposition of one of these punishments in a particular case or at any one moment in history, the Constitution prohibits it. These are punishments that have been barred since the adoption of the Bill of Rights.

*331Second, there are punishments that are unusual, signifying that they were previously unknown as penalties for a given offense. Cf. United States ex rel. Milwaukee Social Democratic Publishing Co. v. Burleson, 255 U. S., at 435 (Brandéis, J., dissenting). If these punishments are intended to serve a humane purpose, they may be constitutionally permissible. In re Kemmler, 136 U. S., at 447; Louisiana ex rel. Francis v. Resweber, 329 U. S., at 464. Prior decisions leave open the question of just how much the word “unusual” adds to the word “cruel.” I have previously indicated that use of the word “unusual” in the English Bill of Rights of 1689 was inadvertent, and there is nothing in the history of the Eighth Amendment to give flesh to its intended meaning. In light of the meager history that does exist, one would suppose that an innovative punishment would probably be constitutional if no more cruel than that punishment which it superseded. We need not decide this question here, however, for capital punishment is certainly not a recent phenomenon.

Third, a penalty may be cruel and unusual because it is excessive and serves no valid legislative purpose. Weems v. United States, supra. The decisions previously discussed are replete with assertions that one of the primary functions of the cruel and unusual punishments clause is to prevent excessive or unnecessary penalties, e. g., Wilkerson v. Utah, 99 U. S., at 134; O’Neil v. Vermont, 144 U. S., at 339-340 (Field, J., dissenting); Weems v. United States, 217 U. S., at 381; Louisiana ex rel. Francis v. Resweber, supra; these punishments are unconstitutional even though popular sentiment may favor them. Both The Chief Justice and Mr. Justice Powell seek to ignore or to minimize this aspect of the Court’s prior decisions. But, since Mr. Justice Field first suggested that “[t]he whole inhibition [of the prohibition against cruel and unusual punish*332ments] is against that which is excessive,” O’Neil v. Vermont, 144 U. S., at 340, this Court has steadfastly maintained that a penalty is unconstitutional whenever it is unnecessarily harsh or cruel. This is what the Founders of this country intended; this is what their fellow citizens believed the Eighth Amendment provided; and this was the basis for our decision in Robinson v. California, supra, for the plurality opinion by Mr. Chief Justice Warren in Trop v. Dulles, supra, and for the Court’s decision in Weems v. United States, supra. See also W. Bradford, An Enquiry How Far the Punishment of Death is Necessary in Pennsylvania (1793), reprinted in 12 Am. J. Legal Hist. 122, 127 (1968). It should also be noted that the "cruel and unusual” language of the Eighth Amendment immediately follows language that prohibits excessive bail and excessive fines. The entire thrust of the Eighth Amendment is, in short, against “that which is excessive.”

Fourth, where a punishment is not excessive and serves a valid legislative purpose, it still may be invalid if popular sentiment abhors it. For example, if the evidence clearly demonstrated that capital punishment served valid legislative purposes, such punishment would, nevertheless, be unconstitutional if citizens found it to be morally unacceptable. A general abhorrence on the part of the public would, in effect, equate a modern punishment with those barred since the adoption of the Eighth Amendment. There are no prior cases in this Court striking down a penalty on this ground, but the very notion of changing values requires that we recognize its existence.

It is immediately obvious, then, that since capital punishment is not a recent phenomenon, if it violates the Constitution, it does so because it is excessive or *333unnecessary, or because it is abhorrent to currently existing moral values.

We must proceed to the history of capital punishment in the United States.

IV

Capital punishment has been used to penalize various forms of conduct by members of society since the beginnings of civilization. Its precise origins are difficult to perceive, but there is some evidence that its roots lie in violent retaliation by members of a tribe or group, or by the tribe or group itself, against persons committing hostile acts toward group members.38 Thus, infliction of death as a penalty for objectionable conduct appears to have its beginnings in private vengeance.39

As individuals gradually ceded their personal prerogatives to a sovereign power, the sovereign accepted the authority to punish wrongdoing as part of its “divine right” to rule. Individual vengeance gave way to the vengeance of the state, and capital punishment became a public function.40 Capital punishment worked its way into the laws of various countries,41 and was inflicted in a variety of macabre and horrific ways.42

It was during the reign of Henry II (1154-1189) that English law first recognized that crime was more than a personal affair between the victim and the per*334petrator.43 The early history of capital punishment in England is set forth in McGautha v. California, 402 U. S. 183, 197-200 (1971), and need not be repeated here.

By 1500, English law recognized eight major capital crimes: treason, petty treason (killing of husband by his wife), murder, larceny, robbery, burglary, rape, and arson.44 Tudor and Stuart kings added many more crimes to the list of those punishable by death, and by 1688 there were nearly 50.45 George II (1727-1760) added nearly 36 more, and George III (1760-1820) increased the number by 60.46

By shortly after 1800, capital offenses numbered more than 200 and not only included crimes against person and property, but even some against the public peace. While England may, in retrospect, look particularly brutal, Blackstone points out that England was fairly civilized when compared to the rest of Europe.47

*335Capital punishment was not as common a penalty in the American Colonies. “The Capitall Lawes of New-England,” dating from 1636, were drawn by the Massachusetts Bay Colony and are the first written expression of capital offenses known to exist in this country. These laws make the following crimes capital offenses: idolatry, witchcraft, blasphemy, murder, assault in sudden anger, sodomy, buggery, adultery, statutory rape, rape, man-stealing, perjury in a capital trial, and rebellion. Each crime is accompanied by a reference to the Old Testament to indicate its source.48 It is not known with any certainty exactly when, or even if, these laws were enacted as drafted; and, if so, just how vigorously these laws were enforced.49 We do know that the other Colonies had a variety of laws that spanned the spectrum of severity.50

By the 18th century, the list of crimes became much less theocratic and much more secular. In the average colony, there were 12 capital crimes.51 This was far fewer than existed in England, and part of the reason was that there was a scarcity of labor in the Colonies.52 Still, there were many executions, because “[w]ith county jails inadequate and insecure, the criminal population seemed best controlled by death, mutilation, and fines.” 53

Even in the 17th century, there was some opposition *336to capital punishment in some of the colonies. In his “Great Act” of 1682, William Penn prescribed death only for premeditated murder and treason,54 although his reform was not long lived.55

In 1776 the Philadelphia Society for Relieving Distressed Prisoners organized, and it was followed 11 years later by the Philadelphia Society for Alleviating the Miseries of Public Prisons.56 These groups pressured for reform of all penal laws, including capital offenses. Dr. Benjamin Rush soon drafted America’s first reasoned argument against capital punishment, entitled An Enquiry into the Effects of Public Punishments upon Criminals and upon Society.57 In 1793, William Bradford, the Attorney General of Pennsylvania and later Attorney General of the United States, conducted “An Enquiry How Ear the Punishment of Death is Necessary in Pennsylvania.”58 He concluded that it was doubtful whether capital punishment was at all necessary, and that until more information could be obtained, it should be immediately eliminated for all offenses except high treason and murder.59

The “Enquiries” of Rush and Bradford and the Pennsylvania movement toward abolition of the death *337penalty had little immediate impact on the practices of other States.60 But in the early 1800’s, Governors George and DeWitt Clinton and Daniel Tompkins unsuccessfully urged the New York Legislature to modify or end capital punishment. During this same period, Edward Livingston, an American lawyer who later became Secretary of State and Minister to France under President Andrew Jackson, was appointed by the Louisiana Legislature to draft a new penal code. At the center of his proposal was “the total abolition of capital punishment.” 61 His Introductory Report to the System of Penal Law Prepared for the State of Louisiana62 contained a systematic rebuttal of all arguments favoring capital punishment. Drafted in 1824, it was not published until 1833. This work was a tremendous impetus to the abolition movement for the next half century.

During the 1830’s, there was a rising tide of sentiment against capital punishment. In 1834, Pennsylvania abolished public executions,63 and two years later, The Report on Capital Punishment Made to the Maine Legislature was published. It led to a law that prohibited the executive from issuing a warrant for execution within one year after a criminal was sentenced by the courts. The totally discretionary character of the law was at odds with almost all prior practices. The “Maine Law” resulted in little enforcement of the death penalty, which was not surprising since the legislature’s idea in passing the law was that the affirmative burden placed on the governor to issue a warrant one full year *338or more after a trial would be an effective deterrent to exercise of his power.64 The law spread throughout New England and led to Michigan's being the first State to abolish capital punishment in 1846.65

Anti-capital-punishment feeling grew in the 1840’s as the literature of the period pointed out the agony of the condemned man and expressed the philosophy that repentance atoned for the worst crimes, and that true repentance derived, not from fear, but from harmony with nature.66

By 1850, societies for abolition existed in Massachusetts, New York, Pennsylvania, Tennessee, Ohio, Alabama, Louisiana, Indiana, and Iowa.67 New York, Massachusetts, and Pennsylvania constantly had abolition bills before their legislatures. In 1852, Rhode Island followed in the footsteps of Michigan and partially abolished capital punishment.68 Wisconsin totally abolished the death penalty the following year.69 Those States that did not abolish the death penalty greatly reduced its scope, and “[f]ew states outside the South had more than one or two . . . capital offenses” in addition to treason and murder.70

But the Civil War halted much of the abolition furor. One historian has said that “[a]fter the Civil War, men’s finer sensibilities, which had once been revolted by the execution of a fellow being, seemed hardened and *339blunted.”71 Some of the attention previously given to abolition was diverted to prison reform. An abolitionist movement still existed, however. Maine abolished the death penalty in 1876, restored it in 1883, and abolished it again in 1887; Iowa abolished capital punishment from 1872-1878; Colorado began an erratic period of de jacto abolition and revival in 1872; and Kansas also abolished it de jacto in 1872, and by law in 1907.72

One great success of the abolitionist movement in the period from 1830-1900 was almost complete elimination of mandatory' capital punishment. Before the legislatures formally gave juries discretion to refrain from imposing the death penalty, the phenomenon of “jury nullification,” in which juries refused to convict in cases in which they believed that death was an inappropriate penalty, was experienced.73 Tennessee was the first State to give juries discretion, Tenn. Laws 1837-1838, c. 29, but other States quickly followed suit. Then, Rep. Curtis of New York introduced a federal bill that ultimately became law in 1897 which reduced the number of federal capital offenses from 60 to 3 (treason, murder, and rape) and gave the jury sentencing discretion in murder and rape cases.74

By 1917 12 States had become abolitionist jurisdictions.75 But, under the nervous tension of World War I, *340four of these States reinstituted capital punishment and promising movements in other States came grinding to a halt.76 During the period following the First World War, the abolitionist movement never regained its momentum.

It is not easy to ascertain why the movement lost its vigor. Certainly, much attention was diverted from penal reform during the economic crisis of the depression and the exhausting years of struggle during World War II. Also, executions, which had once been frequent public spectacles, became infrequent private affairs. The manner of inflicting death changed, and the horrors of the punishment were, therefore, somewhat diminished in the minds of the general public.77

In recent years there has been renewed interest in modifying capital punishment. New York has moved toward abolition,78 as have several other States.79 In 1967, a bill was introduced in the Senate to abolish *341capital punishment for all federal crimes, but it died in committee.80

At the present time, 41 States, the District of Columbia, and other federal jurisdictions authorize the death penalty for at least one crime. It would be fruitless to attempt here to categorize the approach to capital punishment taken by the various States.81 It is sufficient to note that murder is the crime most often punished by death, followed by kidnaping and treason.82 Rape is a capital offense in 16 States and the federal system.83

The foregoing history demonstrates that capital punishment was carried from Europe to America but, once here, was tempered considerably. At times in our history, strong abolitionist movements have existed. But, they have never been completely successful, as no more than one-quarter of the States of the Union have, at any one time, abolished the death penalty. They have had partial success, however, especially in reducing the number of capital crimes, replacing mandatory death sentences with jury discretion, and developing more humane methods of conducting executions.

This is where our historical foray leads. The question now to be faced is whether American society has *342reached a point where abolition is not dependent on a successful grass roots movement in particular jurisdictions, but is demanded by the Eighth Amendment. To answer this question, we must first examine whether or not the death penalty is today tantamount to excessive punishment.

V

In order to assess whether or not death is an excessive or unnecessary penalty, it is necessary to consider the reasons why a legislature might select it as.punishment for one or more offenses, and examine whether less severe penalties would satisfy the legitimate legislative wants as well as capital punishment. If they would, then the death penalty is unnecessary cruelty, and, therefore, unconstitutional.

There are six purposes conceivably served by capital punishment: retribution, deterrence, prevention of repetitive criminal acts, encouragement of guilty pleas and confessions, eugenics, and economy. These are considered seriatim, below.

A. The concept of retribution is one of the most misunderstood in all of our criminal jurisprudence. The principal source of confusion derives from the fact that, in dealing with the concept, most people confuse the question “why do men in fact punish?” with the question “what justifies men in punishing?” 84 Men may punish for any number of reasons, but the one reason that punishment is morally good or morally justifiable is that someone has broken the law. Thus, it can correctly be said that breaking the law is the sine qua non of punishment, or, in other words, that we only *343tolerate punishment as it is imposed on one who deviates from the norm established by the criminal law.

The fact that the State may seek retribution against those who have broken its laws does not mean that retribution may then become the State’s sole end in punishing. Our jurisprudence has always accepted deterrence in general, deterrence of individual recidivism, isolation of dangerous persons, and rehabilitation as proper goals of punishment. See Trop v. Dulles, 356 U. S., at 111 (Brennan, J., concurring). Retaliation, vengeance, and retribution have been roundly condemned as intolerable aspirations for a government in a free society.

Punishment as retribution has been condemned by scholars for centuries,85 and the Eighth Amendment itself was adopted to prevent punishment from becoming synonymous with vengeance.

In Weems v. United States, 217 U. S., at 381, the Court, in the course of holding that Weems’ punishment violated the Eighth Amendment, contrasted it with penalties provided for other offenses and concluded:

“[Tjhis contrast shows more than different exercises of legislative judgment. It is greater than that. It condemns the sentence in this case as cruel and unusual. It exhibits a difference between unrestrained power and that which is exercised under the spirit of constitutional limitations formed to establish justice. The State thereby suffers nothing and loses no power. The purpose of punishment is fulfilled, crime is repressed by penalties of just, not tormenting, severity, its repetition is prevented, and hope is given for the reformation of the criminal.” (Emphasis added.)

*344It is plain that the view of the Weems Court was that punishment for the sake of retribution was not permissible under the Eighth Amendment. This is the only view that the Court could have taken if the “cruel and unusual” language were to be given any meaning. Retribution surely underlies the imposition of some punishment on one who commits a criminal act. But, the fact that some punishment may be imposed does not mean that any punishment is permissible. If retribution alone could serve as a justification for any particular penalty, then all penalties selected by the legislature would by definition be acceptable means for designating society’s moral approbation of a particular act. The “cruel and unusual” language would thus be read out of the Constitution and the fears of Patrick Henry and the other Pounding Fathers would become realities.

To preserve the integrity of the Eighth Amendment, the Court has consistently denigrated retribution as a permissible goal of punishment.86 It is undoubtedly correct that there is a demand for vengeance on the part of many persons in a community against one who is convicted of a particularly offensive act. At times a cry is heard that morality requires vengeance to evi*345dence society’s abhorrence of the act.87 But the Eighth Amendment is our insulation from our baser selves. The “cruel and unusual” language limits the avenues through which vengeance can be channeled. Were this not so, the language would be empty and a return to the rack and other tortures would be possible in a given case.

Mr. Justice Story wrote that the Eighth Amendment’s limitation on punishment “would seem to be wholly unnecessary in a free government, since it is scarcely possible that any department of such a government should authorize or justify such atrocious conduct.”88

I would reach an opposite conclusion — that only in a free society would men recognize their inherent weaknesses and seek to compensate for them by means of a Constitution.

The history of the Eighth Amendment supports only the conclusion that retribution for its own sake is improper.

B. The most hotly contested issue regarding capital punishment is whether it is better than life imprisonment as a deterrent to crime.89

While the contrary position has been argued,90 it is my firm opinion that the death penalty is a more severe sanction than life imprisonment. Admittedly, there are *346some persons who would rather die than languish in prison for a lifetime. But, whether or not they should be able to choose death as an alternative is a far different question from that presented here — i. e., whether the State can impose death as a punishment. Death is irrevocable; life imprisonment is not. Death, of course, makes rehabilitation impossible; life imprisonment does not. In short, death has always been viewed as the ultimate sanction, and it seems perfectly reasonable to continue to view it as such.91

It must be kept in mind, then, that the question to be considered is not simply whether capital punishment is *347a deterrent, but whether it is a better deterrent than life imprisonment.92

There is no more complex problem than determining the deterrent efficacy of the death penalty. “Capital punishment has obviously failed as a deterrent when a murder is committed. We can number its failures. But we cannot number its successes. No one can ever know how many people have refrained from murder because of the fear of being hanged.” 93 This is the nub of the problem and it is exacerbated by the paucity of useful data. The United States is more fortunate than most countries, however, in that it has what are generally considered to be the world’s most reliable statistics.94

The two strongest arguments in favor of capital punishment as a deterrent are both logical hypotheses devoid of evidentiary support, but persuasive nonetheless. The first proposition was best stated by Sir James Stephen in 1864:

“No other punishment deters men so effectually from committing crimes as the punishment of death. This is one of those propositions which it is difficult to prove, simply because they are in themselves more obvious than any proof can make them. It is possible to display ingenuity in arguing against it, but that is all. The whole experience of mankind is in the other direction. The threat of instant death is the one to which resort has always been made when there was an absolute necessity for producing some result. . . . No one goes to certain *348inevitable death except by compulsion. Put the matter the other way. Was there ever yet a criminal who, when sentenced to death and brought out to die, would refuse the offer of a commutation of his sentence for the severest secondary punishment? Surely not. Why is this? It can only be because 'All that a man has will he give for his life.’ In any secondary punishment, however terrible, there is hope; but death is death; its terrors cannot be described more forcibly.” 95

This hypothesis relates to the use of capital punishment as a deterrent for any crime. The second proposition is that "if life imprisonment is the maximum penalty for a crime such as murder, an offender who is serving a life sentence cannot then be deterred from murdering a fellow inmate or a prison officer.”96 This hypothesis advocates a limited deterrent effect under particular circumstances.

Abolitionists attempt to disprove these hypotheses by amassing statistical evidence to demonstrate that there is no correlation between criminal activity and the existence or nonexistence of a capital sanction. Almost all of the evidence involves the crime of murder, since murder is punishable by death in more jurisdictions than are other offenses,97 and almost 90%. of all executions since 1930 have been pursuant to murder convictions.98

Thorsten Sellin, one of the leading authorities on capital punishment, has urged that if the death penalty *349deters prospective murderers, the following hypotheses should be true:

“(a) Murders should be less frequent in states that have the death penalty than in those that have abolished it, other factors being equal. Comparisons of this nature must be made among states that are as alike as possible in all other respects— character of population, social and economic condition, etc. — in order not to introduce factors known to influence murder rates in a serious manner but present in only one of these states.

“(b) Murders should increase when the death penalty is abolished and should decline when it is restored.

“(c) The deterrent effect should be greatest and should therefore affect murder rates most powerfully in those communities where the crime occurred and its consequences are most strongly brought home to the population.

“(d) Law enforcement officers would be safer from murderous attacks in states that have the death penalty than in those without it.”99 (Footnote omitted.)

Sellin’s evidence indicates that not one of these propositions is true. This evidence has its problems, however. One is that there are no accurate figures for capital murders; there are only figures on homicides and they, of course, include noncapital killings.100 A second problem is that certain murders undoubtedly are misinterpreted as accidental deaths or suicides, and there *350is no way of estimating the number of such undetected crimes. A third problem is that not all homicides are reported. Despite these difficulties, most authorities have assumed that the proportion of capital murders in a State’s or nation’s homicide statistics remains reasonably constant,101 and that the homicide statistics are therefore useful.

Sellin’s statistics demonstrate that there is no correlation between the murder rate and the presence or absence of the capital sanction. He compares States that have similar characteristics and finds that irrespective of their position on capital punishment, they have similar murder rates. In the New England States, for example, there is no correlation between executions102 and homicide rates.103 The same is true for Midwestern States,104 and for all others studied. Both the United Nations105 and Great Britain 100 have acknowledged the validity of Sellin’s statistics.

Sellin also concludes that abolition and/or reintroduction of the death penalty had no effect on the homicide rates of the various States involved.107 This conclusion is borne out by others who have made similar *351inquiries108 and by the experience of other countries.109 Despite problems with the statistics,110 Sellin’s evidence has been relied upon in international studies of capital punishment.111

Statistics also show that the deterrent effect of capital punishment is no greater in those communities where executions take place than in other communities.112 In fact, there is some evidence that imposition of capital punishment may actually encourage crime, rather than deter it.113 And, while police and law enforcement offi*352cers are the strongest advocates of capital punishment,114 the evidence is overwhelming that police are no safer in communities that retain the sanction than in those that have abolished it.115

There is also a substantial body of data showing that the existence of the death penalty has virtually no effect on the homicide rate in prisons.110 Most of the persons sentenced to death are murderers, and murderers tend to be model prisoners.117

*353In sum, the only support for the theory that capital punishment is an effective deterrent is found in the hypotheses with which we began and the occasional stories about a specific individual being deterred from doing a contemplated criminal act.118 These claims of specific deterrence are often spurious,119 however, and may be more than counterbalanced by the tendency of capital punishment to incite certain crimes.120

The United Nations Committee that studied capital punishment found that “'[i]t is generally agreed between the retentionists and abolitionists, whatever their opinions about the validity of comparative studies of deterrence, that the data which now exist show no correlation between the existence of capital punishment and lower rates of capital crime.”121

Despite the fact that abolitionists have not proved non-deterrence beyond a reasonable doubt, they have succeeded in showing by clear and convincing evidence that capital punishment is not necessary as a deterrent to crime in our society. This is all that they must do. We would shirk our judicial responsibilities if we failed to accept the presently existing statistics and demanded more proof. It may be that we now possess all the proof that anyone could ever hope to assemble on the subject. But, even if further proof were to be forthcoming, I believe there is more than enough evidence presently available for a decision in this case.

In 1793 William Bradford studied the utility of the death penalty in Pennsylvania and found that it probably had no deterrent effect but that more evidence *354was needed.122 Edward Livingston reached a similiar conclusion with respect to deterrence in 1833 upon completion of his study for Louisiana.123 Virtually every study that has since been undertaken has reached the same result.124

In light of the massive amount of evidence before us, I see no alternative but to conclude that capital punishment cannot be justified on the basis of its deterrent effect.125

*355C. Much of what must be said about the death penalty as a device to prevent recidivism is obvious — if a murderer is executed, he cannot possibly commit another offense. The fact is, however, that murderers are extremely unlikely to commit other crimes either in prison or upon their release.126 For the most part, they are first offenders, and when released from prison they are known to become model citizens.127 Furthermore, most persons who commit capital crimes are not executed. With respect to those who are sentenced to die, it is critical to note that the jury is never asked to determine whether they are likely to be recidivists. In light of these facts, if capital punishment were justified purely on the basis of preventing recidivism, it would have to be considered to be excessive; no general need to obliterate all capital offenders could have been demonstrated, nor any specific need in individual cases.

D. The three final purposes which may underlie utilization of a capital sanction — encouraging guilty pleas and confessions, eugenics, and reducing state "expenditures — may be dealt with quickly. If the death penalty is used to encourage guilty pleas and thus to deter suspects from exercising their rights under the Sixth Amendment to jury trials, it is unconstitutional. United States *356v. Jackson, 390 U. S. 570 (1968).128 Its elimination would do little to impair the State’s bargaining position in criminal cases, since life imprisonment remains a severe sanction which can be used as leverage for bargaining for pleas or confessions in exchange either for charges of lesser offenses or recommendations of leniency.

Moreover, to the extent that capital punishment is used to encourage confessions and guilty pleas, it is not being used for punishment purposes. A State that justifies capital punishment on its utility as part of the conviction process could not profess to rely on capital punishment as a deterrent. Such a State’s system would be structured with twin goals only: obtaining guilty pleas and confessions and imposing imprisonment as the maximum sanction. Since life imprisonment is sufficient for bargaining purposes, the death penalty is excessive if used for the same purposes.

In light of the previous discussion on deterrence, any suggestions concerning the eugenic benefits of capital punishment are obviously meritless.129 As I pointed out above, there is not even any attempt made to discover which capital offenders are likely to be recidivists, let alone which are positively incurable. No test or procedure presently exists by which incurables can be screened from those who would benefit from treatment. On the one hand, due process would seem to require that we have some procedure to demonstrate incurability before execution; and, on the other hand, equal protection would then seemingly require that all incurables be executed, cf. Skinner v. Oklahoma, 316 U. S. 535 (1942). In addition, the “cruel and unusual” language *357would require that life imprisonment, treatment, and sterilization be inadequate for eugenic purposes. More importantly, this Nation has never formally professed eugenic goals, and the history of the world does not look kindly on them. If eugenics is one of our purposes, then the legislatures should say so forthrightly and design procedures to serve this goal. Until such time, I can only conclude, as has virtually everyone else who has looked at the problem,130 that capital punishment cannot be defended on the basis of any eugenic purposes.

As for the argument that it is cheaper to execute a capital offender than to imprison him for life, even assuming that such an argument, if true, would support. a capital sanction, it is simply incorrect. A disproportionate amount of money spent on prisons is attributable to death row.131 Condemned men are not productive members of the prison community, although they could be,132 and executions are expensive.133 Appeals are often automatic, and courts admittedly spend more time with death cases.134

*358At trial, the selection of jurors is likely to become a costly, time-consuming problem in a capital case,135 and defense counsel will reasonably exhaust every possible means to save his client from execution, no matter how long the trial takes.

During the period between conviction and execution, there are an inordinate number of collateral attacks on the conviction and attempts to obtain executive clemency, all of which exhaust the time, money, and effort of the State. There are also continual assertions that the condemned prisoner has gone insane.136 Because there is a formally established policy of not executing insane persons,137 great sums of money may be spent on detecting and curing mental illness in order to perform the execution.138 Since no one wants the responsibility for the execution, the condemned man is likely to be passed back and forth from doctors to custodial officials to courts like a ping-pong ball.139 The entire process is very costly.

When all is said and done, there can be no doubt that it costs more to execute a man than to keep him in prison for life.140

E. There is but one conclusion that can be drawn from all of this — i. e., the death penalty is an excessive and unnecessary punishment that violates the Eighth *359Amendment. The statistical evidence is not convincing beyond all doubt, but it is persuasive. It is not improper at this point to take judicial notice of the fact that for more than 200 years men have labored to demonstrate that capital punishment serves no purpose that life imprisonment could not serve equally well. And they have done so with great success. Little, if any, evidence has been adduced to prove the contrary. The point has now been reached at which deference to the legislatures is tantamount to abdication of our judicial roles as factfinders, judges, and ultimate arbiters of the Constitution. We know that at some point the presumption of constitutionality accorded legislative acts gives way to a realistic assessment of those acts. This point comes when there is sufficient evidence available so that judges can determine, not whether the legislature acted wisely, but whether it had any rational basis whatsoever for acting. We have this evidence before us now. There is no rational basis for concluding that capital punishment is not excessive. It therefore violates the Eighth Amendment.141

*360VI

In addition, even if capital punishment is not excessive, it nonetheless violates the Eighth Amendment because it is morally unacceptable to the people of the United States at this time in their history.

In judging whether or not a given penalty is morally acceptable, most courts have said that the punishment is valid unless “it shocks the conscience and sense of justice of the people.” 142

*361Judge Frank once noted the problems inherent in the use of such a measuring stick:

“[The court,] before it reduces a sentence as 'cruel and unusual/ must have reasonably good assurances that the sentence offends the 'common conscience.’ And, in any context, such a standard — the community’s attitude — is usually an unknowable. It resembles a slithery shadow, since one can seldom learn, at all accurately, what the community, or a majority, actually feels. Even a carefully-taken 'public opinion poll’ would be inconclusive in a case like this.” 143

While a public opinion poll obviously is of some assistance in indicating public acceptance or rejection of a specific penalty,144 its utility cannot be very great. This is because whether or not a punishment is cruel and unusual depends, not on whether its mere mention “shocks the conscience and sense of justice of the people,” but on whether people who were fully informed as to the purposes of the penalty and its liabilities would find the penalty shocking, unjust, and unacceptable.145

*362In other words, the question with which we must deal is not whether a substantial proportion of American citizens would today, if polled, opine that capital punishment is barbarously cruel, but whether they would find it to be so in the light of all information presently available.

This is not to suggest that with respect to this test of unconstitutionality people are required to act rationally; they are not. With respect to this judgment, a violation of the Eighth Amendment is totally dependent on the predictable subjective, emotional reactions of informed citizens.146

It has often been noted that American citizens know almost nothing about capital punishment.147 Some of the conclusions arrived at in the preceding section and the supporting evidence would be critical to an informed judgment on the morality of the death penalty: e. g., that the death penalty is no more effective a deterrent than life imprisonment, that convicted murderers are *363rarely executed, but are usually sentenced to a term in prison; that convicted murderers usually are model prisoners, and that they almost always become law-abiding citizens upon their release from prison; that the costs of executing a capital offender exceed the costs of imprisoning him for life; that while in prison, a convict under sentence of death performs none of the useful functions that life prisoners perform; that no attempt is made in the sentencing process to ferret out likely recidivists for execution; and that the death penalty may actually stimulate criminal activity.

This information would almost surely convince the average citizen that the death penalty was unwise, but a problem arises as to whether it would convince him that the penalty was morally reprehensible. This problem arises from the fact that the public’s desire for retribution, even though this is a goal that the legislature cannot constitutionally pursue as its sole justification for capital punishment, might influence the citizenry’s view of the morality of capital punishment. The solution to the problem lies in the fact that no one has ever seriously advanced retribution as a legitimate goal of our society. Defenses of capital punishment are always mounted on deterrent or other similar theories. This should not be surprising. It is the people of this country who have urged in the past that prisons rehabilitate as well as isolate offenders, and it is the people who have injected a sense of purpose into our penology. I cannot believe that at this stage in our history, the American people would ever knowingly support purposeless vengeance. Thus, I believe that the great mass of citizens would conclude on the basis of the material already considered that the death penalty is immoral and therefore unconstitutional.

But, if this information needs supplementing, I believe that the following facts would serve to convince *364even the most hesitant of citizens to condemn death as a sanction: capital punishment is imposed discrim-inatorily against certain identifiable classes of people; there is evidence that innocent people have been executed before their innocence can be proved; and the death penalty wreaks havoc with our entire criminal justice system. Each of these facts is considered briefly below.

Regarding discrimination, it has been said that “[i]t is usually the poor, the illiterate, the underprivileged, the member of the minority group — the man who, because he is without means, and is defended by a court-appointed attorney — who becomes society’s sacrificial lamb 148 Indeed, a look at the bare statistics regarding executions is enough to betray much of the discrimination. A total of 3,859 persons have been executed since 1930, of whom 1,751 were white and 2,066 were Negro.149 Of the executions, 3,334 were for murder; 1,664 of the executed murderers were white and 1,630 were Negro; 150 455 persons, including 48 whites and 405 Negroes, were executed for rape.151 It is immediately apparent that Negroes were executed far more often than whites in proportion to their percentage of the population. Studies indicate that while the higher rate of execution among Negroes is partially due to a higher rate of crime, there is evidence of racial discrimination.152 *365Racial or other discriminations should not be surprising. In McGautha v. California, 402 U. S., at 207, this Court held “that committing to the untrammeled discretion of the jury the power to pronounce life or death in capital cases is [not] offensive to anything in the Constitution.” This was an open invitation to discrimination.

There is also overwhelming evidence that the death penalty is employed against men and not women. Only 32 women have been executed since 1930, while 3,827 men have met a similar fate.153 It is difficult to understand why women have received such favored treatment since the purposes allegedly served by capital punishment seemingly are equally applicable to both sexes.154

It also is evident that the burden of capital punishment falls upon the poor, the ignorant, and the under*366privileged members of society.155 It is the poor, and the members of minority groups who are least able to voice their complaints against capital punishment. Their impotence leaves them victims of a sanction that the wealthier, better-represented, just-as-guilty person can escape. So long as the capital sanction is used only against the forlorn, easily forgotten members of society, legislators are content to maintain the status quo, because change would draw attention to the problem and concern might develop. Ignorance is perpetuated and apathy soon becomes its mate, and we have today’s situation.

Just as Americans know little about who is executed and why, they are unaware of the potential dangers of executing an innocent man. Our “beyond a reasonable doubt” burden of proof in criminal cases is intended to protect the innocent, but we know it is not foolproof. Various studies have shown that people whose innocence is later convincingly established are convicted and sentenced to death.156

*367Proving one’s innocence after a jury finding of guilt is almost impossible. While reviewing courts are willing to entertain all kinds of collateral attacks where a sentence of death is involved, they very rarely dispute the jury’s interpretation of the evidence. This is, perhaps, as it should be. But, if an innocent man has been found guilty, he must then depend on the good faith of the prosecutor’s office to help him establish his innocence. There is evidence, however, that prosecutors do not welcome the idea of having convictions, which they labored hard to secure, overturned, and that their cooperation is highly unlikely.157

No matter how careful courts are, the possibility of perjured testimony, mistaken honest testimony, and human error remain all too real.158 We have no way of *368judging how many innocent persons have been executed but we can be certain that there were some. Whether there were many is an open question made difficult by the loss of those who were most knowledgeable about the crime for which they were convicted. Surely there will be more as long as capital punishment remains part of our penal law.

While it is difficult to ascertain with certainty the degree to which the death penalty is discriminatorily imposed or the number of innocent persons sentenced to die, there is one conclusion about the penalty that is universally accepted — i. e., it “tends to distort the course of the criminal law.”159 As Mr. Justice Frankfurter said:

“I am strongly against capital punishment . . . . When life is at hazard in a trial, it sensationalizes the whole thing almost unwittingly; the effect on juries, the Bar, the public, the Judiciary, I regard as very bad. I think scientifically the claim of deterrence is not worth much. Whatever proof there may be in my judgment does not outweigh the social loss due to the inherent sensationalism of a trial for fife.” 160

*369The deleterious effects of the death penalty are also felt otherwise than at trial. For example, its very existence “inevitably sabotages a social or institutional program of reformation.” 161 In short “;[t]he presence of the death penalty as the keystone of our penal system bedevils the administration of criminal justice all the way down the line and is the stumbling block in the path of general reform and of the treatment of crime and criminals.” 162

Assuming knowledge of all the facts presently available regarding capital punishment, the average citizen would, in my opinion, find it shocking to his conscience and sense of justice.163 For this reason alone capital punishment cannot stand.

*370YII

To arrive at the conclusion that the death penalty violates the Eighth Amendment, we have had to engage in a long and tedious journey. The amount of information that we have assembled and sorted is enormous. *371Yet, I firmly believe that we have not deviated in the slightest from the principles with which we began.

At a time in our history when the streets of the Nation’s cities inspire fear and despair, rather than pride and hope, it is difficult to maintain objectivity and concern for our fellow citizens. But, the measure of a country’s greatness is its ability to retain compassion in time of crisis. No nation in the recorded history of man has a greater tradition of revering justice and fair treatment for all its citizens in times of turmoil, confusion, and tension than ours. This is a country which stands tallest in troubled times, a country that clings to fundamental principles, cherishes its constitutional heritage, and rejects simple solutions that compromise the values that lie at the roots of our democratic system.

In striking down capital punishment, this Court does not malign our system of government. On the contrary, it pays homage to it. Only in a free society could right triumph in difficult times, and could civilization record its magnificent advancement. In recognizing the humanity of our fellow beings, we pay ourselves the highest tribute. We achieve “a major milestone in the long road up from barbarism” 164 and join the approximately 70 other jurisdictions in the world which celebrate their regard for civilization and humanity by shunning capital punishment.165

I concur in the judgments of the Court.

[Appendices I, II, and III follow.]

*372APPENDIX I TO OPINION OF MARSHALL, J., CONCURRING

ABOLITION OF THE DEATH PENALTY IN THE UNITED STATES: 1846-1968

(States are listed according to year most recent action was taken)

Year of Year of restoration reabolition State Year of partial abolition Year of complete abolition

New York. 19651

Vermont. 19652

West Virginia. — 1965

Iowa . — 1872 1 — l co 05 C71 00 CO 1 — 1

Oregon . — 1914 l_* co 05 o <M 05 1 — Í

Michigan . 18473 1963

1961 Delaware . — 1958

Alaska . — 1957

Hawaii . — 1957

South Dakota. — Kansas . — Missouri. — Tennessee. 1915 4_ Washington . — Arizona . 19165 North Dakota. 19156 1915 1907 1917 1913 SO CO CO CO C0 CO MMMMWCO GO co co co C* CO

Minnesota . — 1911

1883 1887 Colorado. — Maine. — Wisconsin . — 1897 1876 1853 o 05

Rhode Island. 18527

*373APPENDIX II TO OPINION OF MARSHALL, J., CONCURRING

CRUDE HOMICIDE DEATH RATES, PER 100,000 POPULATION, AND NUMBER OF EXECUTIONS IN CERTAIN AMERICAN STATES: 1920-1955

*374APPENDIX III TO OPINION OF MARSHALL, J., CONCURRING

CRUDE HOMICIDE DEATH RATES, PER 100,000 POPULATION, AND NUMBER OF EXECUTIONS IN CERTAIN AMERICAN STATES: 1920-1955

*375Mr. Chief Justice Burger,

with whom Mr. Justice Blackmun, Mr. Justice Powell, and Mr. Justice Rehnquist join, dissenting.

At the outset it is important to note that only two members of the Court, Mr. Justice Brennan and Mr. Justice Marshall, have concluded that the Eighth Amendment prohibits capital punishment for all crimes and under all circumstances. Mr. Justice Douglas has also determined that the death penalty contravenes the Eighth Amendment, although I do not read his opinion as necessarily requiring final abolition of the penalty.1 For the reasons set forth in Parts I-IV of this opinion, I conclude that the constitutional prohibition against “cruel and unusual punishments” cannot be construed to bar the imposition of the punishment of death.

Mr. Justice Stewart and Mr. Justice White have concluded that petitioners’ death sentences must be set aside because prevailing sentencing practices do not comply with the Eighth Amendment. For the reasons set forth in Part V of this opinion, I believe this approach fundamentally misconceives the nature of the Eighth Amendment guarantee and flies directly in the face of controlling authority of extremely recent vintage.

I

If we were possessed of legislative power, I would either join with Mr. Justice Brennan and Mr. Justice Marshall or, at the very least, restrict the use of capital punishment to a small category of the most heinous crimes. Our constitutional inquiry, however, must be divorced from personal feelings as to the morality and efficacy of the death penalty, and be confined to the meaning and applicability of the uncertain language of the Eighth Amendment. There is no novelty in being called upon to interpret a constitutional provision that is less than *376self-defining, but, of all our fundamental guarantees, the ban on “cruel and unusual punishments” is one of the most difficult to translate into judicially manageable terms. The widely divergent views of the Amendment expressed in today’s opinions reveal the haze that surrounds this constitutional command. Yet it is essential to our role as a court that we not seize upon the enigmatic character of the guarantee as an invitation to enact our personal predilections into law.

Although the Eighth Amendment literally reads as prohibiting only those punishments that are both “cruel” and “unusual,” history compels the conclusion that the Constitution prohibits all punishments of extreme and barbarous cruelty, regardless of how frequently or infrequently imposed.

The most persuasive analysis of Parliament’s adoption of the English Bill of Rights of 1689 — the unquestioned source of the Eighth Amendment wording— suggests that the prohibition against “cruel and unusual punishments” was included therein out of aversion to severe punishments not legally authorized and not within the jurisdiction of the courts to impose. To the extent that the term “unusual” had any importance in the English version, it was apparently intended as a reference to illegal punishments.2

*377From every indication, the Framers of the Eighth Amendment intended to give the phrase a meaning far different from that of its English precursor. The records of the debates in several of the state conventions called to ratify the 1789 draft Constitution submitted prior to the addition of the Bill of Rights show that the Framers’ exclusive concern was the absence of any ban on tortures.3 The later inclusion of the “cruel and unusual punishments” clause was in response to these objections. There was no discussion of the interrelationship of the terms “cruel” and “unusual,” and there is nothing in the debates supporting the inference that the Founding Fathers would have been receptive to torturous or excessively cruel punishments even if usual in character or authorized by law.

The cases decided under the Eighth Amendment are consistent with the tone of the ratifying debates. In Wilkerson v. Utah, 99 U. S. 130 (1879), this Court held that execution by shooting was not a prohibited mode of carrying out a sentence of death. Speaking to the mean*378ing of the Cruel and Unusual Punishments Clause, the Court stated,

“[I]t is safe to affirm that punishments of torture . . . and all others in the same line of unnecessary cruelty, are forbidden by that amendment to the Constitution.” Id., at 136.

The Court made no reference to the role of the term “unusual” in the constitutional guarantee.

In the case of In re Kemmler, 136 U. S. 436 (1890), the Court held the Eighth Amendment inapplicable to the States and added the following dictum:

“So that, if the punishment prescribed for an of-fence against the laws of the State were manifestly cruel and unusual, as burning at the stake, crucifixion, breaking on the wheel, or the like, it would be the duty of the courts to adjudge such penalties to be within the . . . [prohibition of the New York constitution]. And we think this equally true of the Eighth Amendment, in its application to Congress.

“. . . Punishments are cruel when they involve torture or a lingering death; but the punishment of death is not cruel, within the meaning of that word as used in the Constitution. It implies there something inhuman and barbarous, something more than the mere extinguishment of life.” Id., at 446-447.

This language again reveals an exclusive concern with extreme cruelty. The Court made passing reference to the finding of the New York courts that electrocution was an “unusual” punishment, but it saw no need to discuss the significance of that term as used in the Eighth Amendment.

Opinions in subsequent cases also speak of extreme cruelty as though that were the sum and substance of the constitutional prohibition. See O’Neil v. Vermont, 144 U. S. 323, 339-340 (1892) (Field, J., dissenting); Weems *379v. United States, 217 U. S. 349, 372-373 (1910); Louisiana ex rel. Francis v. Resweber, 329 U. S. 459, 464 (1947). As summarized by Mr. Chief Justice Warren in the plurality opinion in Trop v. Dulles, 356 U. S. 86, 100 n. 32 (1958):

“Whether the word ‘unusual’ has any qualitative meaning different from ‘cruel’ is not clear. On the few occasions this Court has had to consider the meaning of the phrase, precise distinctions between cruelty and unusualness do not seem to have been drawn. See Weems v. United States, supra; O’Neil v. Vermont, supra; Wilkerson v. Utah, supra. These cases indicate that the Court simply examines the particular punishment involved in light of the basic prohibition against inhuman treatment, without regard to any subtleties of meaning that might be latent in the word ‘unusual.’ ”

I do not suggest that the presence of the word “unusual” in the Eighth Amendment is merely vestigial, having no relevance to the constitutionality of any punishment that might be devised. But where, as here, we consider a punishment well known to history, and clearly authorized by legislative enactment, it disregards the history of the Eighth Amendment and all the judicial comment that has followed to rely on the term “unusual” as affecting the outcome of these cases. Instead, I view these cases as turning on the single question whether capital punishment is “cruel” in the constitutional sense. The term “unusual” cannot be read as limiting the ban on “cruel” punishments or as somehow expanding the meaning of the term “cruel.” For this reason I am unpersuaded by the facile argument that since capital punishment has always been cruel in the everyday sense of the word, and has become unusual due to decreased use, it is, therefore, now “cruel and unusual.”

*380II

Counsel for petitioners properly concede that capital punishment was not impermissibly cruel at the time of the adoption of the Eighth Amendment. Not only do the records of the debates indicate that the Founding Fathers were limited in their concern to the prevention of torture, but it is also clear from the language of the Constitution itself that there was no thought whatever of the elimination of capital punishment. The opening sentence of the Fifth Amendment is a guarantee that the death penalty not be imposed “unless on a presentment or indictment of a Grand Jury.” The Double Jeopardy Clause of the Fifth Amendment is a prohibition against being “twice put in jeopardy of life” for the same offense. Similarly, the Due Process Clause commands “due process of law” before an accused can be “deprived of life, liberty, or property.” Thus, the explicit language of the Constitution affirmatively acknowledges the legal power to impose capital punishment; it does not expressly or by implication acknowledge the legal power to impose any of the various punishments that have been banned as cruel since 1791. Since the Eighth Amendment was adopted on the same day in 1791 as the Fifth Amendment, it hardly needs more to establish that the death penalty was not “cruel” in the constitutional sense at that time.

In the 181 years since the enactment of the Eighth Amendment, not a single decision of this Court has cast the slightest shadow of a doubt on the constitutionality of capital punishment. In rejecting Eighth Amendment attacks on particular modes of execution, the Court has more than once implicitly denied that capital punishment is impermissibly “cruel” in the constitutional sense. Wilkerson v. Utah, 99 U. S. 130 (1879); Louisiana ex rel. Francis v. Resweber, 329 U. S., at 464. In *381re Kemmler, 136 U. S. 436 (1890) (dictum). It is only 14 years since Mr. Chief Justice Warren, speaking for four members of the Court, stated without equivocation:

“Whatever the arguments may be against capital punishment, both on moral grounds and in terms of accomplishing the purposes of punishment — -and they are forceful — the death penalty has been employed throughout our history, and, in a day when it is still widely accepted, it cannot be said to violate the constitutional concept of cruelty.” Trop v. Dulles, 356 U. S., at 99.

It is only one year since Mr. Justice Black made his feelings clear on the constitutional issue:

“The Eighth Amendment forbids 'cruel and unusual punishments.’ In my view, these words cannot be read to outlaw capital punishment because that penalty was in common use and authorized by law here and in the countries from which our ancestors came at the time the Amendment was adopted. It is inconceivable to me that the framers intended to end capital punishment by the Amendment.” Mc-Gautha v. California, 402 U. S. 183, 226 (1971) (separate opinion).

By limiting its grants of certiorari, the Court has refused even to hear argument on the Eighth Amendment claim on two occasions in the last four years. Witherspoon v. Illinois, cert. granted, 389 U. S. 1035, rev’d, 391 U. S. 510 (1968); McGautha v. California, cert. granted, 398 U. S. 936 (1970), aff’d, 402 U. S. 183 (1971). In these cases the Court confined its attention to the procedural aspects of capital trials, it being implicit that the punishment itself could be constitutionally imposed. Nonetheless, the Court has now been asked to hold that a punishment clearly permissible under the Constitution at the time of its adoption and accepted as such by every *382member of the Court until today, is suddenly so cruel as to be incompatible with the Eighth Amendment.

Before recognizing such an instant evolution in the law, it seems fair to ask what factors have changed that capital punishment should now be “cruel” in the constitutional sense as it has not been in the past. It is apparent that there has been no change of constitutional significance in the nature of the punishment itself. Twentieth century modes of execution surely involve no greater physical suffering than the means employed at the time of the Eighth Amendment’s adoption. And although a man awaiting execution must inevitably experience extraordinary mental anguish4 no one suggests that this anguish is materially different from that experienced by condemned men in 1791, even though protracted appellate review processes have greatly increased the waiting time on “death row.” To be sure, the ordeal of the condemned man may be thought cruel in the sense that all suffering is thought cruel. But if the Constitution proscribed every punishment producing severe emotional stress, then capital punishment would clearly have been impermissible in 1791.

However, the inquiry cannot end here. For reasons unrelated to any change in intrinsic cruelty, the Eighth Amendment prohibition cannot fairly be limited to those punishments thought excessively cruel and barbarous at the time of the adoption of the Eighth Amendment. A punishment is inordinately cruel, in the sense we must deal with it in these cases, chiefly as perceived by the society so characterizing it. The standard of extreme cruelty is not merely descriptive, but necessarily embodies a moral judgment. The standard itself remains the same, but its applicability must change as the basic mores of society change. This notion is not *383new to Eighth Amendment adjudication. In Weems v. United States, 217 U. S. 349 (1910), the Court referred with apparent approval to the opinion of the commentators that “[t]he clause of the Constitution . . . may be therefore progressive, and is not fastened to the obsolete but may acquire meaning as public opinion becomes enlightened by a humane justice.” 217 U. S., at 378. Mr. Chief Justice Warren, writing the plurality opinion in Trop v. Dulles, supra, stated, “The Amendment must fidraw its meaning from the evolving standards of decency that mark the progress of a maturing society.” 356 U. S., at 101. Nevertheless, the Court up to now has never actually held that a punishment has become impermis-sibly cruel due to a shift in the weight of accepted social values; nor has the Court suggested judicially manageable criteria for measuring such a shift in moral consensus.

The Court’s quiescence in this area can be attributed to the fact that in a democratic society legislatures, not courts, are constituted to respond to the will and consequently the moral values of the people. For this reason, early commentators suggested that the “cruel and unusual punishments” clause was an unnecessary constitutional provision.5 As acknowledged in the principal brief for petitioners, “both in constitutional contemplation and in fact, it is the legislature, not the Court, which responds to public opinion and immediately reflects the society’s standards of decency.”6 *384Accordingly, punishments such as branding and the cutting off of ears, which were commonplace at the time of the adoption of the Constitution, passed from the penal scene without judicial intervention because they became basically offensive to- the people and the legislatures responded to this sentiment.

Beyond any doubt, if we were today called upon to review such punishments, we would find them excessively cruel because we could say with complete assurance that contemporary society universally rejects such bizarre penalties. However, this speculation on the Court's probable reaction to such punishments is not of itself significant. The critical fact is that this Court has never had to hold that a mode of punishment authorized by a domestic legislature was so cruel as to be fundamentally at odds with our basic notions of decency. Cf. Weems v. United States, supra. Judicial findings of impermissible cruelty have been limited, for the most part, to offensive punishments devised without specific authority by prison officials, not by legislatures. See, e. g., Jackson v. Bishop, 404 F. 2d 571 (CA8 1968); Wright v. McMann, 387 F. 2d 519 (CA2 1967). The paucity of judicial decisions invalidating legislatively prescribed punishments is powerful evidence that in this country legislatures have in fact been responsive — albeit belatedly at times — to changes in social attitudes and moral values.

I do not suggest that the validity of legislatively authorized punishments presents no justiciable issue under the Eighth Amendment, but, rather, that the primacy of the legislative role narrowly confines the scope of judicial inquiry. Whether or not provable, and whether or not true at all times, in a democracy the legislative judgment is presumed to embody the basic standards of decency prevailing in the society. This presumption can only be negated by unambiguous and compelling evidence of legislative default.

*385III

There are no obvious indications that capital punishment offends the conscience of society to such a degree that our traditional deference to the legislative judgment must be abandoned. It is not a punishment such as burning at the stake that everyone would ineffably find to be repugnant to all civilized standards. Nor is it a punishment so roundly condemned that only a few aberrant legislatures have retained it on the statute books. Capital punishment is authorized by statute in 40 States, the District of Columbia, and in the federal courts for the commission of certain crimes.7 On four occasions in the last 11 years Congress has added to the list of federal crimes punishable by death.8 In looking for reliable indicia of contemporary attitude, none more trustworthy has been advanced.

One conceivable source of evidence that legislatures have abdicated their essentially barometric role with respect to community values would be public opinion polls, of which there have been many in the past decade addressed to the question of capital punishment. Without assessing the reliability of such polls, or intimating that any judicial reliance could ever be placed on them, *386it need only be noted that the reported results have shown nothing approximating the universal condemnation of capital punishment that might lead us to suspect that the legislatures in general have lost touch with current social values.9

Counsel for petitioners rely on a different body of empirical evidence. They argue, in effect, that the number of cases in which the death penalty is imposed, as compared with the number of cases in which it is statutorily available, reflects a general revulsion toward the penalty that would lead to its repeal if only it were more generally and widely enforced. It cannot be gainsaid that by the choice of juries — and sometimes judges10— the death penalty is imposed in far fewer than half the cases in which it is available.11 To go further and char*387acterize the rate of imposition as “freakishly rare,” as petitioners insist, is unwarranted hyperbole. And regardless of its characterization, the rate of imposition does not impel the conclusion that capital punishment is now regarded as intolerably cruel or uncivilized.

It is argued that in those capital cases where juries have recommended mercy, they have given expression to civilized values and effectively renounced the legislative authorization for capital punishment. At the same time it is argued that where juries have made the awesome decision to send men to their deaths, they have acted arbitrarily and without sensitivity to prevailing standards of decency. This explanation for the infrequency of imposition of capital punishment is unsupported by known facts, and is inconsistent in principle with everything this Court has ever said about the functioning of juries in capital eases.

In McGautha v. California, supra, decided only one year ago, the Court held that there was no mandate in the Due Process Clause of the Fourteenth Amendment that juries be given instructions as to when the death penalty should be imposed. After reviewing the autonomy that juries have traditionally exercised in capital cases and noting the practical difficulties of framing manageable instructions, this Court concluded that judicially articulated standards were not needed to insure a responsible decision as to penalty. Nothing in McGautha licenses capital juries to act arbitrarily or assumes that they have so acted in the past. On the contrary, the assumption underlying the McGautha ruling is that juries “will act with *388due regard for the consequences of their decision.” 402 U. S., at 208.

The responsibility of juries deciding capital cases in our system of justice was nowhere better described than in Witherspoon v. Illinois, supra:

“[A] jury that must choose between life imprisonment and capital punishment can do little more— and must do nothing less — than express the con science of the community on the ultimate question of life or death.”

“And one of the most important functions any jury can perform in making such a selection is to maintain a link between contemporary community values and the penal system — a link without which the determination of punishment could hardly reflect ‘the evolving standards of decency that mark the progress of a maturing society' ” 391 U. S., at 519 and n. 15 (emphasis added).

The selectivity of juries in imposing the punishment of death is properly viewed as a refinement on, rather than a repudiation of, the statutory authorization for that penalty. Legislatures prescribe the categories of crimes for which the death penalty should be available, and, acting as “the conscience of the community,” juries are entrusted to determine in individual cases that the ultimate punishment is warranted. Juries are undoubtedly influenced in this judgment by myriad factors. The motive or lack of motive of the perpetrator, the degree of injury or suffering of the victim or victims, and the degree of brutality in the commission of the crime would seem to be prominent among these factors. Given the general awareness that death is no longer a routine punishment for the crimes for which it is made available, it is hardly surprising that juries have been increasingly meticulous in their imposition of the penalty. But to *389assume from the mere fact of relative infrequency that only a random assortment of pariahs are sentenced to death, is to cast grave doubt on the basic integrity of our jury system.

It would, of course, be unrealistic to assume that juries have been perfectly consistent in choosing the cases where the death penalty is to be imposed, for no human institution performs with perfect consistency. There are doubtless prisoners on death row who would not be there had they been tried before a different jury or in a different State. In this sense their fate has been controlled by a fortuitous circumstance. However, this element of fortuity does not stand as an indictment either of the general functioning of juries in capital cases or of the integrity of jury decisions in individual cases. There is no empirical basis for concluding that juries have generally failed to discharge in good faith the responsibility described in Witherspoon — that of choosing between life and death in individual cases according to the dictates of community values.12

*390The rate of imposition of death sentences falls far short of providing the requisite unambiguous evidence that the legislatures of 40 States and the Congress have turned their backs on current or evolving standards of decency in continuing to make the death penalty available. For, if selective imposition evidences a rejection of capital punishment in those cases where it is not imposed, it surely evidences a correlative affirmation of the penalty in those cases where it is imposed. Absent some clear indication that the continued imposition of the death penalty on a selective basis is violative of prevailing standards of civilized conduct, the Eighth Amendment cannot be said to interdict its use.

*391In two of these cases we have been asked to rule on the narrower question whether capital punishment offends the Eighth Amendment when imposed as the punishment for the crime of forcible rape.13 It is true that the death penalty is authorized for rape in fewer States than it is for murder,14 and that even in those States it is applied more sparingly for rape than for murder.15 But for the reasons aptly brought out in the opinion of Mr. Justice Powell, post, at 456-461, I do not believe these differences can be elevated to the level of an Eighth Amendment distinction. This blunt constitutional command cannot be sharpened to carve neat distinctions corresponding to the categories of crimes defined by the legislatures.

IV

Capital punishment has also been attacked as viola-tive of the Eighth Amendment on the ground that it is not needed to achieve legitimate penal aims and is thus “unnecessarily cruel.” As a pure policy matter, this approach has much to recommend it, but it seeks to give a dimension to the Eighth Amendment that it was' never intended to have and promotes a line of inquiry that this Court has never before pursued.

The Eighth Amendment, as I have noted, was included in the Bill of Rights to guard against the use of torturous and inhuman punishments, not those of limited efficacy. One of the few to speak out against the adop*392tion of the Eighth Amendment asserted that it is often necessary to use cruel punishments to deter crimes.16 But among those favoring the Amendment, no sentiment was expressed that a punishment of extreme cruelty could ever be justified by expediency. The dominant theme of the Eighth Amendment debates was that the ends of the criminal laws cannot justify the use of measures of extreme cruelty to achieve them. Cf. Rochin v. California, 342 U. S. 165, 172-173 (1952).

The apparent seed of the “unnecessary cruelty” argument is the following language, quoted earlier, found in Wilkerson v. Utah, supra:

“Difficulty would attend the effort to define with exactness the extent of the constitutional provision which provides that cruel and unusual punishments shall not be inflicted; but it is safe to affirm that punishments of torture . . . and all others in the same line of unnecessary cruelty, are forbidden by that amendment to the Constitution.” 99 U. S., at 135-136 (emphasis added).

To lift the italicized phrase from the context of the Wilkerson opinion and now view it as a mandate for assessing the value of punishments in achieving the aims of penology is a gross distortion; nowhere are such aims even mentioned in the Wilkerson opinion. The only fair reading of this phrase is that punishments similar to torture in their extreme cruelty are prohibited by the Eighth Amendment. In Louisiana ex rel. Francis v. Resweber, 329 U. S., at 463, 464, the Court made reference to the Eighth Amendment’s prohibition against the infliction of “unnecessary pain” in carrying out an execution. The context makes abundantly clear that the Court was disapproving the wanton infliction of phys*393ical pain, and once again not advising pragmatic analysis of punishments approved by legislatures.17

Apart from these isolated uses of the word “unnecessary/' nothing in the cases suggests that it is for the courts to make a determination of the efficacy of punishments. The decision in Weems v. United States, supra, is not to the contrary. In Weems the Court held that for the crime of falsifying public documents, the punishment imposed under the Philippine Code of 15 years' imprisonment at hard labor under shackles, followed by perpetual surveillance, loss of voting rights, loss of the right to hold public office, and loss of right to change domicile freely, was violative of the Eighth Amendment. The case is generally regarded as holding that a punishment may be excessively cruel within the meaning of the Eighth Amendment because it is grossly out of proportion to the severity of the crime;18 some view the decision of the Court primarily as *394a reaction to the mode of the punishment itself.19 Under any characterization of the holding, it is readily apparent that the decision grew out of the Court’s overwhelming abhorrence of the imposition of the particular penalty for the particular crime; it was making an essentially moral judgment, not a dispassionate assessment of the need for the penalty. The Court specifically disclaimed “the right to assert a judgment against that of the legislature of the expediency of the laws . . . 217 U. S., at 378. Thus, apart from the fact that the Court in Weems concerned itself with the crime committed as well as the punishment imposed, the case marks no departure from the largely unarticulable standard of extreme cruelty. However intractable that standard may be, that is what the Eighth Amendment is all about. The constitutional provision is not addressed to social utility and does not command that enlightened principles of penology always be followed.

By pursuing the necessity approach, it becomes even more apparent that it involves matters outside the purview of the Eighth Amendment. Two of the several aims of punishment are generally associated with capital punishment — retribution and deterrence. It is argued that retribution can be discounted because that, after all, is what the Eighth Amendment seeks to eliminate. There is no authority suggesting that the Eighth Amendment was intended to purge the law of its retributive elements, and the Court has consistently assumed that retribution is a legitimate dimension of the punishment of crimes. See Williams v. New York, 337 U. S. 241, 248 (1949); United States v. Lovett, 328 U. S. 303, 324 (1946) (Frankfurter, J., concurring). Furthermore, responsible legal thinkers of widely varying *395persuasions have debated the sociological and philosophical aspects of the retribution question for generations, neither side being able to convince the other.20 It would be reading a great deal into the Eighth Amendment to hold that the punishments authorized by-legislatures cannot constitutionally reflect a retributive purpose.

The less esoteric but no less controversial question is whether the death penalty acts as a superior deterrent. Those favoring abolition find no evidence that it does.21 Those favoring retention start from the intuitive notion that capital punishment should act as the most effective deterrent and note that there is no convincing evidence that it does not.22 Escape from this empirical stalemate is sought by placing the burden of proof on the States and concluding that they have failed to demonstrate that capital punishment is a more effective deterrent than life imprisonment. Numerous justifications have been advanced for shifting the burden, and they *396are not without their rhetorical appeal. However, these arguments are not descended from established constitutional principles, but are born of the urge to bypass an unresolved factual question.23 Comparative deterrence is not a matter that lends itself to precise measurement; to shift the burden to the States is to provide an illusory solution to an enormously complex problem. If it were proper to put the States to the test of demonstrating the deterrent value of capital punishment, we could just as well ask them to prove the need for life imprisonment or any other punishment. Yet I know of no convincing evidence that life imprisonment is a more effective deterrent than 20 years' imprisonment, or even that a $10 parking ticket is a more effective deterrent than a $5 parking ticket. In fact, there are some who go so far as to challenge the notion that any punishments deter crime.24 If the States are unable to adduce convincing proof rebutting such assertions, does it then follow that all punishments are suspect as being “cruel and unusual” within the meaning of the Constitution? On the contrary, I submit that the questions raised by the necessity approach are beyond the pale of judicial inquiry under the Eighth Amendment.

y

Today the Court has not ruled that capital punishment is per se violative of the Eighth Amendment; nor has it ruled that the punishment is barred for any particular class or classes of crimes. The substantially similar concurring opinions of Mr. Justice Stewart and Mr. Justice White, which are necessary to support the judgment setting aside petitioners’ sentences, stop *397short of reaching the ultimate question. The actual scope of the Court’s ruling, which I take to be embodied in these concurring opinions, is not entirely clear. This much, however, seems apparent: if the legislatures are to continue to authorize capital punishment for some crimes, juries and judges can no longer be permitted to make the sentencing determination in the same manner they have in the past.25 This approach — not urged in oral arguments or briefs — misconceives the nature of the constitutional command against “cruel and unusual punishments,” disregards controlling case law, and demands a rigidity in capital cases which, if possible of achievement, cannot be regarded as a welcome change. Indeed the contrary seems to be the case.

As I have earlier stated, the Eighth Amendment forbids the imposition of punishments that are so cruel and inhumane as to violate society’s standards of civilized conduct. The Amendment does not prohibit all punishments the States are unable to prove necessary to deter or control crime. The Amendment is not concerned with the process by which a State determines that a particular punishment is to be imposed in a particular case. And the Amendment most assuredly does not speak to the power of legislatures to confer sentencing discretion on juries, rather than to fix all sentences by statute.

The critical factor in the concurring opinions of both Mr. Justice Stewart and Mr. Justice White is the infrequency with which the penalty is imposed. This factor is taken not as evidence of society’s abhorrence *398of capital punishment — the inference that petitioners would have the Court draw — but as the earmark of a deteriorated system of sentencing. It is concluded that petitioners’ sentences must be set aside, not because the punishment is impermissibly cruel, but because juries and judges have failed to exercise their sentencing discretion in acceptable fashion.

To be sure, there is a recitation cast in Eighth Amendment terms: petitioners’ sentences are “cruel” because they exceed that which the legislatures have deemed necessary for all cases;26 petitioners’ sentences are “unusual” because they exceed that which is imposed in most cases.27 This application of the words of the Eighth Amendment suggests that capital punishment can be made to satisfy Eighth Amendment values if its rate of imposition is somehow multiplied; it seemingly follows that the flexible sentencing system created by the legislatures, and carried out by juries and judges, has yielded more mercy than the Eighth Amendment can stand. The implications of this approach are mildly ironical. For example, by this measure of the Eighth Amendment, the elimination of death-qualified juries in Witherspoon v. Illinois, 391 U. S. 510 (1968), can only be seen in retrospect as a setback to “the evolving standards of decency that mark the progress of a maturing society.” Trop v. Dulles, 356 U. S., at 101.

This novel formulation of Eighth Amendment principles — albeit necessary to satisfy the terms of our limited grant of certiorari — does not lie at the heart of these concurring opinions. The decisive grievance of the opinions — not translated into Eighth Amendment terms— is that the present system of discretionary sentencing *399in capital cases has failed to produce evenhanded justice; the problem is not that too few have been sentenced to die, but that the selection process has followed no rational pattern.28 This claim of arbitrariness is not only lacking in empirical support29 but also it manifestly fails to establish that the death penalty is a “cruel and unusual” punishment. The Eighth Amendment was included in the Bill of Rights to assure that certain types of punishments would never be imposed, not to channelize the sentencing process. The approach of these concurring opinions has no antecedent in the Eighth Amendment cases. It is essentially and exclusively a procedural due process argument.

This ground of decision is plainly foreclosed as well as misplaced. Only one year ago, in McGautha v. California, the Court upheld the prevailing system of sentencing in capital cases. The Court concluded:

“In light of history, experience, and the present limitations of human knowledge, we find it quite impossible to say that committing to the untrammeled discretion of the jury the power to pronounce life or death in capital cases is offensive to anything in the Constitution.” 402 U. S., at 207.

In reaching this decision, the Court had the benefit of extensive briefing, full oral argument, and six months of careful deliberations. The Court’s labors are documented by 130 pages of opinions in the United States Reports. All of the arguments and factual contentions accepted *400in the concurring opinions today were considered and rejected by the Court one year ago. McGautha was an exceedingly difficult case, and reasonable men could fairly disagree as to the result. But the Court entered its judgment, and if stare decisis means anything, that decision should be regarded as a controlling pronouncement of law.

Although the Court’s decision in McGautha was technically confined to the dictates of the Due Process Clause of the Fourteenth Amendment, rather than the Eighth Amendment as made applicable to the States through the Due Process Clause of the Fourteenth Amendment, it would be disingenuous to suggest that today’s ruling has done anything less than overrule McGautha in the guise of an Eighth Amendment adjudication. It may be thought appropriate to subordinate principles of stare decisis where the subject is as sensitive as capital punishment and the stakes are so high, but these external considerations were no less weighty last year. This pattern of decisionmaking will do little to inspire confidence in the stability of the law.

While I would not undertake to make a definitive statement as to the parameters of the Court’s ruling, it is clear that if state legislatures and the Congress wish to maintain the availability of capital punishment, significant statutory changes will have to be made. Since the two pivotal concurring opinions turn on the assumption that the punishment of death is now meted out in a random and unpredictable manner, legislative bodies may seek to bring their laws into compliance with the Court’s ruling by providing standards for juries and judges to follow in determining the sentence in capital cases or by more narrowly defining the crimes for which the penalty is to be imposed.30 If such standards can be devised or *401the crimes more meticulously defined, the result cannot be detrimental. However, Mr. Justice Harlan’s opinion for the Court in McGautha convincingly demonstrates that all past efforts “to identify before the fact” the cases in which the penalty is to be imposed have been “uniformly unsuccessful.” 402 U. S., at 197. One problem is that “the factors which determine whether the sentence of death is the appropriate penalty in particular cases are too complex to be compressed within the limits of a simple formula . . . .” Report of Royal Commission on Capital Punishment, 1949-1953, Cmd. 8932, ¶ 498, p. 174 (1953). As the Court stated in McGautha, “[t]he infinite variety of cases and facets to each case would make general standards either meaningless 'boilerplate’ or a statement of the obvious that no jury would need.” 402 U. S., at 208. But even assuming that suitable guidelines can be established, there is no assurance that sentencing patterns will change so long as juries are possessed of the power to determine the sentence or to bring in a verdict of guilt on a charge carrying a lesser sentence; juries have not been inhibited in the exercise of these powers in the past. Thus, unless the Court in McGautha misjudged the experience of history, there is little reason to believe that sentencing standards in any form will substantially alter the discretionary character of the prevailing system of sentencing in capital cases. That system may fall short of perfection, but it is yet to be shown that a different system would produce more satisfactory results.

Real change could clearly be brought about if legislatures provided mandatory death sentences in such a way as to deny juries the opportunity to bring in a verdict on a lesser charge; under such a system, the death sentence could only be avoided by a verdict of acquittal. If this is the only alternative that the legislatures can safely pursue under today’s ruling, I would have preferred that the Court opt for total abolition.

*402It seems remarkable to me that with our basic trust in lay jurors as the keystone in our system of criminal justice, it should now be suggested that we take the most sensitive and important of all decisions away from them. I could more easily be persuaded that mandatory sentences of death, without the intervening and ameliorating impact of lay jurors, are so arbitrary and doctrinaire that they violate the Constitution. The very infrequency of death penalties imposed by jurors attests their cautious and discriminating reservation of that penalty for the most extreme cases. I had thought that nothing was clearer in history, as we noted in McGautha one year ago, than the American abhorrence of “the common-law rule imposing a mandatory death sentence on all convicted murderers.” 402 U. S., at 198. As the concurring opinion of Me. Justice Marshall shows, ante, at 339, the 19th century movement away from mandatory death sentences marked an enlightened introduction of flexibility into the sentencing process. It recognized that individual culpability is not always measured by the category of the crime committed. This change in sentencing practice was greeted by the Court as a humanizing development. See Winston v. United States, 172 U. S. 303 (1899); cf. Calton v. Utah, 130 U. S. 83 (1889). See also Andres v. United States, 333 U. S. 740, 753 (1948) (Frankfurter, J., concurring). I do not see how this history can be ignored and how it can be suggested that the Eighth Amendment demands the elimination of the most sensitive feature of the sentencing system.

As a general matter, the evolution of penal concepts in this country has not been marked by great progress, nor have the results up to now been crowned with significant success. If anywhere in the whole spectrum of criminal justice fresh ideas deserve sober analysis, the sentencing and porrectional area ranks high on the list. But it has been widely accepted that mandatory sentences for *403crimes do not best serve the ends of the criminal justice system. Now, after the long process of drawing away from the blind imposition of uniform sentences for every person convicted of a particular offense, we are confronted with an argument perhaps implying that only the legislatures may determine that a sentence of death is appropriate, without the intervening evaluation of jurors or judges. This approach threatens to turn back the progress of penal reform, which has moved until recently at too slow a rate to absorb significant setbacks.

VI

Since there is no majority of the Court on the ultimate issue presented in these cases, the future of capital punishment in this country has been left in an uncertain limbo. Rather than providing a final and unambiguous answer on the basic constitutional question, the collective impact of the majority’s ruling is to demand an undetermined measure of change from the various state legislatures and the Congress. While I cannot endorse the process of decisionmaking that has yielded today’s result and the restraints that that result imposes on legislative action, I am not altogether displeased that legislative bodies have been given the opportunity, and indeed unavoidable responsibility, to make a thorough re-evaluation of the entire subject of capital punishment. If today’s opinions demonstrate nothing else, they starkly show that this is an area where legislatures can act far more effectively than courts.

The legislatures are free to eliminate capital punishment for specific crimes or to carve out limited exceptions to a general abolition of the penalty, without adherence to the conceptual strictures of the Eighth Amendment. The legislatures can and should make an assessment of the deterrent influence of capital punishment, both generally and as affecting the commission of specific types of *404crimes. If legislatures come to doubt the efficacy of capital punishment, they can abolish it, either completely or on a selective basis. If new evidence persuades them that they have acted unwisely, they can reverse their field and reinstate the penalty to the extent it is thought warranted. An Eighth Amendment ruling by judges cannot be made with such flexibility or discriminating precision.

The world-wide trend toward limiting the use of capital punishment, a phenomenon to which we have been urged to give great weight, hardly points the way to a judicial solution in this country under a written Constitution. Rather, the change has generally come about through legislative action, often on a trial basis and with the retention of the penalty for certain limited classes of crimes.31 Virtually nowhere has change been wrought by so crude a tool as the Eighth Amendment. The complete and unconditional abolition of capital punishment in this country by judicial fiat would have undermined the careful progress of the legislative trend and foreclosed further inquiry on many as yet unanswered questions in this area.

Quite apart from the limitations of the Eighth Amendment itself, the preference for legislative action is justified by the inability of the courts to participate in the *405debate at the level where the controversy is focused. The case against capital punishment is not the product of legal dialectic, but rests primarily on factual claims, the truth of which cannot be tested by conventional judicial processes. The five opinions in support of the judgments differ in many respects, but they share a willingness to make sweeping factual assertions, unsupported by empirical data, concerning the manner of imposition and effectiveness of capital punishment in this country. Legislatures will have the opportunity to make a more penetrating study of these claims with the familiar and effective tools available to them as they are not to us.

The highest judicial duty is to recognize the limits on judicial power and to permit the democratic processes to deal with matters falling outside of those limits. The “hydraulic pressure [s]” 32 that Holmes spoke of as being generated by cases of great import have propelled the Court to go beyond the limits of judicial power, while fortunately leaving some room for legislative judgment.

Me. Justice Blackmun,

dissenting.

I join the respective opinions of The Chief Justice, Me. Justice Powell, and Me. Justice Rehnquist, and add only the following, somewhat personal, comments.

1. Cases such as these provide for me an excruciating agony of the spirit. I yield to no one in the depth of my distaste, antipathy, and, indeed, abhorrence, for the death penalty, with all its aspects of physical distress and fear and of moral judgment exercised by finite minds. That distaste is buttressed by a belief that capital punishment serves no useful purpose that can be demonstrated. For me, it violates childhood’s training and life’s experiences, and is not compatible *406with the philosophical convictions I have been able to develop. It is antagonistic to any sense of “reverence for life.” Were I a legislator, I would vote against the death penalty for the policy reasons argued by counsel for the respective petitioners and expressed and adopted in the several opinions filed by the Justices who vote to reverse these judgments.

2. Having lived for many years in a State that does not have the death penalty,1 that effectively abolished it in 1911,2 and that carried out its last execution on February 13, 1906,3 capital punishment had never been a part of life for me. In my State, it just did not exist. So far as I can determine, the State, purely from a statistical deterrence point of view, was neither the worse nor the better for its abolition, for, as the concurring opinions observe, the statistics prove little, if anything. But the State and its citizens accepted the fact that the death penalty was not to be in the arsenal of possible punishments for any crime.

3. I, perhaps alone among the present members of the Court, am on judicial record as to this. As a member of the United States Court of Appeals, I first struggled silently with the issue of capital punishment in Feguer v. United States, 302 F. 2d 214 (CA8 1962), cert. denied, 371 U. S. 872 (1962). The defendant in that case may have been one of the last to be executed under federal auspices. I struggled again with the issue, and once more refrained from comment, in my writing for an en banc court in Pope v. United States, 372 F. 2d 710 (CA8 1967), vacated (upon acknowledgment by the Solicitor General of error revealed by the subsequently decided United States v. Jackson, 390 U. S. 570 (1968)) and remanded, 392 U. S. 651 (1968). Finally, in Max*407well v. Bishop, 398 F. 2d 138 (CA8 1968), vacated and remanded, sua sponte, by the Court on grounds not raised below, 398 U. S. 262 (1970), I revealed, solitarily and not for the panel, my distress and concern. 398 F. 2d, at 153-154.4 And in Jackson v. Bishop, 404 F. 2d 571 (CA8 1968), I had no hesitancy in writing a panel opinion that held the use of the strap by trusties upon fellow Arkansas prisoners to be a violation of the Eighth Amendment. That, however, was in-prison punishment imposed by inmate-foremen.

4. The several concurring opinions acknowledge, as they must, that until today capital punishment was accepted and assumed as not unconstitutional per se under the Eighth Amendment or the Fourteenth Amendment. This is either the flat or the implicit holding of a unanimous Court in Wilkerson v. Utah, 99 U. S. 130, 134-135, in 1879; of a unanimous Court in In re Kemmler, 136 U. S. 436, 447, in 1890; of the Court in Weems v. United States, 217 U. S. 349, in 1910; of all those members of the Court, a majority, who addressed the issue in Louisiana ex rel. Francis v. Resweber, 329 U. S. 459, 463-464, 471-472, in 1947; of Mr. Chief Justice Warren, speaking for himself and three others (Justices Black, Doug*408las, and Whittaker) in Trop v. Dulles, 356 U. S. 86, 99, in 1958;5 in the denial of certiorari in Rudolph v. Alabama, 375 U. S. 889, in 1963 (where, however, Justices Douglas, Brennan, and Goldberg would have heard argument with respect to the imposition of the ultimate penalty on a convicted rapist who had “neither taken nor endangered human life”); and of Mr. Justice Black in McGautha v. California, 402 U. S. 183, 226, decided only last Term on May 3, 1971.6

Suddenly, however, the course of decision is now the opposite way, with the Court evidently persuaded that somehow the passage of time has taken us to a place of greater maturity and outlook. The argument, plausible and high-sounding as it may be, is not persuasive, for it is only one year since McGautha, only eight and one-half years since Rudolph, 14 years since Trop, and 25 years since Francis, and we have been presented with nothing that demonstrates a significant movement of any kind in these brief periods. The Court has just decided that it is time to strike down the death penalty. There would have been as much reason to do this *409when any of the cited cases were decided. But the Court refrained from that action on each of those occasions.

The Court has recognized, and I certainly subscribe to the proposition, that the Cruel and Unusual Punishments Clause “may acquire meaning as public opinion becomes enlightened by a humane justice.” Weems v. United States, 217 U. S., at 378. And Mr. Chief Justice Warren, for a plurality of the Court, referred to “the evolving standards of decency that mark the progress of a maturing society.” Trop v. Dulles, 356 U. S., at 101. Mr. Jefferson expressed the same thought well.7

*410My problem, however, as I have indicated, is the suddenness of the Court’s perception of progress in the human attitude since decisions of only a short while ago.

5. To reverse the judgments in these cases is, of course, the easy choice. It is easier to strike the balance in favor of life and against death. It is comforting to relax in the thoughts — perhaps the rationalizations— that this is the compassionate decision for a maturing society; that this is the moral and the “right” thing to do; that thereby we convince ourselves that we are moving down the road toward human decency; that we value life even though that life has taken another or others or has grievously scarred another or others and their families; and that we are less barbaric than we were in 1879, or in 1890, or in 1910, or in 1947, or in 1958, or in 1963, or a year ago, in 1971, when Wilkerson, Kemmler, Weems, Francis, Trop, Rudolph, and Mc-Gautha were respectively decided.

This, for me, is good argument, and it makes some sense. But it is good argument and it makes sense only in a legislative and executive way and not as a judicial expedient. As I have said above, were I a legislator, I would do all I could to sponsor and to vote for legislation abolishing the death penalty. And were I the chief executive of a sovereign State, I would be sorely tempted to exercise executive clemency as Governor Rockefeller of Arkansas did recently just before he departed from office. There — on the Legislative Branch of the State or Federal Government, and secondarily, on the Executive Branch — is where the authority and responsibility for this kind of action lies. The authority should not be taken over by the judiciary in the modern guise of an Eighth Amendment issue.

I do not sit on these cases, however, as a legislator, responsive, at least in part, to the will of constituents. *411Our task here, as must so frequently be emphasized and re-emphasized, is to pass upon the constitutionality of legislation that has been enacted and that is challenged. This is the sole task for judges. We should not allow our personal preferences as to the wisdom of legislative and congressional action, or our distaste for such action, to guide our judicial decision in cases such as these. The temptations to cross that policy line are very great. In fact, as today’s decision reveals, they are almost irresistible.

6. The Court, in my view, is somewhat propelled toward its result by the interim decision of the California Supreme Court, with one justice dissenting, that the death penalty is violative of that State’s constitution. People v. Anderson, 6 Cal. 3d 628, 493 P. 2d 880 (Feb. 18, 1972). So far as I am aware, that was the first time the death penalty in its entirety has been nullified by judicial decision. Cf. Ralph v. Warden, 438 F. 2d 786, 793 (CA4 1970), cert. denied, post, p. 942. California’s moral problem was a profound one, for more prisoners were on death row there than in any other State. California, of course, has the right to construe its constitution as it will. Its construction, however, is hardly a precedent for federal adjudication.

7. I trust the Court fully appreciates what it is doing when it decides these cases the way it does today. Not only are the capital punishment laws of 39 States and the District of Columbia struck down, but also all those provisions of the federal statutory structure that permit the death penalty apparently are voided. No longer is capital punishment possible, I suspect, for, among other crimes, treason, 18 U. S. C. § 2381; or assassination of the President, the Vice President, or those who stand elected to those positions, 18 U. S. C. § 1751; or assassination of a Member or member-elect of Congress, 18 U. S. C. § 351; or espionage, 18 U. S. C. § 794; *412or rape within the special maritime jurisdiction, 18 U. S. C. § 2031; or aircraft or motor vehicle destruction where death occurs, 18 U. S. C. § 34; or explosives offenses where death results, 18 U. S. C. §§ 844 (d) and (f); or train wrecking, 18 U. S. C. § 1992; or aircraft piracy, 49 U. S. C. § 1472 (i). Also in jeopardy, perhaps, are the death penalty provisions in various Articles of the Uniform Code of Military Justice. 10 U. S. C. §§ 885, 890, 894, 899, 901, 904, 906, 913, 918, and 920. All these seem now to be discarded without a passing reference to the reasons, or the circumstances, that prompted their enactment, some very recent, and their retention in the face of efforts to repeal them.

8. It is of passing interest to note a few voting facts with respect to recent federal death penalty legislation:

A. The aircraft piracy statute, 49 U. S. C. § 1472 (i), was enacted September 5, 1961. The Senate vote on August 10 was 92-0. It was announced that Senators Chavez, Fulbright, Neuberger, and Symington were absent but that, if present, all four would vote yea. It was also announced, on the other side of the aisle, that Senator Butler was ill and that Senators Beall, Carlson, and Morton were absent or detained, but that those four, if present, would vote in the affirmative. These announcements, therefore, indicate that the true vote was 100-0. 107 Cong. Rec. 15440. The House passed the bill without recorded vote. 107 Cong. Rec. 16849.

B. The presidential assassination statute, 18 U. S. C. § 1751, was approved August 28, 1965, without recorded votes. Ill Cong. Rec. 14103, 18026, and 20239.

C. The Omnibus Crime Control Act of 1970 was approved January 2, 1971. Title IV thereof added the congressional assassination statute that is now 18 U. S. C. § 351. The recorded House vote on October 7, 1970, was 341-26, with 63 not voting and 62 of those paired. 116 Cong. Rec. 35363-35364. The Senate vote on October 8 *413was 59-0, with 41 not voting, but with 21 of these announced as favoring the bill. 116 Cong. Rec. 35743. Final votes after conference were not recorded. 116 Cong. Rec. 42150, 42199.

It is impossible for me to believe that the many lawyer-members of the House and Senate — including, I might add, outstanding leaders and prominent candidates for higher office — were callously unaware and insensitive of constitutional overtones in legislation of this type. The answer, of course, is that in 1961, in 1965, and in 1970 these elected representatives of the' people — far more conscious of the temper of the times, of the maturing of society, and of the contemporary demands for man’s dignity, than are we who sit cloistered on this Court— took it as settled that the death penalty then, as it always had been, was not in itself unconstitutional. Some of those Members of Congress, I suspect, will be surprised at this Court’s giant stride today.

9. If the reservations expressed by my Brother Stewart (which, as I read his opinion, my Brother White shares) were to command support, namely, that capital punishment may not be unconstitutional so long as it be man-datorily imposed, the result, I fear, will be that statutes struck down today will be re-enacted by state legislatures to prescribe the death penalty for specified crimes without any alternative for the imposition of a lesser punishment in the discretion of the judge or jury, as the case may be. This approach, it seems to me, encourages legislation that is regressive and of an antique mold, for it eliminates the element of mercy in the imposition of punishment. I thought we had passed beyond that point in our criminology long ago.

10. It is not without interest, also, to note that, although the several concurring opinions acknowledge the heinous and atrocious character of the offenses committed by the petitioners, none of those opinions makes *414reference to the misery the petitioners’ crimes occasioned to the victims, to the families of the victims, and to the communities where the offenses took place. The arguments for the respective petitioners, particularly the oral arguments, were similarly and curiously devoid of reference to the victims. There is risk, of course, in a comment such as this, for it opens one to the charge of emphasizing the retributive. But see Williams v. New York, 337 U. S. 241, 248 (1949). Nevertheless, these cases are here because offenses to innocent victims were perpetrated. This fact, and the terror that occasioned it, and the fear that stalks the streets of many of our cities today perhaps deserve not to be entirely overlooked. Let us hope that, with the Court’s decision, the terror imposed will be forgotten by those up'on whom it was visited, and that our society will reap the hoped-for benefits of magnanimity.

Although personally I may rejgice at the Court’s result, I find it difficult to accept or to justify as a matter of history, of law, or of constitutional pronouncement. I fear the Court has overstepped. It has sought and has achieved an end.

Mr. Justice Powell,

with whom The Chief Justice, Mr. Justice Blackmun, and Mr. Justice Rehnquist join, dissenting.

The Court granted certiorari in these cases to consider whether the death penalty is any longer a permissible form of punishment. 403 U. S. 952 (1971). It is the judgment of five Justices that the death penalty, as customarily prescribed and implemented in this country today, offends the constitutional prohibition against cruel and unusual punishments. The reasons for that judgment are stated in five separate opinions, expressing as many separate rationales. In my view, none of these opinions provides a constitutionally adequate foundation for the Court’s decision.

*415Mr. Justice Douglas concludes that capital punishment is incompatible with notions of "equal protection” that he finds to be “implicit” in the Eighth Amendment. Ante, at 257. Mr. Justice Brennan bases his judgment primarily on the thesis that the penalty “does not comport with human dignity.” Ante, at 270. Mr. Justice Stewart concludes that the penalty is applied in a “wanton” and “freakish” manner. Ante, at 310. For Mr. Justice White it is the “infrequency” with which the penalty is imposed that renders its use unconstitutional. Ante, at 313. Mr. Justice Marshall finds that capital punishment is an impermissible form of punishment because it is “morally unacceptable” and “excessive.” Ante, at 360, 358.

Although the central theme of petitioners’ presentations in these cases is that the imposition of the death penalty is per se unconstitutional, only two of today’s opinions explicitly conclude that so sweeping a determination is mandated by the Constitution. Both Mr. Justice Brennan and Mr. Justice Marshall call for the abolition of all existing state and federal capital punishment statutes. They intimate as well that no capital statute could be devised in the future that might comport with the Eighth Amendment. While the practical consequences of the other three opinions are less certain, they at least do not purport to render impermissible every possible statutory scheme for the use of capital punishment that legislatures might hereafter devise.1 Insofar as these latter opinions fail, at least ex*416plicitly, to go as far as petitioners’ contentions would carry them, their reservations are attributable to a willingness to accept only a portion of petitioners’ thesis. For the reasons cogently set out in the Chief Justice’s dissenting opinion (ante, at 396-403), and for reasons stated elsewhere in this opinion, I find my Brothers’ less-than-absolute-abolition judgments unpersuasive. Because those judgments are, for me, not dispositive, I shall focus primarily on the broader ground upon which the petitions in these cases are premised. The foundations of my disagreement with that broader thesis are equally applicable to each of the concurring opinions. I will, therefore, not endeavor to treat each one separately. Nor will I attempt to predict what forms of capital statutes, if any, may avoid condemnation in the future under the variety of views expressed by the collective majority today. That difficult task, not performed in any of the controlling opinions, must go unanswered until other cases presenting these more limited inquiries arise.

Whatever uncertainties may hereafter surface, several of the consequences of today’s decision are unmistakably clear. The decision is plainly one of the greatest im*417portance. The Court's judgment removes the death sentences previously imposed on some 600 persons awaiting punishment in state and federal prisons throughout the country. At least for the present, it also bars the States and the Federal Government from seeking sentences of death for defendants awaiting trial on charges for which capital punishment was heretofore a potential alternative. The happy event for these countable few constitutes, however, only the most visible consequence of this decision. Less measurable, but certainly of no less significance, is the shattering effect this collection of views has on the root principles of stare decisis, federalism, judicial restraint and — most importantly — separation of powers.

The Court rejects as not decisive the clearest evidence that the Framers of the Constitution and the authors of the Fourteenth Amendment believed that those documents posed no barrier to the death penalty. The Court also brushes aside an unbroken line of precedent reaffirming the heretofore virtually unquestioned constitutionality of capital punishment. Because of the pervasiveness of the constitutional ruling sought by petitioners, and accepted in varying degrees by five members of the Court, today’s departure from established precedent invalidates a staggering number of state and federal laws. The capital punishment laws of no less than 39 States2 and the District of Columbia are nullified. In addition, numerous provisions of the Criminal Code of the United States and of the Uniform Code of Mili*418tary Justice also are voided. The Court’s judgment not only wipes out laws presently in existence, but denies to Congress and to the legislatures of the 50 States the power to adopt new policies contrary to the policy selected by the Court. Indeed, it is the view of two of my Brothers that the people of each State must be denied the prerogative to amend their constitutions to provide for capital punishment even selectively for the most heinous crime.

In terms of the constitutional role of this Court, the impact of the majority’s ruling is all the greater because the decision encroaches upon an area squarely within the historic prerogative of the legislative branch — both state and federal — to protect the citizenry through the designation of penalties for prohibitable conduct. It is the very sort of judgment that the legislative branch is competent to make and for which the judiciary is ill-equipped. Throughout our history, Justices of this Court have emphasized the gravity of decisions invalidating legislative judgments, admonishing the nine men who sit on this bench of the duty of self-restraint, especially when called upon to apply the expansive due process and cruel and unusual punishment rubrics. I can recall no case in which, in the name of deciding constitutional questions, this Court has subordinated national and local democratic processes to such an extent. Before turning to address the thesis of petitioners’ case against capital punishment — a thesis that has proved, at least in large measure, persuasive to a majority of this Court— I first will set out the principles that counsel against the Court’s sweeping decision.

I

The Constitution itself poses the first obstacle to petitioners’ argument that capital punishment is per se unconstitutional. The relevant provisions are the Fifth, *419Eighth, and Fourteenth Amendments. The first of these provides in part:

“No person shall be held to answer for a capital, or otherwise infamous crime, unless on a presentment or indictment of a Grand Jury . . . ; nor shall any person be subject for the same offence to be twice put in jeopardy of life or limb; . . . nor be deprived of life, liberty, or property, without due process of law . . . .”

Thus, the Federal Government’s power was restricted in order to guarantee those charged with crimes that the prosecution would have only a single opportunity to seek imposition of the death penalty and that the death penalty could not be exacted without due process and a grand jury indictment. The Fourteenth Amendment, adopted about 77 years after the Bill of Rights, imposed the due process limitation of the Fifth Amendment upon the States’ power to authorize capital punishment.

The Eighth Amendment, adopted at the same time as the Fifth, proscribes “cruel and unusual” punishments. In an effort to discern its meaning, much has been written about its history in the opinions of this Court and elsewhere.3 That history need not be restated here since, whatever punishments the Framers of the Constitution may have intended to prohibit under the “cruel and unusual” language, there cannot be the slightest doubt that they intended no absolute bar on the Government’s authority to impose the death penalty. McGautha v. *420California, 402 U. S. 183, 226 (1971) (separate opinion of Black, J.). As much is made clear by the three references to capital punishment in the Fifth Amendment. Indeed, the same body that proposed the Eighth Amendment also provided, in the first Crimes Act of 1790, for the death penalty for a number of offenses. 1 Stat. 112.

Of course, the specific prohibitions within the Bill of Rights are limitations on the exercise of power; they are not an affirmative grant of power to the Government. I, therefore, do not read the several references to capital punishment as foreclosing this Court from considering whether the death penalty in a particular case offends the Eighth and Fourteenth Amendments. Nor are “cruel and unusual punishments” and “due process of law” static concepts whose meaning and scope were sealed at the time of their writing. They were designed to be dynamic and to gain meaning through application to specific circumstances, many of which were not contemplated by their authors. While flexibility in the application of these broad concepts is one of the hallmarks of our system of government, the Court is not free to read into the Constitution a meaning that is plainly at variance with its language. Both the language of the Fifth and Fourteenth Amendments and the history of the Eighth Amendment confirm beyond doubt that the death penalty was considered to be a constitutionally permissible punishment. It is, however, within the historic process of constitutional adjudication to challenge the imposition of the death penalty in some barbaric manner or as a penalty wholly disproportionate to a particular criminal act. And in making such a judgment in a case before it, a court may consider contemporary standards to the extent they are relevant. While this weighing of a punishment against the Eighth Amendment standard on a case-by-case basis is consonant with history and precedent, it is not what *421petitioners demand in these cases. They seek nothing less than the total abolition of capital punishment by judicial fiat.

II

Petitioners assert that the constitutional issue is an open one uncontrolled by prior decisions of this Court. They view the several cases decided under the Eighth Amendment as assuming the constitutionality of the death penalty without focusing squarely upon the issue. I do not believe that the case law can be so easily cast aside. The Court on numerous occasions has both assumed and asserted the constitutionality of capital punishment. In several cases that assumption provided a necessary foundation for the decision, as the issue was whether a particular means of carrying out a capital sentence would be allowed to stand. Each of those decisions necessarily was premised on the assumption that some method of exacting the penalty was permissible.

The issue in the first capital case in which the Eighth Amendment was invoked, Wilkerson v. Utah, 99 U. S. 130 (1879), was whether carrying out a death sentence by public shooting was cruel and unusual punishment. A unanimous Court upheld that form of execution, noting first that the punishment itself, as distinguished from the mode of its infliction, was “not pretended by the counsel of the prisoner” (id., at 137) to be cruel and unusual. The Court went on to hold that:

“Cruel and unusual punishments are forbidden by the Constitution, but the authorities . . . are quite sufficient to show that the punishment of shooting as a mode of executing the death penalty for the crime of murder in the first degree is not included in that category . . . .” Id., at 134-135.

Eleven years later, in In re Kemmler, 136 U. S. 436 (1890), the Court again faced a question involving the *422method of carrying out a capital sentence. On review of a denial of habeas corpus relief by the Supreme Court of New York, this Court was called on to decide whether electrocution, which only very recently had been adopted by the New York Legislature as a means of execution, was impermissibly cruel and unusual in violation of the Fourteenth Amendment.4 Chief Justice Fuller, speaking for the entire Court, ruled in favor of the State. Electrocution had been selected by the legislature, after careful investigation, as “the most humane and practical method known to modern science of carrying into effect the sentence of death.” Id., at 444. The Court drew a clear line between the penalty itself and the mode of its execution:

“Punishments are cruel when they involve torture or a lingering death; but the punishment of death *423is not cruel, within the meaning of that word as used in the Constitution. It implies there something inhuman and barbarous, something more than the mere extinguishment of life.” Id., at 447.

More than 50 years later, in Louisiana ex rel. Francis v. Resweber, 329 U. S. 459 (1947), the Court considered a case in which, due to a mechanical malfunction, Louisiana’s initial attempt to electrocute a convicted murderer had failed. Petitioner sought to block a second attempt to execute the sentence on the ground that to do so would constitute cruel and unusual punishment. In the plurality opinion written by Mr. Justice Reed, concurred in by Chief Justice Vinson and Justices Black and Jackson, relief was denied. Again the Court focused on the manner of execution, never questioning the propriety of the death sentence itself.

“The case before us does not call for an examination into any punishments except that of death. . . . The traditional humanity of modern Anglo-American law forbids the infliction of unnecessary pain in the execution of the death sentence. . . .

“. . . The cruelty against which the Constitution protects a convicted man is cruelty inherent in the method of punishment, not the necessary suffering involved in any method employed to extinguish life humanely.” Id., at 463-464.

Mr. Justice Frankfurter, unwilling to dispose of the case under the Eighth Amendment’s specific prohibition, approved the second execution attempt under the Due Process Clause. He concluded that “a State may be found to deny a person due process by treating even one guilty of crime in a manner that violates standards of *424decency more or less universally accepted though not when it treats him by a mode about which opinion is fairly divided.” Id., at 469-470.

The four dissenting Justices, although finding a second attempt at execution to be impermissibly cruel, expressly recognized the validity of capital punishment:

“In determining whether the proposed procedure is unconstitutional, we must measure it against a lawful electrocution. . . . Electrocution, when instantaneous, can be inflicted by a state in conformity with due process of law. . . .

“The all-important consideration is that the execution shall be so instantaneous and substantially painless that the punishment shall be reduced, as nearly as possible, to no more than that of death itself.” Id., at 474 (original emphasis).

Each of these cases involved the affirmance of a death sentence where its validity was attacked as violating the Eighth Amendment. Five opinions were written in these three cases, expressing the views of 23 Justices. While in the narrowest sense it is correct to say that in none was there a frontal attack upon the constitutionality of the death penalty, each opinion went well beyond an un-articulated assumption of validity. The power of the States to impose capital punishment was repeatedly and expressly recognized.

In addition to these cases in which the constitutionality of the death penalty was a necessary foundation for the decision, those who today would have this Court undertake the absolute abolition of the death penalty also must reject the opinions of other cases stipulating or assuming the constitutionality of capital punishment. Trop v. Dulles, 356 U. S. 86, 99, 100 (1958) ; Weems v. United States, 217 U. S. 349, 382, 409 (1910) *425(White, J., joined by Holmes, J., dissenting).5 See also McGautha v. California, 402 U. S., at 226 (separate opinion of Black, J.); Robinson v. California, 370 U. S. 660, 676 (1962) (Douglas, J., concurring).

The plurality opinion in Trop v. Dulles, supra, is of special interest since it is this opinion, in large measure, that provides the foundation for the present attack on the death penalty.6 It is anomalous that the standard urged by petitioners — “evolving standards of decency that mark the progress of a maturing society” (356 U. S., at 101) — should be derived from an opinion that so un-qualifiedly rejects their arguments. Chief Justice Warren, joined by Justices Black, Douglas, and Whittaker, stated flatly:

“At the outset, let us put to one side the death penalty as an index of the constitutional limit on punishment. Whatever the arguments may be against capital punishment, both on moral grounds and in terms of accomplishing the purposes of punishment — and they are forceful — the death penalty has been employed throughout our history, and, in a day when it is still widely accepted, it cannot be said to violate the constitutional concept of cruelty.” Id., at 99.

The issue in Trop was whether forfeiture of citizenship was a cruel and unusual punishment when imposed on *426a wartime deserter who had gone “over the hill” for less than a day and had willingly surrendered. In examining the consequences of the relatively novel punishment of denationalization,7 Chief Justice Warren drew a line between “traditional” and “unusual” penalties:

“While the State has the power to punish, the [Eighth] Amendment stands to assure that this power be exercised within the limits of civilized standards. Fines, imprisonment and even execution may be imposed depending upon the enormity of the crime, but any technique outside the bounds of these traditional penalties is constitutionally suspect.” Id., at 100.

The plurality’s repeated disclaimers of any attack on capital punishment itself must be viewed as more than offhand dicta since those views were written in direct response to the strong language in Mr. Justice Frankfurter’s dissent arguing that denationalization could not be a disproportionate penalty for a concededly capital offense.8

The most recent precedents of this Court — Witherspoon v. Illinois, 391 U. S. 510 (1968), and McGautha v. California, supra — are also premised to a significant degree on the constitutionality of the death penalty. While the scope of review in both cases was limited to questions involving the procedures for selecting juries *427and regulating their deliberations in capital cases,9 those opinions were “singularly academic exercise [s]” 10 if the members of this Court were prepared at those times to find in the Constitution the complete prohibition of the death penalty. This is especially true of Mr. Justice Harlan’s opinion for the Court in Mc-Gautha, in which, after a full review of the history of capital punishment, he concluded that “we find it quite impossible to say that committing to the untrammeled discretion of the jury the power to pronounce life or death in capital cases is offensive to anything in the Constitution.” Id., at 207.11

*428Perhaps enough has been said to demonstrate the unswerving position that this Court has taken in opinions spanning the last hundred years. On virtually every occasion that any opinion has touched on the question of' the constitutionality of the death penalty, it has been asserted affirmatively, or tacitly assumed, that the Constitution does not prohibit the penalty. No Justice of the Court, until today, has dissented from this consistent reading of the Constitution. The petitioners in these cases now before the Court cannot fairly avoid the weight of this substantial body of precedent merely by asserting that there is no prior decision precisely in point. Stare decisis, if it is a doctrine founded on principle, surely applies where there exists a long line of cases endorsing or necessarily assuming the validity of a particular matter of constitutional interpretation. Green v. United States, 356 U. S. 165, 189-193 (1958) (Frankfurter, J., concurring). While these oft-repeated expressions of unchallenged belief in the constitutionality of capital punishment may not justify a summary disposition of the constitutional question before us, they are views expressed and joined in over the years by no less than 29 Justices of this Court and therefore merit the greatest respect.12 Those who now resolve to set those views aside indeed have a heavy burden.

Ill

Petitioners seek to avoid the authority of the foregoing cases, and the weight of express recognition in the Constitution itself, by reasoning which will not withstand analysis. The thesis of petitioners’ case derives from several opinions in which members of this Court *429have recognized the dynamic nature of the prohibition against cruel and unusual punishments. The final meaning of those words was not set in 1791. Rather, to use the words of Chief Justice Warren speaking for a plurality of the Court in Trop v. Dulles, 356 U. S., at 100-101:

“[T]he words of the Amendment are not precise, and . . . their scope is not static. The Amendment must draw its meaning from the evolving standards of decency that mark the progress of a maturing society.”

But this was not new doctrine. It was the approach to the Eighth Amendment taken by Mr. Justice McKenna in his opinion for the Court in Weems v. United States, 217 U. S. 349 (1910). Writing for four Justices sitting as the majority of the six-man Court deciding the case, he concluded that the clause must be “progressive”; it is not “fastened to the obsolete but may acquire meaning as public opinion becomes enlightened by a humane justice.” Id., at 378. The same test was offered by Mr. Justice Frankfurter in his separate concurrence in Louisiana ex rel. Francis v. Resweber, 329 U. S., at 469. While he rejected the notion that the Fourteenth Amendment made the Eighth Amendment fully applicable to the States, he nonetheless found as a matter of due process that the States were prohibited from “treating even one guilty of crime in a manner that violates standards of decency more or less universally accepted.”

Whether one views the question as one of due process or of cruel and unusual punishment, as I do for convenience in this case, the issue is essentially the same.13 The fundamental premise upon which either standard is based is that notions of what constitutes cruel and unusual punishment or due process do evolve. *430Neither the Congress nor any state legislature would today tolerate pillorying, branding, or cropping or nailing of the ears — punishments that were in existence during our colonial era.14 Should, however, any such punishment be prescribed, the courts would certainly enjoin its execution. See Jackson v. Bishop, 404 F. 2d 571 (CA8 1968). Likewise, no court would approve any method of implementation of the death sentence found to involve unnecessary cruelty in light of presently available alternatives. Similarly, there may well be a process of evolving attitude with respect to the application of the death sentence for particular crimes.15 See McGautha v. California, 402 U. S., at 242 (Douglas, J., dissenting).

But we are not asked to consider the permissibility of any of the several methods employed in carrying out the death sentence. Nor are we asked, at least as part of the core submission in these cases, to determine whether the penalty might be a grossly excessive punishment for some specific criminal conduct. Either inquiry would call for a discriminating evaluation of particular means, or of the relationship between particular conduct and its punishment. Petitioners’ principal argument goes far beyond the traditional process of case-by-case inclusion and exclusion. Instead the argument insists on an unprecedented constitutional rule of absolute prohibition of capital punishment for any crime, regardless of its depravity and impact on society. In calling for a precipitate and final judicial end to this form of penalty as offensive to evolving standards of decency, petitioners would have this Court abandon the traditional and more refined approach consistently followed in its prior Eighth Amendment precedents. What they are saying, in effect, is that the evolutionary *431process has come suddenly to an end; that the ultimate wisdom as to the appropriateness of capital punishment under all circumstances, and for all future generations, has somehow been revealed.

The prior opinions of this Court point with great clarity to reasons why those of us who sit on this Court at a particular time should act with restraint before assuming, contrary to a century of precedent, that we now know the answer for all time to come. First, where as here, the language of the applicable provision provides great leeway and where the underlying social policies are felt to be of vital importance, the temptation to read personal preference into the Constitution is understandably great. It is too easy to propound our subjective standards of wise policy under the rubric of more or less universally held standards of decency. See Trop v. Dulles, 356 U. S., at 103 (Warren, C. J.), 119-120 (Frankfurter, J., dissenting); Louisiana ex rel. Francis v. Resweber, 329 U. S., at 470-471 (Frankfurter, J., concurring) ; Weems v. United States, 217 U. S., at 378-379 (McKenna, J.).

The second consideration dictating judicial self-restraint arises from a proper recognition of the respective roles of the legislative and judicial branches. The designation of punishments for crimes is a matter peculiarly within the sphere of the state and federal legislative bodies. See, e. g., In re Kemmler, 136 U. S., at 447; Trop v. Dulles, 356 U. S., at 103. When asked to encroach on the legislative prerogative we are well counseled to proceed with the utmost reticence. The review of legislative choices, in the performance of our duty to enforce the Constitution, has been characterized most appropriately by Mr. Justice Holmes as “the gravest and most delicate duty that this Court is called on to perform.” Blodgett v. Holden, 275 U. S. 142, 147-148 (1927) (separate opinion).

*432How much graver is that duty when we are not asked to pass on the constitutionality of a single penalty under the facts of a single case but instead are urged to overturn the legislative judgments of 40 state legislatures as well as those of Congress. In so doing is the majority able to claim, as did the Court in Weems, that it appreciates “to the fullest the wide range of power that the legislature possesses to adapt its penal laws to conditions as they may exist and punish the crimes of men according to their forms and frequency”? 217 U. S., at 379. I think not. No more eloquent statement of the essential separation of powers limitation on our prerogative can be found than the admonition of Mr. Justice Frankfurter, dissenting in Trop. His articulation of the traditional view takes on added significance where the Court undertakes to nullify the legislative judgments of the Congress and four-fifths of the States.

“What is always basic when the power of Congress to enact legislation is challenged is the appropriate approach to judicial review of congressional legislation .... When the power of Congress to pass a statute is challenged, the function of this Court is to determine whether legislative action lies clearly outside the constitutional grant of power to which it has been, or may fairly be, referred. In making this determination, the Court sits in judgment on the action of a co-ordinate branch of the Government while keeping unto itself — as it must under our constitutional system — the final determination of its own power to act. . . .

“Rigorous observance of the difference between limits of power and wise exercise of power — between questions of authority and questions of prudence — requires the most alert appreciation of this decisive but subtle relationship of two concepts that too easily coalesce. No less does it require a *433disciplined will to adhere to the difference. It is not easy to stand aloof and allow want of wisdom to prevail, to disregard one’s own strongly held view of what is wise in the conduct of affairs. But it is not the business of this Court to pronounce policy. It must observe a fastidious regard for limitations on its own power, and this precludes the Court’s giving effect to its own notions of what is wise or politic. That self-restraint is of the essence in the observance of the judicial oath, for the Constitution has not authorized the judges to sit in judgment on the wisdom of what Congress and the Executive Branch do.” 356 U. S., at 119-120.

See also Mr. Justice White’s dissenting opinion in Weems v. United States, 217 U. S., at 382.

IV

Although determining the range of available punishments for • a particular crime is a legislative function, the very presence of the Cruel and Unusual Punishments Clause within the Bill of Rights requires, in the context of a specific case, that courts decide whether particular acts of the Congress offend that Amendment. The Due Process Clause of the Fourteenth Amendment imposes on the judiciary a similar obligation to scrutinize state legislation. But the proper exercise of that constitutional obligation in the cases before us today must be founded on a full recognition of the several considerations set forth above — the affirmative references to capital punishment in the Constitution, the prevailing precedents of this Court, the limitations on the exercise of our power imposed by tested principles of judicial self-restraint, and the duty to avoid encroachment on the powers conferred upon state and federal legislatures. In the face of these considerations, only the most con-*434elusive of objective demonstrations could warrant this Court in holding capital punishment per se unconstitutional. The burden of seeking so sweeping a decision against such formidable obstacles is almost insuperable. Viewed from this perspective, as I believe it must be, the case against the death penalty falls far short.

Petitioners’ contentions are premised, as indicated above, on the long-accepted view that concepts embodied in the Eighth and Fourteenth Amendments evolve. They present, with skill and persistence, a list of “objective indicators” which are said to demonstrate that prevailing standards of human decency have progressed to the final point of requiring the Court to hold, for all cases and for all time, that capital punishment is unconstitutional.

Briefly summarized, these proffered indicia of contemporary standards of decency include the following: (i) a worldwide trend toward the disuse of the death penalty;16 (ii) the reflection in the scholarly literature of a progressive rejection of capital punishment founded essentially on moral opposition to such treatment;17 (iii) the decreasing numbers of executions over the last 40 years and especially over the last decade;18 (iv) the *435small number of death sentences rendered in relation to the number of cases in which they might have been imposed; 19 and (v) the indication of public abhorrence of *436the penalty reflected in the circumstance that executions are no -longer public affairs.20 The foregoing is an incomplete summary but it touches the major bases of petitioners’ presentation. Although they are not appropriate for consideration as objective evidence, petitioners strongly urge two additional propositions. They contend, first, that the penalty survives public condemnation only through the infrequency, arbitrariness, and discriminatory nature of its application, and, second, that there no longer exists any legitimate justification for the utilization of the ultimate penalty. These contentions, which have proved persuasive to several of the Justices constituting the majority, deserve separate consideration and will be considered in the ensuing sections. Before turning to those arguments, I first address the argument based on “objective” factors.

Any attempt to discern contemporary standards of decency through the review of objective factors must take into account several overriding considerations which petitioners choose to discount or ignore. In a democracy *437the first indicator of the public’s attitude must always be found in the legislative judgments of the people’s chosen representatives. Mr. Justice Marshall’s opinion today catalogues the salient statistics. Forty States,21 the District of Columbia, and the Federal Government still authorize the death penalty for a wide variety of crimes. That number has remained relatively static since the end of World War I. Ante, at 339-341. That does not mean, however, that capital punishment has become a forgotten issue in the legislative arena. As recently as January, 1971, Congress approved the death penalty for congressional assassination. 18 U. S. C. § 351. In 1965 Congress added the death penalty for presidential and vice presidential assassinations. 18 U. S. C. § 1751. Additionally, the aircraft piracy statute passed in 1961 also carries the death penalty. 49 U. S. C. § 1472 (i). Mr. Justice Blackmun’s dissenting opinion catalogues the impressive ease with which each of these statutes was approved. Ante, at 412-413. On the converse side, a bill proposing the abolition of capital punishment for all federal crimes was introduced in 1967 but failed to reach the Senate floor.22

At the state level, New York, among other States, has recently undertaken reconsideration of its capital crimes. A law passed in 1965 restricted the use of capital punishment to the crimes of murder of a police officer and murder by a person serving a sentence of life imprisonment. N. Y. Penal Code § 125.30 (1967).

I pause here to state that I am at a loss to under*438stand how those urging this Court to pursue a course of absolute abolition as a matter of constitutional judgment can draw any support from the New York experience. As is also the case with respect to recent legislative activity in Canada23 and Great Britain,24 New York’s decision to restrict the availability of the death penalty is a product of refined and discriminating legislative judgment, reflecting, not the total rejection of capital punishment as inherently cruel, but a desire to limit it to those circumstances in which legislative judgment deems retention to be in the public interest. No such legislative flexibility is permitted by the contrary course petitioners urge this Court to follow.25

In addition to the New York experience, a number of other States have undertaken reconsideration of capital punishment in recent years. In four States the penalty has been put to a vote of the people through public referenda — a means likely to supply objective evidence of community standards. In Oregon a referendum seeking abolition of capital punishment failed in 1958 but was subsequently approved in 1964.26 Two years later the penalty was approved in Colorado by a wide margin.27 *439In Massachusetts in 1968, in an advisory referendum, the voters there likewise recommended retention of the penalty. In 1970, approximately 64% of the voters in Illinois approved the penalty.28 In addition, the National Commission on Reform of Federal Criminal Laws reports that legislative committees in Massachusetts, Pennsylvania, and Maryland recommended abolition, while committees in New Jersey and Florida recommended retention.29 The legislative views of other States have been summarized by Professor Hugo Bedau in his compilation of sources on capital punishment entitled The Death Penalty in America:

“What our legislative representatives think in the two score states which still have the death penalty may be inferred from the fate of the bills to repeal or modify the death penalty filed during recent years in the legislatures of more than half of these states. In about a dozen instances, the bills emerged from committee for a vote. But in none except Delaware did they become law. In those states where these bills were brought to the floor of the legislatures, the vote in most instances wasn’t even close.” 30

This recent history of activity with respect to legislation concerning the death penalty abundantly refutes the abolitionist position.

The second and even more direct source of information *440reflecting the public’s attitude toward capital punishment is the jury. In Witherspoon v. Illinois, 391 U. S. 510 (1968), Mr. Justice Stewart, joined by Justices Brennan and Marshall, characterized the jury’s historic function in the sentencing process in the following terms:

“[T]he jury is given broad discretion to decide whether or not death is ‘the proper penalty’ in a given case, and a juror’s general views about capital punishment play an inevitable role in any such decision.

“A man who opposes the death penalty, no less than one who favors it, can make the discretionary judgment entrusted to him by the State and can thus obey the oath he takes as a juror. . . . Guided by neither rule nor standard, ... a jury that must choose between life imprisonment and capital punishment can do little more — and must do nothing less — than express the conscience of the community on the ultimate question of life or death.”

“[0]ne of the most important functions any jury can perform in making such a selection is to maintain a link between contemporary community values and the penal system — a link without which the determination of punishment could hardly reflect ‘the evolving standards of decency that mark the progress of a maturing society.’ Trop v. Dulles, . . .”31

Any attempt to discern, therefore, where the prevailing standards of decency lie must take careful account of *441the jury’s response to the question of capital punishment. During the 1960’s juries returned in excess of a thousand death sentences, a rate of approximately two per week. Whether it is true that death sentences were returned in less than 10% of the cases as petitioners estimate or whether some higher percentage is more accurate,32 these totals simply do not support petitioners’ assertion at oral argument that “the death penalty is virtually unanimously repudiated and condemned by the conscience of contemporary society.” 33 It is also worthy of note that the annual rate of death sentences has remained relatively constant over the last 10 years and that the figure for 1970 — 127 sentences — is the highest annual total since 1961.34 It is true that the sentencing rate might be expected to rise, rather than remain constant, when the number of violent crimes increases as it has in this country.35 And it may be conceded that the constancy in these statistics indicates the unwillingness of juries to demand the ultimate penalty in many cases where it might be imposed. But these considerations fall short of indicating that juries are imposing the death penalty with such rarity as to justify this Court in reading into this circumstance a public rejection of capital punishment.36

*442One must conclude, contrary to petitioners’ submission, that the indicators most likely to reflect the public’s view — legislative bodies, state referenda and the juries which have the actual responsibility — do not support the contention that evolving standards of decency require total abolition of capital punishment.37 Indeed, *443the weight of the evidence indicates that the public generally has not accepted either the morality or the social merit of the views so passionately advocated by the articulate spokesmen for abolition. But however one may assess the amorphous ebb and flow of public opinion generally on this volatile issue, this type of inquiry lies at the periphery — not the core — of the judicial process in constitutional cases. The assessment of popular opinion is essentially a legislative, not a judicial, function.

V

Petitioners seek to salvage their thesis by arguing that the infrequency and discriminatory nature of the actual resort to the ultimate penalty tend to diffuse public opposition. We are told that the penalty is imposed exclusively on uninfluential minorities — “the poor and powerless, personally ugly and socially unacceptable.” 38 It is urged that this pattern of application assures that large segments of the public will be either uninformed or unconcerned and will have no reason to measure the punishment against prevailing moral standards.

Implicitly, this argument concedes the unsoundness of petitioners’ contention, examined above under Part IV, that objective evidence shows a present and widespread community rejection of the death penalty. It is now said, *444in effect, not that capital punishment presently offends our citizenry, but that the public would be offended if the penalty were enforced in a nondiscriminatory manner against a significant percentage of those charged with capital crimes, and if the public were thereby made aware of the moral issues surrounding capital punishment. Rather than merely registering the objective indicators on a judicial balance, we are asked ultimately to rest a far-reaching constitutional determination on a prediction regarding the subjective judgments of the mass of our people under hypothetical assumptions that may or may not be realistic.

Apart from the impermissibility of basing a constitutional judgment of this magnitude on such speculative assumptions, the argument suffers from other defects. If, as petitioners urge, we are to engage in speculation, it is not at all certain that the public would experience deep-felt revulsion if the States were to execute as many sentenced capital offenders this year as they executed in the mid-1930’s.39 It seems more likely that public reaction, rather than being characterized by undifferentiated rejection, would depend upon the facts and circumstances surrounding each particular case.

Members of this Court know, from the petitions and appeals that come before us regularly, that brutish and revolting murders continue to occur with disquieting frequency. Indeed, murders are so commonplace *445in our society that only the most sensational receive significant and sustained publicity. It could hardly be suggested that in any of these highly publicized murder cases — the several senseless assassinations or the too numerous shocking multiple murders that have stained this country’s recent history — the public has exhibited any signs of “revulsion” at the thought of executing the convicted murderers. The public outcry, as we all know, has been quite to the contrary. Furthermore, there is little reason to suspect that the public’s reaction would differ significantly in response to other less publicized murders. It is certainly arguable that many such murders, because of their senselessness or barbarousness, would evoke a public demand for the death penalty rather than a public rejection of that alternative. Nor is there any rational basis for arguing that the public reaction to any of these crimes would be muted if the murderer were “rich and powerful.” The demand for the ultimate sanction might well be greater, as a wealthy killer is hardly a sympathetic figure. While there might be specific cases in which capital punishment would be regarded as excessive and shocking to the conscience of the "community, it can hardly be argued that the public’s dissatisfaction with the penalty in particular cases would translate into a demand for absolute abolition.

In pursuing the foregoing speculation, I do not suggest that it is relevant to the appropriate disposition of these cases. The purpose of the digression is to indicate that judicial decisions cannot be founded on such speculations and assumptions, however appealing they may seem.

But the discrimination argument does not rest alone on a projection of the assumed effect on public opinion of more frequent executions. Much also is made of the undeniable fact that the death penalty has a greater impact on the lower economic strata of society, which *446include a relatively higher percentage of persons of minority racial and ethnic group backgrounds. The argument drawn from this fact is two-pronged. In part it is merely an extension of the speculative approach pursued by petitioners, %. e., that public revulsion is suppressed in callous apathy because the penalty does not affect persons from the white middle class which constitutes the majority in this country. This aspect, however, adds little to the infrequency rationalization for public apathy which I have found unpersuasive.

As Mr. Justice Marshall’s opinion today demonstrates, the argument does have a more troubling aspect. It is his contention that if the average citizen were aware of the disproportionate burden of capital punishment borne by the “poor, the ignorant, and the underprivileged,” he would find the penalty “shocking to his conscience and sense of justice” and would not stand for its further use. Ante, at 365-366, 369. This argument, like the apathy rationale, calls for further speculation on the part of the Court. It also illuminates the quicksands upon which we are asked to base this decision. Indeed, the two contentions seem to require contradictory assumptions regarding the public’s moral attitude toward capital punishment. The apathy argument is predicated on the assumption that the penalty is used against the less influential elements of society, that the public is fully aware of this, and that it tolerates use of capital punishment only because of a callous indifference to the offenders who are sentenced. Mr. Justice Marshall’s argument, on the other hand, rests on the contrary assumption that the public does not know against whom the penalty is enforced and that if the public were educated to this fact it would find the punishment intolerable. Ante, at 369. Neither assumption can claim to be an entirely accurate portrayal of public attitude; for some acceptance of capital punishment might be a conse*447quence of hardened apathy based on the knowledge of infrequent and uneven application, while for others acceptance may grow only out of ignorance. More significantly, however, neither supposition acknowledges what, for me, is a more basic flaw.

Certainly the claim is justified that this criminal sanction falls more heavily on the relatively impoverished and underprivileged elements of society. The “have-nots” in every society always have been subject to greater pressure to commit crimes and to fewer constraints than their more affluent fellow citizens. This is, indeed, a tragic byproduct of social and economic deprivation, but it is not an argument of constitutional proportions under the Eighth or Fourteenth Amendment. The same discriminatory impact argument could be made with equal force and logic with respect to those sentenced to prison terms. The Due Process Clause admits of no distinction between the deprivation of “life” and the deprivation of “liberty.” If discriminatory impact renders capital punishment cruel and unusual, it likewise renders invalid most of the prescribed penalties for crimes of violence. The root causes of the higher incidence of criminal penalties on “minorities and the poor” will not be cured by abolishing the system of penalties. Nor, indeed, could any society have a viable system of criminal justice if sanctions were abolished or ameliorated because most of those who commit crimes happen to be underprivileged. The basic problem results not from the penalties imposed for criminal conduct but from social and economic factors that have plagued humanity since the beginning of recorded history, frustrating all efforts to create in any country at any time the perfect society in which there are no “poor,” no “minorities” and no “underprivileged.” 40 *448The causes underlying this problem are unrelated to the constitutional issue before the Court.

Finally, yet another theory for abolishing the death penalty — reflected in varying degrees in each of the concurring opinions today — is predicated on the discriminatory impact argument. Quite apart from measuring the public’s acceptance or rejection of the death penalty under the “standards of decency” rationale, Mr. Justice Douglas finds the punishment cruel and unusual because it is “arbitrarily” invoked. He finds that “the basic theme of equal protection is implicit” in the Eighth Amendment, and that the Amendment is violated when jury sentencing may be characterized as arbitrary or discriminatory. Ante, at 249. While Mr. Justice Stewart does not purport to rely on notions of equal protection, he also rests primarily on what he views to. be a history of arbitrariness. Ante, at 309-310.41 Whatever may be the facts with respect to jury sentencing, this argument calls for a reconsideration of the “standards” aspects of the Court’s decision in McGautha v. California, 402 U. S. 183 (1971). Although that is the unmistakable thrust of these opinions today, I see no reason to reassess the standards question considered so carefully in Mr. Justice Harlan’s opinion for the Court *449last Term. Having so recently reaffirmed our historic dedication to entrusting the sentencing function to the jury’s “untrammeled discretion” (id., at 207), it is difficult to see how the Court can now hold the entire process constitutionally defective under the Eighth Amendment. For all of these reasons I find little merit in the various discrimination arguments, at least in the several lights in which they have been cast in these cases.

Although not presented by any of the petitioners today, a different argument, premised on the Equal Protection Clause, might well be made. If a Negro defendant, for instance, could demonstrate that members of his race were being singled out for more severe punishment than others charged with the same offense, a constitutional violation might be established. This was the contention made in Maxwell v. Bishop, 398 F. 2d 138 (CA8 1968), vacated and remanded on other grounds, 398 U. S. 262 (1970), in which the Eighth Circuit was asked to issue a writ of habeas corpus setting aside a death sentence imposed on a Negro defendant convicted of rape. In that case substantial statistical evidence was introduced tending to show a pronounced disproportion in the number of Negroes receiving death sentences for rape in parts of Arkansas and elsewhere in the South. That evidence was not excluded but was found to be insufficient to show discrimination in sentencing in Maxwell’s trial. Me. Justice Blackmun, then sitting on the Court of Appeals for the Eighth Circuit, concluded:

“The petitioner’s argument is an interesting one and we are not disposed to say that it could not have some validity and weight in certain situations. Like the trial court, however ... we feel that the argument does not have validity and pertinent application to Maxwell’s case.

*450“We are not yet ready to condemn and upset the result reached in every case of a Negro rape defendant in the State of Arkansas on the basis of broad theories of social and statistical injustice. . . .

“We do not say that there is no ground for suspicion that the death penalty for rape may have been discriminatorily applied over the decades in that large area of states whose statutes provide for it. There are recognizable indicators of this. But . . . improper state practice of the past does not automatically invalidate a procedure of the present. . . .” Id., at 146-148.

I agree that discriminatory application of the death penalty in the past, admittedly indefensible, is no justification for holding today that capital punishment is invalid in all cases in which sentences were handed out to members of the class discriminated against. But Maxwell does point the way to a means of raising the equal protection challenge that is more consonant with precedent and the Constitution’s mandates than the several courses pursued by today’s concurring opinions.

A final comment on the racial discrimination problem seems appropriate. The possibility of racial bias in the trial and sentencing process has diminished in recent years. The segregation of our society in decades past, which contributed substantially to the severity of punishment for interracial crimes, is now no longer prevalent in this country. Likewise, the day is past when juries do not represent the minority group elements of the community. The assurance of fair trials for all citizens is greater today than at any previous time in our history. Because standards of criminal justice have “evolved” in a manner favorable to the accused, discriminatory imposition of capital punishment is far less likely today than in the past.

*451VI

Petitioner in Branch v. Texas, No. 69-5031, and to a lesser extent the petitioners in the other cases before us today, urge that capital punishment is cruel and unusual because it no longer serves any rational legislative interests. Before turning to consider whether any of the traditional aims of punishment justify the death penalty, I should make clear the context in which I approach this aspect of the cases.

First, I find no support — in the language of the Constitution, in its history, or in the cases arising under it — for the view that this Court may invalidate a category of penalties because we deem less severe penalties adequate to serve the ends of penology. While the cases affirm our authority to prohibit punishments that are cruelly inhumane (e. g., Wilkerson v. Utah, 99 U. S., at 135-136; In re Kemmler, 136 U. S., at 447), and punishments that are cruelly excessive in that they are disproportionate to particular crimes (see Part VII, infra), the precedents of this Court afford no basis for striking- down a particular form of punishment because we may be persuaded that means less stringent would be equally efficacious.

Secondly, if we were free to question the justifications for the use of capital punishment, a heavy burden would rest on those who attack the legislatures’ judgments to prove the lack of rational justifications. This Court has long held that legislative decisions in this area, which lie within the special competency of that branch, are entitled to a presumption of validity. See, e. g., Trop v. Dulles, 356 U. S., at 103; Louisiana ex rel. Francis v. Resweber, 329 U. S., at 470 (Frankfurter, J., concurring) ; Weems v. United States, 217 U. S., at 378-379; In re Kemmler, 136 U. S., at 449.

*452I come now to consider, subject to the reservations above expressed, the two justifications most often cited for the retention of capital punishment. The concept of retribution — though popular for centuries — is now criticized as unworthy of a civilized people. Yet this Court has acknowledged the existence of a retributive element in criminal sanctions and has never heretofore found it impermissible. In Williams v. New York, 337 U. S. 241 (1949), Mr. Justice Black stated that,

“Retribution is no longer the dominant objective of the criminal law. Reformation and rehabilitation of offenders have become important goals of criminal jurisprudence.” Id., at 248.

It is clear, however, that the Court did not reject retribution altogether. The record in that case indicated that one of the reasons why the trial judge imposed the death penalty was his sense of revulsion at the “shocking details of the crime.” Id., at 244. Although his motivation was clearly retributive, the Court upheld the trial judge’s sentence.42 Similarly, Mr. Justice Marshall noted in his plurality opinion in Powell v. Texas, 392 U. S. 514, 530 (1968), that this Court “has never held that anything in the Constitution requires that penal sanctions be designed solely to achieve therapeutic or rehabilitative effects.”43

*453While retribution alone may seem an unworthy justification in a moral sense, its utility in a system of criminal justice requiring public support has long been recognized. Lord Justice Denning, now Master of the Rolls of the Court of Appeal in England, testified on this subject before the British Royal Commission on Capital Punishment:

“Many are inclined to test the efficacy of punishment solely by its value as a deterrent: but this is too narrow a view. Punishment is the way in which society expresses its denunciation of wrong doing: and, in order to maintain respect for law, it is essential that the punishment inflicted for grave crimes should adequately reflect the revulsion felt by the great majority of citizens for them. It is a mistake to consider the objects of punishment as being deterrent or reformative or preventive and nothing else. If this were so, we should not send to prison a man who was guilty of motor manslaughter, but only disqualify him from driving; but would public opinion be content with this? The truth is that some crimes are so outrageous that society insists on adequate punishment, because the wrong-doer deserves it, irrespective of whether it is a deterrent or not.” 44

The view expressed by Lord Denning was cited approvingly in the Royal Commission’s Report, recognizing “a *454strong and widespread demand for retribution.” 45 Mr. Justice Stewart makes much the same point in his opinion today when he concludes that expression of man’s retributive instincts in the sentencing process “serves an important purpose in promoting the stability of a society governed by law.” Ante, at 308. The view, moreover, is not without respectable support in the jurisprudential literature in this country,46 despite a substantial body of opinion to the contrary.47 And it is conceded on all sides that, not infrequently, cases arise that are so shocking or offensive that the public demands the ultimate penalty for the transgressor.

Deterrence is a more appealing justification, although opinions again differ widely. Indeed, the deterrence issue lies at the heart of much of the debate between the abolitionists and retentionists.48 Statistical studies, based primarily on trends in States that have abolished the penalty, tend to support the view that the death penalty has not been proved to be a superior deterrent.49 Some dispute the validity of this conclusion,50 pointing *455out that the studies do not show that the death penalty has no deterrent effect on any categories of crimes. On the basis of the literature and studies currently available, I find myself in agreement with the conclusions drawn by the Royal Commission following its exhaustive study of this issue:

“The general conclusion which we reach, after careful review of all the evidence we have been able to obtain as to the deterrent effect of capital punishment, may be stated as follows. Prima facie the penalty of death is likely to have a stronger effect as a deterrent to normal human beings than any other form of punishment, and there is some evidence (though no convincing statistical evidence) that this is in fact so. But this effect does not operate universally or uniformly, and there are many offenders on whom it is limited and may often be negligible. It is accordingly important to view this question in a just perspective and not base a penal policy in relation to murder on exaggerated estimates of the uniquely deterrent force of the death penalty.” 51

Only recently this Court was called on to consider the deterrence argument in relation to punishment by fines i for public drunkenness. Powell v. Texas, 392 U. S. 514 (1968). The Court was unwilling to strike down the Texas statute on grounds that it lacked a rational foundation. What Mb. Justice Marshall said there would seem to have equal applicability in this case:

“The long-standing and still raging debate over the validity of the deterrence justification for penal sanctions has not reached any sufficiently clear conclusions to permit it to be said that such sanctions are ineffective in any particular context or for any *456particular group of people who are able to appreciate the consequences of their acts. . . Id., at 531.

As I noted at the outset of this section, legislative judgments as to the efficacy of particular punishments are presumptively rational and may not be struck down under the Eighth Amendment because this Court may think that some alternative sanction would be more appropriate. Even if such judgments were within the judicial prerogative, petitioners have failed to show that there exist no justifications for the legislative enactments challenged in these cases.52 While the evidence and arguments advanced by petitioners might have proved profoundly persuasive if addressed to a legislative body, they do not approach the showing traditionally required before a court declares that the legislature has acted irrationally.

VII

In two of the cases before us today juries imposed sentences of death after convictions for rape.53 In these cases we are urged to hold that even if capital punishment is permissible for some crimes, it is a cruel and unusual punishment for this crime. Petitioners in these cases rely on the Court's opinions holding that the Eighth Amendment, in addition to prohibiting punishments *457deemed barbarous and inhumane, also condemns punishments that are greatly disproportionate to the crime charged. This reading of the Amendment was first expressed by Mr. Justice Field in his dissenting opinion in O’Neil v. Vermont, 144 U. S. 323, 337 (1892), a case in which a defendant charged with a large number of violations of Vermont’s liquor laws received a fine in excess of $6,600, or a 54-year jail sentence if the fine was not paid. The majority refused to consider the question on the ground that the Eighth Amendment did not apply to the States. The dissent, after carefully examining the history of that Amendment and the Fourteenth, concluded that its prohibition was binding on Vermont and that it was directed against “all punishments which by their excessive length or severity are greatly dispro-portioned to the offences charged.” Id., at 339-340.54

The Court, in Weems v. United States, 217 U. S. 349 (1910), adopted Mr. Justice Field’s view. The defendant, in Weems, charged with falsifying Government documents, had been sentenced to serve 15 years in cadena temporal, a punishment which included carrying chains at the wrists and ankles and the perpetual loss of the right to vote and hold office. Finding the sentence grossly excessive in length and condition of imprisonment, the Court struck it down. This notion of disproportionality — that particular sentences may be cruelly excessive for particular crimes — has been cited with approval in more recent decisions of this Court. See Robinson v. California, 370 U. S., at 667; Trop v. Dulles, 356 U. S., at 100; see also Howard v. Fleming, 191 U. S. 126, 135-136 (1903).

These cases, while providing a rationale for gauging the constitutionality of capital sentences imposed for rape, *458also indicate the existence of necessary limitations on the judicial function. The use of limiting terms in the various expressions of this test found in the opinions — grossly excessive, greatly disproportionate — emphasizes that the Court’s power to strike down punishments as excessive must be exercised with the greatest circumspection. As I have noted earlier, nothing in the history of the Cruel and Unusual Punishments Clause indicates that it may properly be utilized by the judiciary to strike down punishments — authorized by legislatures and imposed by juries — in any but the extraordinary case. This Court is not empowered to sit as a court of sentencing review, implementing the personal views of its members on the proper role of penology. To do so is to usurp a function committed to the Legislative Branch and beyond the power and competency of this Court.

Operating within these narrow limits, I find it quite impossible to declare the death sentence grossly excessive for all rapes. Rape is widely recognized as among the most serious of violent crimes, as witnessed by the very fact that it is punishable by death in 16 States and by life imprisonment in most other States.55 The several reasons why rape stands so high on the list of serious crimes are well known: It is widely viewed as the most atrocious of intrusions upon the privacy and dignity of the victim; never is the crime committed accidentally; rarely can it be said to be unpremeditated; *459often the victim suffers serious physical injury; the psychological impact can often be as great as the physical consequences; in a real sense, the threat of both types of injury is always present.56 For these reasons, and for the reasons arguing against abolition of the death penalty altogether, the excessiveness rationale provides no basis for rejection of the penalty for rape in all cases.

The argument that the death penalty for rape lacks rational justification because less severe punishments might be viewed as accomplishing the proper goals of penology is as inapposite here as it was in considering per se abolition. See Part VI supra. The state of knowledge with respect to the deterrent value of the sentence for this crime is inconclusive.57 Moreover, what has been said about the concept of retribution applies with equal force where the crime is rape. There are many cases in which the sordid, heinous nature of a particular crime, demeaning, humiliating, and often physically or psychologically traumatic, will call for public condemnation. In a period in our country’s history when the frequency of this crime is increasing alarmingly,58 it is indeed a grave event for the Court to take from the States whatever deterrent and retributive weight the death penalty retains.

Other less sweeping applications of the dispropor-tionality concept have been suggested. Recently the Fourth Circuit struck down a death sentence in Ralph v. Warden, 438 F. 2d 786 (1970), holding that the death penalty was an appropriate punishment for rape *460only where life is “endangered.” Chief Judge Hayns-worth, who joined in the panel’s opinion, wrote separately in denying the State of Maryland’s petition for rehearing in order to make clear the basis for his joinder. He stated that, for him, the appropriate test was not whether life was endangered, but whether the victim in fact suffered “grievous physical or psychological harm.” Id., at 794. See Rudolph v. Alabama, 375 U. S. 889 (1963) (dissent from the denial of certiorari).

It seems to me that both of these tests depart from established principles and also raise serious practical problems. How are those cases in which the victim’s life is endangered to be distinguished from those in which no danger is found? The threat of serious injury is implicit in the definition of rape; the victim is either forced into submission by physical violence or by the threat of violence. Certainly that test would provide little comfort for either of the rape defendants in the cases presently before us. Both criminal acts were accomplished only after a violent struggle. Petitioner Jackson held a scissors blade against his victim’s neck. Petitioner Branch had less difficulty subduing his 65-year-old victim. Both assailants threatened to kill their victims. See Mr. Justice Douglas’ opinion, ante, at 252-253. The alternate test, limiting the penalty to cases in which the victim suffers physical or emotional harm, might present even greater problems of application. While most physical effects may be seen and objectively measured, the emotional impact may be impossible to gauge at any particular point in time. The extent and duration of psychological trauma may not be known or ascertainable prior to the date of trial.

While I reject each of these attempts to establish specific categories of cases in which the death penalty may be deemed excessive, I view them as groping *461toward what is for me the appropriate application of the Eighth Amendment. While in my Anew the dispro-portionality test may not be used either to strike down the death penalty for rape altogether or to install the Court as a tribunal for sentencing review, that test may find its application in the peculiar circumstances of specific cases. Its utilization should be limited to the rare case in which the death penalty is rendered for a crime technically falling within the legislatively defined class but factually falling outside the likely legislative intent in creating the category. Specific rape cases (and specific homicides as well) can be imagined in which the conduct of the accused would render the ultimate penalty a grossly excessive punishment. Although this case-by-case approach may seem painfully slow and inadequate to those who wish the Court to assume an activist legislative role in reforming criminal punishments, it is the approach dictated both by our prior opinions and by a due recognition of the limitations of judicial power. This approach, rather than the majority’s more pervasive and less refined judgment, marks for me the appropriate course under the Eighth Amendment.

VIII

I now return to the overriding question in these cases: whether this Court, acting in conformity with the Constitution, can justify its judgment to abolish capital punishment as heretofore known in this country. It is important to keep in focus the enormity of the step undertaken by the Court today. Not only does it invalidate hundreds of state and federal laws, it deprives those jurisdictions of the power to legislate with respect to capital punishment in the future, except in a manner consistent with the cloudily outlined views of those Justices who do not purport to undertake total abolition. *462Nothing short of an amendment to the United States Constitution can reverse the Court’s judgments. Meanwhile, all flexibility is foreclosed. The normal democratic process, as well as the opportunities for the several States to respond to the will of their people expressed through ballot referenda (as in Massachusetts, Illinois, and Colorado),59 is now shut off.

The sobering disadvantage of constitutional adjudication of this magnitude is the universality and permanence of the judgment. The enduring merit of legislative action is its responsiveness to the democratic process, and to revision and change: mistaken judgments may be corrected and refinements perfected. In England 60 and Canada61 critical choices were made after studies canvassing all competing views, and in those countries revisions may be made in light of experience.62

As recently as 1967 a presidential commission did consider, as part of an overall study of crime in this country, whether the death penalty should be abolished. *463The commission’s unanimous recommendation was as follows:

“The question whether capital punishment is an appropriate sanction is a policy decision to be made by each State. Where it is retained, the types of offenses for which it is available should be strictly limited, and the law should be enforced in an evenhanded and nondiscriminatory manner, with procedures for review of death sentences that are fair and expeditious. When a State finds that it cannot administer the penalty in such a manner, or that the death penalty is being imposed but not carried into effect, the penalty should be abandoned.” 63

The thrust of the Commission’s recommendation, as presently relevant, is that this question “is a policy-decision to be made by each State.” There is no hint that this decision could or should be made by the judicial branch.

The National Commission on Reform of Federal Criminal Laws also considered the capital punishment issue. The introductory commentary of its final report states that “a sharp division [existed] within the Commission on the subject of capital punishment,” although a *464majority favored its abolition.64 Again, consideration of the question was directed to the propriety of retention or abolition as a legislative matter. There was no suggestion that the difference of opinion existing among commission members, and generally across the country, could or should be resolved in one stroke by a decision of this Court.65 Similar activity was, before today, evident at the state level with re-evaluation having been undertaken by special legislative committees in some States and by public ballot in others.66

With deference and respect for the views of the Justices who differ, it seems to me that all these studies — both in this country and elsewhere — suggest that, as a matter of policy and precedent, this is a classic case for the exercise of our oft-announced allegiance to judicial restraint. I know of no case in which greater gravity and delicacy have attached to the duty that this Court is called on to perform whenever legislation — state or federal — is challenged on constitutional grounds.67 It seems to me that the sweeping judicial action undertaken today reflects a *465basic lack of faith and confidence in the democratic process. Many may regret, as I do, the failure of some legislative bodies to address the capital punishment issue with greater frankness or effectiveness. Many might decry their failure either to abolish the penalty entirely or selectively, or to establish standards for its enforcement. But impatience with the slowness, and even the unresponsiveness, of legislatures is no justification for judicial'intrusion upon their historic powers. Rarely has there been a more appropriate opportunity for this Court to heed the philosophy of Mr. Justice Oliver Wendell Holmes. As Mr. Justice Frankfurter reminded the Court in Trop:

“[T]he whole of [Mr. Justice Holmes’] work during his thirty years of service on this Court should be a constant reminder that the power to invalidate legislation must not be exercised as if, either in constitutional theory or in the art of government, it stood as the sole bulwark against unwisdom or excesses of the moment.” 356 U. S., at 128.

Me. Justice Rehnquist,

with whom The Chief Justice, Mr. Justice Blackmun, and Me. Justice Powell join, dissenting.

The Court’s judgments today strike down a penalty that our Nation’s legislators have thought necessary since our country was founded. My Brothers Douglas, Brennan, and Marshall would at one fell swoop invalidate laws enacted by Congress and 40 of the 50 state legislatures, and would consign to the limbo of unconstitutionality under a single rubric penalties for offenses as varied and unique as murder, piracy, mutiny, highjacking, and desertion in the face of the enemy. My Brothers Stewart and White, asserting reliance on a more limited rationale — the reluctance of judges and juries actually to impose the death penalty in the majority of capital *466cases — join in the judgments in these cases. Whatever its precise rationale, today’s holding necessarily brings into sharp relief the fundamental question of the role of judicial review in a democratic society. How can government by the elected representatives of the people co-exist with the power of the federal judiciary, whose members are constitutionally insulated from responsiveness to the popular will, to declare invalid laws duly enacted by the popular branches of government?

The answer, of course, is found in Hamilton’s Federalist Paper No. 78 and in Chief Justice Marshall’s classic opinion in Marbury v. Madison, 1 Cranch 137 (1803). An oft-told story since then, it bears summarization once more. Sovereignty resides ultimately in the people as a whole and, by adopting through their States a written Constitution for the Nation and subsequently adding amendments to that instrument, they have both granted certain powers to the National Government, and denied other powers to the National and the State Governments. Courts are exercising no more than the judicial function conferred upon them by Art. Ill of the Constitution when they assess, in a case before them, whether or not a particular legislative enactment is within the authority granted by the Constitution to the enacting body, and whether it runs afoul of some limitation placed by the Constitution on the authority of that body. For the theory is that the people themselves have spoken in the Constitution, and therefore its commands are superior to the commands of the legislature, which is merely an agent of the people.

The Founding Fathers thus wisely sought to have the best of both worlds, the undeniable benefits of both democratic self-government and individual rights protected against possible excesses of that form of govern merit

The courts in cases properly before them have been entrusted under the Constitution with the last word, short of constitutional amendment, as to whether a law passed *467by the legislature conforms to the Constitution. But just because courts in general, and this Court in particular, do have the last word, the admonition of Mr. Justice Stone dissenting in United States v. Butler must be constantly borne in mind:

“[W]hile unconstitutional exercise of power by the executive and legislative branches of the government is subject to judicial restraint, the only check upon our own exercise of power is our own sense of self-restraint." 297 U. S. 1, 78-79 (1936).

Rigorous attention to the limits of this Court’s authority is likewise enjoined because of the natural desire that beguiles judges along with other human beings into imposing their own views of goodness, truth, and justice upon others. Judges differ only in that they have the power, if not the authority, to enforce their desires. This is doubtless why nearly two centuries of judicial precedent from this Court counsel the sparing use of that power. The most expansive reading of the leading constitutional cases does not remotely suggest that this Court has been granted a roving commission, either by the Founding Fathers or by the framers of the Fourteenth Amendment, to strike down laws that are based upon notions of policy or morality suddenly found unacceptable by a majority of this Court. The Framers of the Constitution would doubtless have agreed with the great English political philosopher John Stuart Mill when he observed:

“The disposition of mankind, whether as rulers or as fellow-citizens, to impose their own opinions and inclinations as a rule of conduct on others, is so energetically supported by some of the best and by some of the worst feelings incident to human nature, that it is hardly ever kept under restraint by anything but want of power.” On Liberty 28 (1885).

*468A separate reason for deference to the legislative judgment is the consequence of human error on the part of the judiciary with respect to the constitutional issue before it. Human error there is bound to be, judges being men and women, and men and women being what they are. But an error in mistakenly sustaining the constitutionality of a particular enactment, while wrongfully depriving the individual of a right secured to him by the Constitution, nonetheless does so by simply letting stand a duly enacted law of a democratically chosen legislative body. The error resulting from a mistaken upholding of an individual’s constitutional claim against the validity of a legislative enactment is a good deal more serious. For the result in such a case is not to leave standing a law duly enacted by a representative assembly, but to impose upon the Nation the judicial fiat of a majority of a court of judges whose connection with the popular will is remote at best.

The task of judging constitutional cases imposed by Art. Ill cannot for this reason be avoided, but it must surely be approached with the deepest humility and genuine deference to legislative judgment. Today’s decision to invalidate capital punishment is, I respectfully submit, significantly lacking in those attributes. For the reasons well stated in the opinions of The Chief Justice, Mr. Justice Blackmun, and Mr. Justice Powell, I conclude that this decision holding unconstitutional capital punishment is not an act of judgment, but rather an act of will. It completely ignores the strictures of Mr. Justice Holmes, writing more than 40 years ago in Baldwin v. Missouri:

“I have not yet adequately expressed the more than anxiety that I feel at the ever increasing scope given to the Fourteenth Amendment in cutting down what I believe to be the constitutional rights of the States. As the decisions now stand, I see hardly *469any limit but the sky to the invalidating of those rights if they happen to strike a majority of this Court as for any reason undesirable. I cannot believe that the Amendment was intended to give us carte blanche to embody our economic or moral beliefs in its prohibitions. Yet I can think of no narrower reason that seems to me to justify the present and the earlier decisions to which I have referred. Of course the words 'due process of law/ if taken in their literal meaning, have no application to this case; and while it is too late to deny that they have been given a much more extended and artificial signification, still we ought to remember the great caution shown by the Constitution in limiting the power of the States, and should be slow to construe the clause in the Fourteenth Amendment as committing to the Court, with no guide but the Court’s own discretion, the validity of whatever laws the States may pass.” 281 U. S. 586, 595 (1930) (dissenting opinion).

More than 20 years ago, Justice Jackson made a similar observation with respect to this Court’s restriction of the States in the enforcement of their own criminal laws:

“The use of the due process clause to disable the States in protection of society from crime is quite as dangerous and delicate a use of federal judicial power as to use it to disable them from social or economic experimentation.” Ashcraft v. Tennessee, 322 U. S. 143, 174 (1944) (dissenting opinion).

If there can be said to be one dominant theme in the Constitution, perhaps more fully articulated in the Federalist Papers than in the instrument itself, it is the notion of checks and balances. The Framers were well aware of the natural desire of office holders as well as others to seek to expand the scope and authority of their *470particular office at the expense of others. They sought to provide against success in such efforts by erecting adequate checks and balances in the form of grants of authority to each branch of the government in order to counteract and prevent usurpation on the part of the others.

This philosophy of the Framers is best described by one of the ablest and greatest of their number, James Madison, in Federalist No. 51:

“In framing a government which is to be administered by men over men, the great difficulty lies in this: You must first enable the government to controul the governed; and in the next place, oblige it to controul itself.”

Madison’s observation applies to the Judicial Branch with at least as much force as ti> the Legislative and Executive Branches. While overreaching by the Legist lative and Executive Branches may result in the sacrifice of individual protections that the Constitution was designed to secure against action of the State, judicial overreaching may result in sacrifice of the equally important right of the people ti> govern themselves. The Due Process and Equal Protection Clauses of the Fourteenth Amendment were “never intended to destroy the States’ power to govern themselves.” Black, J., in Oregon v. Mitchell, 400 U. S. 112, 126 (1970).

The very nature of judicial review, as pointed out by Justice Stone in his dissent in the Butler case, makes the courts the least subject to Madisonian check in the'event that they shall, for the best of motives, expand judicial authority beyond the limits contemplated by the Framers. It is for this reason that judicial self-restraint is surely an implied, if not an expressed, condition of the grant of authority of judicial review. The Court’s holding in these cases has been reached, I believe, in complete disregard of that implied condition.

MEMORANDUM AND OPINION SETTING OUT FINDINGS OF FACT AND CONCLUSIONS OF LAW

Lee H. Rosenthal, Chief United States District Judge

Introduction... 1057

I.Findings of Fact... 1060

A. Procedural Background... 1060

B. The Evidence in the Record... 1061

1. The Parties... 1062

2. The Fact Witnesses... 1064

3. The Expert Witnesses... 1066

4. Overview of the Factual and Legal Issues,..1067

C. The Historical Development of Bail in the United States and in Hams County. ..1068

1. The Constitutionalization of Bail. ..1068

2. Statutory and Judicial Bail Reform: Pretrial Services, Probable Cause Hearings, and “Meaningful” Alternatives to Secured Money Bail... 1070

3. Bail at the Federal Level... 1073

4. Bail under Texas Law... 1076

5. Recent Distinctions Drawn Between Bail and Preventive Detention.. .1078

a. Washington, D.C.... 1078

b. New Mexico... 1079

c. New Jersey., .1079

d. New Orleans.,. 1080

e. Maryland... 1080

f. Alabama.. .1081

g. Calhoun, Georgia.. .1082

h. Conclusion... 1084

D. The Use of Bail in Hands County Misdemeanor Pretrial Detention.. .1084

1. The Statutory Framework... 1084

2. Arrest and Booking... 1087

3. The Probable Cause and Bail-Setting Hearing.. .1092.

4. The First Appearance Before a County Judge... 1101

5. Disposition of Misdemeanor Cases... 1104

6. The Use of Bail to Detain... 1107

E. The Population Statistics of Misdemeanor Detainees at Each Stage in the Post-arrest Process... 1111

*10571. Arrestees Detained More than 24 Hours Before the Probable Cause Hearing... 1111

2. Arrestees Detained More than 48 Hours Before a Bail Review... 1112

3. Arrestees Detained Until Case Disposition. ..1114

4. Arrestees Detained “Because of’ Indigence.. .1114

5. Bond Forfeitures and Re-Arrests for New Criminal Activity... 1117

F. The Effects of Pretrial Detention on Misdemeanor Defendants Who Cannot Pay Secured Money Bail... 1121

G. Comparisons to Other Jurisdictions... 1122

H. Proposed Bail Reforms... 1124

I. Changes to Risk Assessment... 1124

2. Changes to the System’s Efficiency. ..1125

3. Changes to the Probable Cause Hearings... 1126

4. Texas House Bill 3011 / Senate Bill 1338...1127

I.Conclusions . on Findings of Fact... 1129 .

II.Conclusions of Law.. .1132

A. The Legal Standards... 1132

B. Likelihood of Success on the Merits... 1138-

1. The Standard of Review... 1133

a. Equal Protection... 1134

b. Due Process... 1139

2. The Constitutional Requirements .".. 1140

a. Equal Protection... 1140

b. Due Process... 1140

c. Excessive Bail.., 1147

3. Harris County Policies that Violate Constitutional Requirements... 1148

a. Municipal Liability under § 1983...1148

b. The County Judges’ Policies and Customs: Equal Protection.. .1149

c. The County Judges’ Policies and Customs: Due Process... 1153

d. The Sheriffs Policies under Equal Protection and Due Process ... 1154

4. Judicial Conduct that Violates Constitutional Requirements.. .1155

5. Conclusion' on Likelihood of Success on the Merits... 1156

C. Irreparable Injury .. .1157

D. Balancing the Harms. .■. 1158

E. The Public Interest... 1159

F. Bond...1159

III. Remedy.. .1160

IV. Conclusion.. .1166

A. Summary Judgment... 1166

B. Preliminary Injunction... 1167

Introduction

“Twenty years ago, not quite one-third of [Texas’s] jail population was awaiting trial. Now the number is three-fourths. Liberty is precious to Americans, and any deprivation'must be scrutinized. To protect public safety and ensure that those accused of a crime will appear at trial, persons charged with breaking the law may be detained before their guilt or innocence can be adjudicated, but that detention must not extend beyond its justifications. Many who are arrested cannot afford a bail bond and remain in jail awaiting a hearing. Though presumed innocent, they lose their jobs and families, and- are more likely to re-offend. And if all this weren’t bad enough, taxpayers must shoulder the cost—a staggering $1 billion per year.” The Honorable Nathan L, Hecht, Chief Justice of the Texas Supreme Court, Remarks Delivered to the 85th Texas Legislature, Feb. 1, 2017.

This case requires the court to decide the constitutionality of a bail system that *1058detains 40 percent of all those arrested only on misdemeanor charges, many of whom are indigent and cannot pay the amount needed for release on secured money bail. These indigent arrestees are otherwise eligible for pretrial release* yet they are detained for days or weeks until their cases are resolved, creating the problems that Chief Justice Hecht identified. The question addressed in this Memorandum and Opinion is narrow: whether the plaintiffs'have met their burden of showing a likelihood of success on the merits of their claims and the other factors necessary for a preliminary injunction against Harris County’s policies and practices of imposing secured money bail on indigent misdemeanor defendants. Maranda- Lynn ODonnell, Robert Ryan Ford, and Loetha McGruder sued while detained in the Harris County Jail on misdemeanor charges. They allege that they were detained because they were too poor to pay the amount needed for release on the secured money bail imposed by the County’s policies and practices. (Docket Entry Nos, 3, 41, 54). They ask this court to certify a Rule 23(b)(2) class and preliminarily enjoin Harris County, the Harris County Sheriff, and—to the extent they are State enforcement officers or County policymakers—the Harris County Criminal Court at Law Judges, from maintaining a “wealth-based post-arrest , detention scheme.” (Docket Entry No. 143 at 2). .

This case is difficult and complex. The Harris County Jail is the third largest jail in the United States. Pls. Ex. 12(aa) at 1. Although misdemeanor arrestees awaiting trial make up' about 5.5 percent of the Harris County Jail population on any given day, see id. at 13, about 50,000 people are arrested in Harris County, on Class A and Class B misdemeanor charges each year. Pis. Ex. 10(c), SOW Pretrial Services Annual Reportat 8.1 The arrests are made by a number of law-enforcement agencies, including the Houston Police Department and the police forces of smaller municipalities, the Texas Department of Public Safety, and the Harris County Sheriffs Office. Id. Harris County’s bail system is regulated by State law, local municipal- codes, informal, rules, unwritten customary practices, and the actions of judges in particular cases. The legal issues implicate intertwined Supreme Court and Fifth Circuit precedents on the level of judicial scrutiny in equal protection and due process cases and on the tailoring of sufficient means to legitimate ends.

Bail has a longstanding presence in the Anglo-American common law tradition. Despite this pedigree, the modern bail-bond industry and the mass incarceration on which it thrives present important questions that must be examined against current law and recent developments. Extrajudicial reforms ' have caused a sea change in American bail practices within the last few years. Harris County is also in the midst of commendable and. important efforts to reform its bail- system for misdemeanor arrests. The reform effort follows similar work in other cities and counties around the country. This work is informed *1059by recent empirical data about the effects of secured money bail on a misdemeanor defendant’s likely appearance at hearings and other law-abiding conduct before trial, as well as the harmful effects on the defendant’s life.

The plaintiffs contend that certainly before, and even with, the implemented reforms, Harris County’s bail system for misdemeanor arrests will continue to violate the Constitution, This case is one of many similar cases recently filed around the country challenging long-established bail practices. Most have settled because the parties have agreed to significant reform. This case is one of the first, although not the only one, that requires a court to examine in detail the constitutionality of a specific bail system for misdemeanor ar-restees. This case is also one of the most thoroughly and skillfully presented by able counsel on all sides, giving the court the best information available to decide these difficult issues.

One other complication is worth noting at the outset. Since this case was filed, the 2016 election replaced , the Harris County Sheriff and the presiding County Judge of Criminal Court at Law No. 16. (Docket Entry Nos. 158, 168). The new Sheriff and County Judge have taken positions adverse to their codefendants, although each continues to oppose certain aspects of the plaintiffs’ request for preliminary injunc-tive relief.2 Nonparty County officials, including the newly elected Harris County District Attorney and one of the Harris County Commissioners, have filed amicus briefs supporting the plaintiffs. (Docket Entry Nos. 206, 272). Harris County’s Chief Public Defender has filed a declaration supporting the defendants. Def. Ex. 23. The lines of affinity and adversity between'the defendants and their nonparty County colleagues are not always clear.

Even with the factual and legal complexities, at the heart of this ease are two straightforward questions: Can a jurisdiction impose secured money bail on misdemeanor arrestees who cannot pay it, who would otherwise be released, effectively ordering their pretrial detention? If so, what do due process and equal protection require for that to be lawful? Based on the extensive record and briefing, the fact and expert witness testimony, the arguments of able counsel, and the applicable legal standards, the answers are that, under federal and state law, secured money bail may serve to detain indigent misdemeanor arrestees only in the narrowest' of cases, and only when, in those c’ases, due process safeguards the rights of the indigent accused.

Because Harris County does not currently supply those safeguards or protect those rights, the court will grant the- plaintiffs’ motion for preliminary injunctive relief. The reasons and the precise, limited relief granted are set out in detail below.

More specifically, the court finds that;

• Harris County has a consistent and systematic policy and practice of imposing secured money bail as de facto orders of pretrial detention in misdemeanor cases.

*1060• These de facto detention orders effectively operate only against the indigent, who would be released if they could pay at least a bondsman’s premium, but who cannot. Those who can pay are released, even if they present similar risks of nonappearance or of new arrests.

• These de facto detention orders are not accompanied by the protections federal due process requires for pretrial detention orders.

• Harris County has an inadequate basis to conclude that releasing misdemeanor defendants on secured financial conditions is more effective to assure a defendant’s appearance or law-abiding behavior before trial than release on unsecured or nonfinancial conditions, or that secured financial conditions of release are reasonably necessary to assure a defendant’s appearance or to deter new criminal activity before trial.

• Harris County’s policy and practice violates the Equal Protection and Due Process Clauses of the United States Constitution.

The court accordingly orders that:

• Harris County and its policymakers— the County Judges in their legislative and rulemaking capacity and the Harris County Sheriff in his law-enforcement capacity—are enjoined from detaining misdemeanor defendants who are otherwise eligible for release but cannot pay a secured financial condition of release.

• Harris County Pretrial Services must verify a misdemeanor arrestee’s inability to pay bail on a secured basis by affidavit.

• The Harris County Sheriff must release on unsecured bail those misdemeanor defendants whose inability to pay is shown by affidavit, who would be released on secured bail if they could pay, and who have not been released after a probable cause hearing held within 24 hours after arrest.

The court does not order: relief in cases involving felony charges or a mix of misdemeanor and felony charges; the elimination of secured money bail; changes to Texas State law; changes to the written Harris County Criminal Courts at Law Rules of Court; modification of prior federal court orders, including the consent decree in Roberson v. Richardson; or a right to “affordable bail” under the Eighth Amendment. Instead, the relief ordered is consistent with Texas state and Harris County law as written, is required by the Equal Protection and Due Process Clauses, and is justified by the plaintiffs’ evidence. The relief is narrow so as not to interfere with the improvements the County is working to implement by July 1, 2017.

The reasons for these rulings are set out in the detailed findings and conclusions below.

I. Findings of Fact

A. Procedural Background

Ms. ODonnell filed suit while she was in custody in the Harris County Jail on May 19, 2016. (Docket Entry No. 3). Ms. McGruder and Mr. Ford filed suit while they were in custody on May 21, 2016. Civil No. 16-1436. The court consolidated the actions in August 2016. (Docket Entry No. 41). The plaintiffs filed an amended complaint on September 1, 2016. (Docket Entry No. 54). After extensive briefing and two lengthy hearings on August 18 and November 28, 2016, the court issued a Memorandum and Opinion on the defendants’ motions to dismiss. (Docket Entry No. 125); ODonnell v. Harris Cty., Tex., 227 F.Supp.3d 706, 2016 WL 7337549 (S.D. Tex. Dec. 16, 2016). The court dismissed *1061the claims against the Harris County Sheriff and the sixteen Harris County Criminal Court at Law Judges in their personal capacities. The court denied the motions to dismiss the claims against the County, the personal-capacity claims against five Harris County Hearing Officers, and the official-capacity claims against the Sheriff and the County Judges. (Id.). The court reset the preliminary injunction hearing scheduled for December 15, 2016 at the parties’ request, to facilitate settlement negotiations between the parties and newly elected Harris County officials. (Docket Entry No. 109). The parties did not settle. The court held an eight-day hearing in March 2017, and the parties filed voluminous records, lengthy video recordings, and numerous briefs.

The pending motions are the plaintiffs’ motion for class certification, (Docket Entry No. 146), the defendants’ motion for summary judgment, (Docket Entry Nos. 101, 104, 108), the plaintiffs’ motion for a preliminary injunction, (Docket Entry No. 143), and the defendants’ contingent motion for a stay pending appeal should the court grant preliminary injunctive relief, (Docket Entry No. 252). The defendants argue, principally, that there is no constitutional right to “affordable bad,” that Harris County’s post-arrest policies are subject to rational basis review, and that Harris County’s policies are constitutional under any level of judicial scrutiny. (See Docket Entry Nos. 101, 161, 162, 166, 193, 256, 286). The plaintiffs argue that Harris County’s system of pretrial bail and detention in misdemeanor cases violates the Equal Protection and Due Process Clauses of the United States Constitution. (See Docket Entry Nos. 143, 145, 188, 189). They do not believe their claims raise an Eighth Amendment challenge, but they argue in the alternative that the County’s bail system for misdemeanor arrestees fails under the Eighth Amendment as well. (Docket Entry No. 92 at 18 n.19; No. 188 at 14 n.13).

This Memorandum and Opinion addresses the parties’ disputes on summary judgment and the plaintiffs’ entitlement to preliminary injunctive relief. Separate orders address class certification and the defendants’ motion to stay.

B. The Evidence in the Record

The motion for a preliminary injunction requires balancing the- expediency demanded by the request for emergency relief with a full and fair consideration of the voluminous record. The parties submitted nearly 300 written exhibits, in addition to 2,300 video recordings of bail-setting hearings conducted within the last year in Harris County, all admitted without objection. (Docket Entry Nos. 244, 267). Thirteen witnesses testified at the eight-day hearing, including four expert witnesses. The court admitted depositions and declarations from many other witnesses as well.

The parties largely agree on the facts of the procedures Harris County follows after the arrest of. misdemeanor defendants. Both parties’ statistical experts used the same data from the County’s administrative sources and largely agreed on the raw numbers produced by, and the gaps found in, the Harris County data. The parties’ experts disagree about how to interpret the data. The parties disagree about the constitutional significance of the evidence about the County’s bail procedures in misdemeanor cases and their effects.

The court reviews the factual record under the applicable legal framework to resolve these disagreements and to enter the findings of fact and conclusions of law.3

*10621. The Parties

Maranda Lynn ODonnell, a 22-year-old single mother, was arrested on May 18, 2016 at 5:00 p.m. and charged with driving with'an invalid license. Pis. Ex. 7(a). After she was booked into- the Harris County Jail, she was informed that she would be released promptly if she paid a secured money bail of $2,500 set according to the County’s bail schedule, but that she would remain in jail if she did not pay either the full bail amount to the County or a premium to a bail bondsman up front. Id. Ms, ODonnell and her child struggled to meet the basic. necessities of life. She received benefits from the federal government’s Women, Infants, and Children program to feed her daughter. She could not afford housing, so she stayed with a friend. Id. At the time of her arrest, Ms. ODonnell was working, but it was at a new job she had held for only seven days. Id. She had no money to buy her release from detention. Id. She was otherwise eligible for release.

Harris County Pretrial Services interviewed Ms. ODonnell at 11:52 p.m. on May 18. Pls. Ex. 8(c)(1), ODonnell Pretrial Services Report. At 3:00 a.m., on May 19, Pretrial Services completed a risk-assessment report recommending her release on a personal bond—that is, an unsecured appearance bond requiring no up-front payment for release. Id. Ms. ODonnell appeared before a Hearing Officer at 7:00 a.m., by videolink from the Harris County Jail, Pls. Ex. 4(c)(1), ODonnell Docket Sheet. The Sheriffs deputies present ordered her not to speak. Pis. Ex. 7(a). Without explanation,' the Hearing Officer told her that she did not “qualify” for release on personal bond and imposed the $2,500 scheduled amount as secured bail, meaning that she had to pay the full bail amount or a bondman’s premium to be released. Pls. Ex. 8(c), ODonnell Hearing Video. When asked if she would hire her own lawyer or would be seeking help from a court-appointed lawyer, Ms. ODonnell responded, “Seeking help.” These were her only words during her 50-second hearing. Id.

On the morning of May 20, Ms. ODonnell appeared before a County Criminal Court at Law Judge. (Docket Entry No. 31, Ex. 1). She completed an .affidavit declaring her lack of assets and was found indigent for the. purpose of appointing counsel- (Id.). Her bail amount was not changed or set on an unsecured basis, even though she declared on her affidavit that she remained in jail. (Id.). That, same day, but after Ms. ODonnell filed this suit, an insurance underwriter for a commercial bondsman posted her bail amount. Pls. Ex. 11 at *5. This third-party payment looks like an attempt to moot her claim. See id. Ms. ODonnell was released from jail after three days in pretrial detention- on the charge of driving with an invalid license. Pls. Ex. 8(c), ODonnell Docket Sheet.

Robert Ryan Ford.was arrested on May 18, 2016 at 8:00 p.m. He was charged with shoplifting from a Wal-Mart. Pls. Ex. 7(c). Mr. Ford could not pay the $5,000 secured money bail imposed as the condition for his release from pretrial detention. Id. This was the amount specified in the bail schedule. Mr. Ford was interviewed by Pretrial Services at 10:00 a.m. the morning after his arrest, but Pretrial Services did not complete Mr. Ford’s risk assessment until the next day, May 20, at 2:00 a.m. Pls. Ex. 8(c)(iii), Ford Pretrial Services Report. The risk-assessment report recommended “Detain,” stating that Mr. Ford had “[sjafety issues that conditions can’t mitigate.” Id. at *16. The form did not explain these issues nor why some combination of conditions of release could not address them. Notwithstanding the recommenda*1063tion to detain, had Mr. Ford paid the $5,000 bail—or paid a bondsman a $500 premium4—he would have been promptly released, regardless of “safety issues.” He could not pay the $5,000 secured money bail or the bondsman’s premium, so he remained in jail. Pls. Ex. 7(c). As intended by Pretrial Services, the secured money bail served as a pretrial detention order because Mr, Ford was too poor to pay.

Mr. Ford did not see a Hearing Officer until May 20, 2016 at 4:00 a.m., 32 hours after his arrest. Pis. Ex. 8(c)(iii), Ford Docket Sheet. His hearing lasted less than 50 seconds. Pls. Ex. 8(c), Ford Hearing Video. He did not speak except to ask for a court-appointed lawyer. Id. His bail was confirmed at $5,000 on a secured basis. Id.

On May 23, 2016,- Mr. Ford appeared before a County Criminal Court at Law Judge, pleaded guilty, and was sentenced to time served. Pls. Ex. 8(c)(iii), Ford Docket Sheet. He was released at 12:30 a.m. on May 24, 2016. Id. Mr. Ford was continuously detained on his misdemeanor charge for over five days, until the final disposition of his case.

Loetha Shanta McGruder, a pregnant 22-year-old mother of two, was arrested on May 19, 2016 at 5:20 p.m. She was charged with failing to identify herself to a police officer, Pls. Ex. 7(b). Ms. McGruder was indigent. Id. She depended on federal benefits to care for her older son, who has Down’s Syndrome and other medical needs, and she depended on child-support payments for her other children. Id. Ms. McGruder was not working when she was arrested. She avoided homelessness by living with her boyfriend. Id. She could not pay the $5,000 secured money bail imposed as the condition for her release from pretrial detention. Id.

Ms. McGruder was interviewed by Pretrial Services the morning after her arrest, at 8:40 a.m. Pls. Ex. 8(c)(ii), McGruder Pretrial Services Report. Pretrial Services completed its risk-assessment report around 1:00 p.m. With no recommendation for either release or detention. Id. Ms. McGruder appeared-before a Hearing Officer at 1:00 p.m. on May 20. Pls. Ex. 8(c)(ii), McGruder Docket Sheet - She did not speak at her hearing except to discuss her need for a court-appointed lawyer. Pls. Ex. 8(c)(ii), McGruder Hearing Video. Her bail was confirmed at $5,000 on a secured basis. Id.

After about 87 hours in jail, Ms. McGru-der appeared before a County Criminal Court at Law Judge. Pls. Ex. 8(e)(ii), McGruder Docket Sheet. She was ready to enter a guilty plea because she bélieved it was the fastest way to be released. Hearing Tr. 2-1:80-81, 108. Her lawyer convinced her to seek’a personal bond instead. Id. At her first counseled hearing before a County Judge, Ms. McGruder was granted a personal bond—an unsecured $5,000 bond with no up-front payment required. She was released at 7:30 p.m. the same day. Pls. Ex. 8(c)(ii), McGruder Docket Sheet. Ms. McGruder spent four full days in pretrial detention on her misdemeanor charge of failing to identify herself to a police officer.

The plaintiffs sued Harris County under 42 U.S.C. § 1983, alleging that the County’s policies have deprived them and others similarly situated of due process and equal protection by detaining them 'before trial on misdemeanor charges because of their *1064inability to pay a secured money bail, and without a meaningful or timely inquiry into their inability to pay. (Docket Entry No. 54). The motions to dismiss resulted in earlier rulings on the claims against the various defendants.

• The court denied Harris County’s motion to dismiss. The County may face municipal liability under § 1983 for the law-enforcement policies of its Sheriff, to the extent the Sheriff knowingly enforces invalid detention orders, and for the legislative and administrative policies of the County Judges to the extent those policies are not directly mandated by Texas law. ODonnell, 227 F.Supp.3d at 738-50, 2016 WL 7337549 at *22-31.

• The court dismissed the plaintiffs’ personal-capacity claim against the Harris County Sheriff but denied the motion to dismiss the official-capacity claim. Id. at 751-52, 2016 WL 7337549 at *32. To the extent the Sheriff enforces facially valid but unconstitutional detention orders, the Sheriff may be liable for prospective relief under Ex parte Young, 209 U.S. 123, 28 S.Ct. 441, 52 L.Ed. 714 (1908). Id.

• The court dismissed personal-capacity claims against the sixteen Harris County Criminal Court at Law Judges, but denied the motion to dismiss the official-capacity claims against them. ODonnell, 227 F.Supp.3d at 745-47, 2016 WL 7337549 at *27-28. To the extent the County Judges administratively enforce facially constitutional Texas laws, such as the Texas Code of Criminal Procedure, in an unconstitutional manner, the County Judges may be liable for prospective relief. Id. at 746—47, 756-58, 2016 WL 7337549 at *28, 36-37.

• The court granted the motion to dismiss the official-capacity claims against five Harris County Hearing Officers. Id. at 753-55, 2016 WL 7337549 at *34-35. They remain in the suit in their personal capacities for declaratory relief only. Id.

2. The Fact Witnesses

The fact witnesses testified about the post-arrest process for misdemeanor defendants in Harris County, as well as the reforms to the bail system the County expects to implement by July 1, 2017. The fact witnesses and their testimony are summarized below.

• Assistant District Attorney JoAnne Musick. Ms. Musick was appointed the Sex Crime Unit Chief at the Harris County District Attorney’s Office in January 2017. She has practiced criminal defense privately for over thirteen years and has served as the vice-chair of the Criminal Law & Procedure Committee of the Houston Bar Association and as a board member of the Texas Criminal Defense Lawyers Association. Ms. Musick testified about her extensive experience with Harris County pretrial processes, both as a criminal defense lawyer and as an Assistant District Attorney. Ms. Musick filed a declaration stating her observation that Harris County consistently detains misdemeanor arrestees, who are otherwise eligible to be released, because they cannot pay a secured financial condition of release. As a consequence, many indigent misdemeanor arrestees plead guilty at their first appearance as the only way to be released from pretrial detention without waiting days or weeks for another hearing. Pls. Ex. 7(g) at 4-5.

• Sheriff Ed Gonzalez. Sheriff Gonzalez was elected Harris County Sheriff in November 2016 and assumed office in January 2017. He served eighteen *1065years with the Houston Police Department and was a Houston City Council member for three terms before his election as Sheriff. Sheriff Gonzalez testified about his experience with the post-arrest process in Harris County. Sheriff Gonzalez also filed a declaration stating his observation that Harris County consistently detains misdemeanor arrestees, who are otherwise eligible to be released, because they are too poor to pay a secured financial condition of release. Pls.' Ex. 7(r) at 1-2.

• Major Patrick Dougherty. Major Dougherty was appointed as a major with the Harris County Sheriffs Office in January 2017 after serving thirty-five years with the Houston Police Department. Major Dougherty testified about his experiences with the post-arrest processes in the City of Houston and in Harris County. Major Dougherty reviewed the technology limits and overcrowded conditions in the Harris County Jail that complicate the timely transfer and presentment of misdemeanor arrestees.

• Director of Pretrial Services Kelvin Banks. Mr. Banks began work as the Director of Pretrial Services for Harris County in October 2016. He served previously as the Director of Pretrial Services for the Third Circuit Court in Wayne County, Michigan, primarily overseeing pretrial services for the City of Detroit. Mr. Banks testified about the County’s current Pretrial Services program, the planned changes to Pretrial Services’s risk-assessment tool, and other changes impacting the use of secured money bail in misdemeanor cases. These changes are expected to be implemented by July 1, 2017.

• Chief Hearing Officer Blanca Villago-mez. Judge Villagomez has been a Harris County Hearing Officer since the position was created in 1993. She testified about her own and others’ practices as Hearing Officers.

• Hearing Officer Eric Hagstette. Judge Hagstette has been a Harris County Hearing Officer for over eleven years. He was a Harris County Assistant District Attorney for ten years and a criminal defense attorney for ten years. Judge Hagstette testified about his practices as a Hearing Officer and his impressions of the pretrial process from his time as a practicing criminal lawyer.

• County Judge Darrell Jordan. Judge Jordan was elected to be the presiding judge of County Criminal Court at Law No. 16 in November 2016. He assumed office in January 2017. Judge Jordan previously practiced as a criminal defense attorney for eight years. Judge Jordan testified about his practices as a County Judge and about his past experiences as a lawyer defending misdemeanor arrestees in Harris County.

• County Judge Paula Goodhart. Judge Goodhart was appointed to be the presiding judge of County Criminal Court at Law No. 1 in 2010. She was an Assistant District Attorney for Harris County for fourteen years and a criminal defense attorney for three years. Judge Goodhart testified about her practices as a County Judge and about her past experiences practicing in the Harris County Criminal Courts at Law.

• County Judge Margaret Harris. The defendants offered the testimony of Judge Harris, the presiding judge of County Criminal Court at Law No. 5 since 2003. The parties stipulated that Judge Harris’s testimony would be consistent in material respects with *1066Judge Goodhart’s testimony. Hearing Tr. 5:152.

• Dr. Marie VanNostrand. Dr. VanNost-rand is a project manager for Luminosity, a consulting firm that advises pretrial services programs. Dr. Van-Nostrand is a former probation and parole officer and pretrial services provider, She began working as a consultant for pretrial services agencies in 2003 and through Luminosity has been consulting with Harris County to reform its pretrial processes and services since February 2015. Hearing Tr. 6-1:131. Dr. VanNostrand testified about her statistical studies on pretrial detention and about the reforms to the Harris County pretrial process planned for implementation by July 1, 2017.

3. The Expert Witnesses

The plaintiffs presented Dr. Stephen Demuth to testify under Rule 702 of the Federal Rules of Evidence on sociology and criminal pretrial procedure. Dr. De-muth has a doctorate in sociology with a concentration in chronology and quantitative methods of research. He is a professor of Sociology at Bowling Green State University in Ohio. He has published extensively in peer-reviewed journals on pretrial criminal processes and on the appropriate use of large data sets. Dr. Demuth testified that he received no compensation for his consultation and testimony in this case. He has invested at least 150 hours of work analyzing the data Harris County has produced since the plaintiffs retained him on February 9, 2017.

The plaintiffs also presented Judge Truman Morrison to testify under Rule 702. Judge Morrison is a Senior Judge of 'the Superior Court of the District of Columbia. He has served on that court for over thirty-seven years. After taking senior status in 2000, Judge Mbrrison has focused on misdemeanor cases. Since the late 1980s, he has led reform efforts in his court to eliminate the use of secured money bail in the D.O. criminal justice system. He has also worked to educate judicial officers and others around the country on the benefits of eliminating money bail and the harms of continuing to use it in misdemeanor cases.

The defendants offered the testimony of Dr. Robert Morris as a Rule 702 witness in criminology. Dr. Morris holds a doctorate in criminal justice and was a professor of criminology at the University of Texas in Dallas for nine years. Since August 2016, he has been the cofounder and chief executive officer of Predicto, a company that uses machine learning to predict failures in industrial equipment. Dr. Morris testified that he has worked 45 to 50 hours analyzing data produced by Harris County since his retention and has invoiced the County $325 per hour. Hearing Tr. 4-2:156.

The defendants-also offered the testimony- of Mr. Bob Wessels as a Rule 702 witness with specialized knowledge in court administration and pretrial procedures, particularly in Harris County..Mr. Wessels was the court manager of the Harris County Criminal Courts at Law for thirty-five years, until he retired in 2011. He has received numerous awards and national recognition for his' work on court administration and is a former president of the National Association of Court Administrators.

The court finds that'Drs. Demuth and Morris meet the Rule 702 requirements to testify about Harris County’s pretrial arrest data and system and that Judge Morrison and Mr. Wessels are qualified to testify about court administration. Specific findings about the reliability, helpfulness, and credibility of their opinions are set out in detail below.

*10674. Overview of the Factual . and Legal Issues

The parties dispute three broad issues: (1) whether Harris County impermissibly sets secured money bail to serve as de facto orders of pretrial detention in misdemeanor cases; (2) whether Harris County provides misdemeanor defendants due process and equal protection in their bail settings; and (3) whether planned reforms will sufficiently address the plaintiffs’ allegations of' constitutional violations. Each issue raises complex questions of fact‘and law.

The defendants argue that Harris County judicial officers do not intentionally use secured money bail to detain and are not recklessly indifferent to that effect of secured money bail. Instead, the defendants argue, Hearing Officers and County Judges apply the Texas Code of Criminal Procedure’s requirement to consider five factors—only one of which relates to a defendant's ability to pay—in setting bail. See Tex. Code Crim. Pro. art. 17.15. The plaintiffs respond that the evidence shows Harris County judicial officers do not in fact give individualized consideration of the five factors in setting bail in each misdemeanor case, but instead routinely set secured money bail to conform to a predetermined schedule, even when it is clear that the effect will be pretrial detention.

The parties’ disputes • extend beyond whether the facts show rare, occasional, or frequent individual consideration of bail in particular cases. The parties also dispute whether imposing secured money bail on an indigent or impecunious misdemeanor arrestee is a but-for cause or a proximate cause of pretrial detention if an arrestee with financial means could pay and secure prompt release. The defendants argue that virtually no misdemeanor defendant is detained before trial “solely by” or “because of’ an inability to pay secured money bail. Instead, the defendant’s past criminal history, prior failures to appear, or other risk factors all contribute to a judicial officer’s decision to impose secured money bail at a particular amount. Under this view, the arrestee’s criminal history, prior failures to appear, or other risk factors—not just the bail amount—are among the reasons for pretrial detention. (See, e.g., Docket Entry No. 162 at 15-16; No. 164 at 8-9); Hearing Tr. 1:99-100.

The plaintiffs counter with a but-for argument. A judicial officer’s decision to set secured money -bail means that the misdemeanor defendant has.been found eligible for release and would be released but for their inability to make the up-front payment of the secured money bail bond. The plaintiffs argue that detaining misdemean- or defendants before trial solely because of their inability to pay violates the Equal Protection Clause, because defendants with similar histories and risks but with access to money are able to purchase pretrial release. The plaintiffs contend that all rigorous studies of pretrial release in misdemeanor cases show that release on secured money bail does no more to mitigate the risk of nonappearance or of new criminal activity during pretrial release than release on, unsecured or nonfinancial conditions. (See, e.g., Docket Entry No. 143 at 15-17; No. 188 at 4-7); Hearing Tr. 4-2:15-16.

For' the reasons set out below, the court finds and concludes that the plaintiffs have the better understanding of the case. A misdemeanor defendant’s criminal background or risk factors may give the County a persuasive reason to detain that defendant. But an order imposing secured money bail is effectively a pretrial preventive detention order only against those who cannot afford to pay. It is not a detention order as to defendants who can pay, even* if they present a similar risk of fail*1068ing to appear or of committing new offenses before trial as those who cannot pay. And the reliable record evidence shows that release on secured money bail does not mitigate those risks for misdemeanor defendants better than release on unsecured or nonfinancial conditions, in Harris County or elsewhere. The issue is not a right to “affordable bail,” as the defendants insist, but a violation of the Equal Protection and Due Process Clauses.

The plaintiffs allege that Harris County’s pretrial misdemeanor bail system violates procedural due process because: (1) misdemeanor arrestees who cannot pay the up-front amount for release on secured money bail are frequently held longer than 24 hours before any meaningful bail review, contrary to Texas law and to federal court orders; (2) misdemeanor arrestees are not able even to ask for a review of their bail in a counseled, adversarial proceeding, with an opportunity to present evidence and a right to findings on the record, until at least two or three days and often up to two weeks after their arrests; (3) misdemeanor arrestees who cannot afford their secured money bail are jailed for more than 48 hours if they do not plead guilty at their first court appearances; and (4) the County imposes secured money bail to serve as de facto detention orders, without affording misdemeanor defendants the due process protections the Constitution requires for detention orders. (See, e.g., Docket Entry No. 144 at 13-14; No. 145 at 7-9; No. 188 at 13-15).

The defendants argue that Harris County’s pretrial process is among the fastest in the nation and that its bail practices are not out of step with the majority of other United States jurisdictions. (See, e.g., Docket Entry No. 286). While insisting that the County’s current system is legal, the defendants acknowledge that it needs improvement. They argue that the planned reforms will make the County’s pretrial system more efficient, more rational, and more equal across classifications of wealth and risk factors. (Docket Entry No. 162 at 23); Hearing Tr. 1:100-01, 8-2:30-32. The plaintiffs respond that until the expected reforms are implemented, serious constitutional violations will continue to occur, affecting hundreds of individuals every day. Even under the reforms, the plaintiffs contend, Harris County will continue its policy of imposing secured money bail as de facto pretrial detention orders in violation of federal due process requirements. (Docket Entry No. 188 at 26-27).

The court finds and concludes that, based on the credible, reliable evidence in the present record, the plaintiffs are likely to succeed on the merits of at least some of their claims that the present system violates due process and equal protection, and that the plaintiffs are likely to succeed in part in their challenges to the new pretrial system as currently proposed by Harris County. The plaintiffs are entitled to a preliminary injunction, as set out in detail below.

C. The Historical Development of Bail in the United States and in Harris County

1. The Constitutionalization of Bail

Bail originated in medieval England “as a device to free untried prisoners.” Daniel J. Freed & Patricia M. Wald, Bail in the United States: 1964 1 (1964); see 4 William Blackstone, Commentaries on the Laws of England (Rees Welsh & Co. [1769] 1902) (“By the ancient common law, before and since the [Norman] conquest, all felonies were bailable, till murder was excepted by statute; so that persons might be admitted to bail before conviction almost in every case.” (footnotes omitted)); see generally William F. Duker, The Right to Bail: A *1069 Historical Inquiry, 42 Alb. L. Rev. 33, 34-66 (1977); Elsa de Haas, Antiquities of Bail 128 (1940). In 1276, the English Parliament enacted the Statute of Westminster, which defined bailable offenses and provided criteria for determining, whether a particular person should be released, including the strength of the evidence against the accused and. the accused’s criminal history. See Note, Bail: An Ancient Practice Reexamined, 70 Yale L.J. 966, 966 (1961); June Carbone, Seeing Through the Emperor’s New Clothes: Rediscovery of Basic Principles in the Administration of Bail, 34 Syracuse L. Rev. 617, 523-26 (1983). In 1679, Parliament adopted the Habeas Corpus Act to ensure that an accused could obtain a timely bail hearing. In 1689, Parliament enacted an English Bill of Rights that prohibited excessive bail. See Carbone, supra, at 528.

Early American constitutions codified a right to bail as a presumption that defendants should be released pending trial. See Note, Bail, supra, at 967. One commentator who surveyed the bail laws in each state found that forty-eight states have protected, by constitution or statute, a right to bail, “by sufficient sureties, except for capital offenses when the proof is evident or the presumption great.” Matthew J. Hegreness, America’s Fundamental and Vanishing Right to Bail, 55 Ariz. L. Rev. 909, 916 (2013). States modeled these provisions on the Pennsylvania Constitution of 1682. See Carbone, supra, at 531-32. Texas substantially incorporated that language into its Constitution in 1845, and it remains.5 Tex. Const. art. 1 § 11.

Texas law interprets Article I, § 11 to prohibit preventive pretrial detention except in specific and narrow circumstances set out in constitutional amendments. Id. § 11a et seq. “The exceptions contained in Article I, § 11a, supra, to the constitutional right to bail proclaimed by Article I, § 11, supra, include the seeds of preventive detention urged by many to be abhorrent to the American system of justice. It is obvious that for these reasons the provisions of said § 11a contain strict limitations and other safeguards.” Ex parte Davis, 574 S.W.2d 166, 169 (Tex. Cr. App. 1978). The exceptions are narrow. All but one are limited to felonies. The one exception is under §§ lib and 11c, which permit a denial of bail and pretrial preventive detention for those accused of a crime of family violence, including misdemeanors, if: (1) the accused has violated a condition of pretrial release or a protective order; and (2) a magistrate determines at an adversary hearing by a preponderance of the evidence that the accused violated the condition of release or protective order in a manner “related to the safety of a victim of the alleged offense or the safety of the community.” Tex. Const. art. 1, §§ 11b-11c.

Historians and jurists confirm that from the medieval period until the early American republic, a bail bond was typically based on an individualized assessment of what the arrestee or his surety could pay to assure appearance and secure release. In medieval England, an arrestee was forbidden to pay his sureties for obtaining his release. If an accused failed to appear, the sureties were' “amerced” with a fine, but there were “maximum amercements depending on the wrong-doer’s rank; the baron [did] not have to pay more than a hundred pounds, nor the routier more than five shillings.” 2 Frederick William Polluck & Frederic William Maitland, The History of English Law Before the Time of Edward I 514 (2d ed. 1984 [1898]). Joseph Chitty, an eminent proceduralist, summa*1070rized the English practice when the United States Constitution was-- ratified: “The rule is, where the offence is - prima facie great, to require good bail; moderation nevertheless is to be observed, and such bail only is to be required as the party is able to procure; for otherwise the allowance of bail would be a mere colour for imprisoning the party on the charge.” 1 J. Chitty, A Practical Treatise on the Criminal Law 88-89 (Philadelphia ed. 1819); see also Bates v. Pilling, 149 Eng. Rep. 805, 805 (K.B. 1834) (“a defendant might be subjected to as much inconvenience by being.compelled to put in bail to an excessive amount, as if he had been actually arrested”); Rex v. Bowes, 99 Eng. Rep. 1327, 1329 (K.B. 1787) (per curiam) (“[ejxcessive bail is a relative term; it depends on the nature of the charge for which bail is required, Upon the- situation in life of the parties, and on various other circumstances”) & (Archbald, J.) (permitting a “lessening” of bail if there were “difficulty” procuring the decreed sum); Neal v. Spencer, 88 Eng. Rep. 1305, 1305-06 (K.B. 1698) (collecting cases showing a diversity of bail amounts given for the same .offense). The pre-Texas history of bail confirms the modern holdings of Texas courts, that bail is a mechanism for pretrial release and not. for continued pretrial.preventive detention.

2. Statutory and Judicial Bail Reform: Pretrial Services, Probable Cause Hearings, and “Meaningful” Alternatives to Secured Money Bail

In the mid-nineteenth century, bail reform was crucial to abolishing imprisonment for. debt. In Massachusetts, the 1831 survey of the Prison Discipline Society noted that the availability of bail in debtors’ prison's created class distinctions between- “poor seamen, poor laborers, and poor mechanics” who could not find sureties and remained -in jail, “while there is scarcely an instance on record of a poor minister, a poor physician, or a poor lawyer in Prison for debt.” Sixth Annual Rep. of the Prison Discipline Society 22 (1831). After Massachusetts abolished imprisonment for debt in 1855, the State permitted those jailed on mesne6 process in contract cases to swear an oath of indigence and to be released on personal recognizance as an alternative to secured money bail. 1857 Mass. L. 489-97. From 1831 to 1833; Congress passed legislation abolishing imprisonment for debt at the federal level. 4 Stat. 467, 594, 676. Ultimately, forty-one States, including Texas, constitutionally banned imprisonment for debt.7 See Tex. Const. art. 1 § 18.

Another wave of bail reform began with the 1960s Manhattan Bail Project, conducted by -the Vera Foundation in New York City. See Wayne H. Thomas, Jr., Bail Reform in America 3, 20-27 (1976); Ronald Goldfarb, Ransom 150-72 (1965). The Project interviewed- defendants before their first court appearance to evaluate whether they were good candidates for pretrial release -on recognizance; that is, release “on one’s honor pending trial.” Goldfarb, supra, at 153-54. The standard interview questions asked about a defendant’s personal background, community ties, and criminal history. Id. The’ interviewer scored a defendant’s answers using a *1071point-weighting system and verified the answers, usually by telephone, with references the defendant provided. Id. at 154-55, 174-75. The interviewers gave the information to the court and recommended which defendants should be released on nonfinancial conditions. Id. at 155. During the first three years of the Project, defendants released on nonfinancial conditions at the recommendation of the Vera Foundation were about three times more likely to appear for trial than were defendants in control groups who were found eligible for release on nonfinancial conditions but who were instead released on secured money bail. Id. at 155, 157. The success of the Manhattan Bail Project inspired the creation of pretrial services programs across the country. See Timothy R. Schnacke et al, Pretrial Justice Inst., The History of Bail and Pretrial Release 10 (2010).

In the 1970s, a major prisoners’ class action challenged the facial and as-applied constitutionality of Florida’s pretrial detention system. The litigation led to two foundational opinions, one by the United States Supreme Court and one by the former Fifth Circuit. In Gerstein v. Pugh, 420 U.S. 103, 95 S.Ct. 854, 43 L.Ed.2d 54 (1975), the Supreme Court, ruled that criminal defendants arrested without a warrant and then detained before trial had to be taken “promptly” before a judicial officer to determine probable cause for the arrest. 420 U.S. at 127, 95 S.Ct. 854. The Court did not specify what would meet the promptness standard, instead noting that “the nature of the probable cause determination usually will be shaped to accord with a State’s pretrial procedure viewed as a whole.... It may be found desirable, for example, to make the probable cause determination at the suspect’s first appear-anee before a judicial officer, or the determination may be incorporated into the procedure for setting bail or fixing other conditions of pretrial release.” Id. at 124, 95 S.Ct. 854 (internal citations omitted).

In Pugh v. Rainwater, 572 F.2d 1053 (1978) (en banc), the Fifth Circuit considered the same class’s challenge to Florida’s pretrial bail system. The en banc court vacated as moot the panel decision finding the system unconstitutional, because Florida had aménded its rules, while the appeal was pending. Id. at 1058-59. The en banc court ruled that the Constitution did not require the statute to include a presumption that indigent arrestees would be released without financial conditions to be facially valid. Id. at 1057-58. But the court noted that while “[ujtilization of a master bond schedule provides speedy and convenient release for those who have no difficulty in meeting its requirements!;, t]he incarceration of those who cannot, without meaningful consideration of other possible alternatives, infringes on both due process and equal protection requirements.” Id. at 1057.

In the decade following Gerstein and Rainwater, the City of Houston and Harris County were sued in two lawsuits disputing how to apply those precedents locally. In Sanders v. City of Houston, 543 F.Supp. 694 (S.D. Tex. 1982), the court ruled after a bench trial that Gerstein’s promptness standard required a probable cause hearing for those arrested without a warrant by the City of Houston within 24 hours of arrest. Id. at 702. The court also ruled that bail had to be set within 24 hours of arrest to avoid an unconstitutional denial of bail under the Texas Constitution.8 Id. at 704.

*1072In Roberson v. Richardson, Agreed Final Judgment, Civil No. 84-2974 (S.D. Tex. Nov. 25, 1987), the court entered a final agreed judgment that applied the 24-hour time limit to misdemeanor cases throughout the County.9 The Roberson order’s stated purpose was to ensure that misdemeanor arrestees in Harris County had “the right to a prompt, fair and reliable determination of Probable Cause as set out in Gerstein v. Pugh, 420 U.S. 103 [95 S.Ct. 854, 43 L.Ed.2d 54] (1978 [1975]), a meaningful review of alternatives to pre-scheduled bail amounts as set out in Rainwater v. Pugh [Pugh v. Rainwater ], 572 F.2d 1053 (5th Cir. 1978) (en banc), and the right to the prompt appointment of counsel.” Id. at 1. The Roberson order required the County Criminal Courts at Law Judges to provide probable cause hearings within 24 hours of misdemeanor arrests, allowing the hearings to be by videolink father than in person. Id. at 2 (videolink), 3 (24 hours).

The Roberson order required judicial officers at the probable cause hearing to “set the amount of bail required of the accused for release and determine the accused’s eligibility for release on personal bond or alternatives to prescheduled bail amounts.” Id. at 3. Substantially repeating Article 17.15 of the Texas Code of Criminal Procedure, Section D of the Roberson order stated:

Such bail determinations shall be according to the following criteria:

1. The bail shall be sufficiently high to give reasonable assurance that the undertaking will be complied with;

2. The nature of the offense for which Probable Cause has been found and the circumstances under which the offense was allegedly committed are to be considered, including both aggravating and mitigating factors for which there is reasonable ground to believe shown, if any;

3. The ability to make bail is to be regarded, and proof may be taken upon this point;

4. The future safety of the victim may be considered, and if this be a factor, release to a third person should also be considered; and

5. The Judicial Officer shall also consider the accused’s employment history, residency, family affiliations, prior criminal record, previous court appearance performance and any outstanding bonds.

Id. at 3.

The Roberson order required the County Judges to “implement and maintain a bond schedule for all misdemeanor offenses within their jurisdiction.” Id. at 4. The schedule had to “establish the initial amounts of bail required in each type or category of offense.” Id. The Roberson order required that:

At the Probable Cause hearing' the [Hearing] Officer shall use the Bail Schedule, in addition to the criteria in Section D, in determining the appropriate bail in a given case. The [Hearing] Officer shall have the authority to order the accused released on personal bond or released on other alternatives to pre-scheduled bail amounts. The [County] Judges shall direct the Pretrial Services Agency to make every effort to insure that sufficient information is available at the time of the hearings required herein for the [Hearing] Officer to determine an accused’s eligibility for a personal *1073bond or alternatives to prescheduled bail amounts.

Id.

Nothing in the Roberson order contemplated detention based on a misdemeanor arrestee’s inability to pay the scheduled bail amount set on a secured basis. Rather, the order required Hearing Officers to make individualized adjustments to the bail schedule in each case to provide a mechanism for release, either by lowering the scheduled amount when setting a secured bond; setting nonfínancial conditions of release; or granting release on unsecured “personal bonds” without additional conditions. See id. at 4, 1 (the purpose of the order is to provide “a meaningful review of alternatives to pre-scheduled bail amounts” (emphasis added)).

Finally, the Roberson order required the County Judges to appoint counsel “prior to any adversarial judicial proceedings” or “where the Judge concludes that the interests of justice require representation, for all accused indigents who do not refuse the appointment of counsel.” Id. at 4. In determining indigency for the purpose of appointing counsel, the Roberson order required the County Judges to consider the accused’s income and expenses, assets and debts, dependents, and “whether the accused -has posted or is capable of posting bail.” Id. In no case could a County Judge “deny appointed counsel to an accused solely because the accused has posted, or is capable of posting bail.” Id.

Efforts to comply with the Roberson order have produced the system Harris County has in place today, examined in greater detail below.10 (Docket Entry No. 101 at 11); Hearing Tr. 4-2:222-29; 5:6-20. Harris County operates a Pretrial Services Agency that interviews misdemeanor ar-restees to provide criminal risk and financial background information to the Hearing Officers. The Hearing Officers hold videolink hearings for those arrested, charged, and booked into the Harris County Jail on misdemeanor charges.11 These hearings usually, but far from always, are held within 24 hours of arrest. The Hearing Officers usually jointly determine probable cause and set bail.

3. Bail at the Federal Level

At the federal level, the Judiciary Act of 1789 provided an absolute right to bail in noncapital cases and bail at the judge’s discretion in capital cases. See 1 Stat. 73, 91. The first Congress also proposed the Eighth Amendment to the United States Constitution, which, like the Texas Constitution and the English Bill of Rights, prohibits excessive bail. See U.S. Const. amend. VIII; Tex. Const. art. 1, § 13. But, unlike the Texas Constitution, the United States Constitution does not explicitly state a right to bail. The Eighth Amendment guarantees only that “[ejxcessive bail shall not be required.” U.S. Const. amend. VIII; see Carlson v. Landon, 342 U.S. 524, 545-46, 72 S.Ct. 525, 96 L.Ed. 547, (1952) (the Eighth Amendment does not provide a “right to bail”). But the United States Supreme Court has made clear that “[b]ail set at a figure higher than an amount reasonably calculated to fulfill [the] purpose [of assuring the defendant’s appearance at trial] is ‘excessive’ under the Eighth Amendment.” Stack v. Boyle, 342 *1074U.S. 1, 5, 72 S.Ct. 1, 96 L.Ed. 3 (1951). As the Court explained,

From the passage of the Judiciary Act of 1789, 1 Stat. 73, 91, to the present Federal Rules of Criminal Procedure, Rule 46(a)(1), 18 U.S.C.A., federal law has unequivocally provided that a person arrested for a noncapital offense shall be admitted to bail.1 This traditional right to freedom before conviction permits the unhampered preparation of a defense, and serves to prevent the infliction of punishment prior to conviction. See Hudson v. Parker, 1895, 156 U.S. 277, 285, 15 S.Ct. 450, 39 L.Ed. 424. Unless this right to bail before trial is preserved, the presumption of innocence,secured only after centuries of struggle, would lose its meaning.

Id. at 4, 72 S.Ct. 1; see also Bandy v. United States, 81 S.Ct. 197, 198, 5 L.Ed.2d 218 (1960) (Douglas, J., in chambers) (“It would be unconstitutional to fix excessive bail to assure that a defendant will not’ gain his freedom. Yet in the case of an indigent defendant, the fixing of bail in even a modest amount may have the practical effect of denying him release." (citing Stack, 342 U.S. at 1, 72 S.Ct. 1)); United States v. Leathers, 412 F.2d 169, 171 (D.C. Cir. 1969) (“the setting of bond unreachable because of its amount would be tantamount to setting no conditions at all”).

The Bail Reform Act of 1966 became' “the first major reform of the federal bail system since the Judiciary Act of 1789.” State v. Brown, 338 P.3d 1276, 1286 (N.M. 2014); see Bail Reform Act of 1966, 80 Stat. 214 (repealed 1984). The stated purpose of the Bail Reform Act of 1966 was “to assure that all persons, regardless of their' financial status, shall not needlessly be detained pending their appearance to answer charges .,, when detention serves neither the ends of justice nor the public interest.” Id. § 2. The Act required: (1) a presumption of release on -personal recognizance unless the court determined that release would not reasonably assure the defendant’s- appearance in court;. (2) the option of conditional pretrial release under supervision. or other terms designed to decrease the flight risk; and (3) a prohibition on using money bail when nonfinancial release options such as supervisory custody or restrictions on “travel ... or place of abode” .could reasonably, assure the defendant’s appearance. See id. § 3, § 3146(a).

Congress again revised federal bail procedures with the Bail Reform Act of 1984, enacted as part of the Comprehensive Crime Control Act of 1984. See Bail Reform Act of 1984, 98 Stat. 1837, 1976 (codified at 18 U.S.C. §§ 3141-3150 (2012)). The legislative history of the 1984 Act states that Congress -wanted to “address the alarming- problem of crimes committed by persons on release”-and to “give the courts adequate authority to- make release decisions that give appropriate recognition to the danger a person may pose to others if released.” S. Rep. 98-225, at 3 (1983), reprinted in 1984 U.S.C.C.A.N. 3182, 3185. The 1984 Act, as amended, retains most of the 1966 Act but “allows a federal court to detain an arrestee pending trial if the Government demonstrates by clear and ‘ convincihg evidence after an adversary hearing that -no release conditions' ‘will reasonably assure ... the safety of any other- person and the community.’ ” United States v. Salerno, 481 U.S. 739, 741, 107 S.Ct. 2095, 95 L.Ed.2d 697 (1987) (omission in original) (quoting the Bail Reform Act of 1984) (upholding the preventive detention provisions in the 1984 Act).

The federal history of bail reform confirms that bail is a mechanism of pretrial release, riot of preventive detention. Pretrial preventive detention in federal cases requires counseled, adversarial hearings with findings stated on the record that, by *1075clear and convincing evidence, no less restrictive alternative can reasonably assure the defendant’s presence at trial. See id. In United States v. McConnell, 842 F.2d 105 (5th Cir. 1988), the Fifth Circuit held that “a bail setting is not constitutionally excessive merely because a defendant is financially unable to satisfy the requirement.” Id. at. 107. The magistrate judge in McConnell had initially ordered the felony defendant detained before trial with no release condition under the provision of the Bail Reform Act recently upheld in Salerno. See id. at 106. The district court replaced the detention order with a set of conditions for release, including weekly check-ins with pretrial services, travel restrictions, and secured money bail of $750,000. Id. The defendant moved for reconsideration, alleging that he did not have the assets to pay the secured money bail. The defendant appealed the district'court’s denial of his motion for reconsideration, and the Fifth Circuit remanded to the district court for a written opinion with findings on the record. Id. The court issued written findings, and a second appeal followed. Id.

The Fifth Circuit recognized that under the Bail Reform Act, Congress “pro-scrib[ed] the setting of a high bail as a de facto automatic detention practice.” Id. at 109. The Fifth Circuit relied on the Senate Report of the Bail Reform Act, which explained that:

section'3142(c) provides' that a judicial officer may not impose a financial condition of release that results in the pretrial detention of the defendant. The purpose of this provision is to preclude the sub rosa use of money bond to detain dangerous defendants. However, its application does not necessarily require the release of a person who says he is-unable to meet a financial condition of release which the judge has determined is the only form of conditional release that will assure the person’s future appearance. Thus, for example, if a judicial officer determines that a $50,000 bond is the only means, short of detention, of assuring the appearance of a defendant who poses a serious risk of flight, and the defendant asserts that, despite the judicial officer’s finding to the contrary, he cannot meet the bond, the judicial officer may reconsider the amount of the bond. If he still concludes that the initial amount is reasonable and nece'ssary then it would appear that there is no available condition of release that will assure the defendant’s appearance. This is the very finding which, under section 3142(e), is the basis for an order of detention, and therefore the judge may proceed with a detention hearing pursuant to section 3142(f) and order the defendant detained, if appropriate. The reasons for the judicial officer’s conclusion that the bond was the only condition that could reasonably assure the appearance of the defendant, the judicial officer’s finding that the amount of the bond was reasonable, and the fact that the defendant stated that he was unable to meet this condition, would be set out in the detention order as provided in section 3142(i)(l). The defendant, eould then appeal the resulting detention pursuant to section 3145.

Id. at 108-09 (quoting S.Rep. No. 225, 98th Cong.2d Sess. 16, reprinted in 1984 U.S.C.C.A.N. 3182, 3199), The Fifth Circuit concluded that the district court could set a secured money bail amount beyond the defendant’s ability to pay, but' “[i]n such an instance, the court must explain its reasons for concluding that the particular financial requirement is a necessary part of the conditions for release. It is sufficient for the court to find by a preponderance of evidence that the defendant poses a serious risk of flight.” Id. at 110, When federal *1076bail functions as an order of detention because of the defendant’s inability to pay, the court must treat the bail as an order of detention under § 3142(e) and must provide the procedural protections that section requires, with a preponderance-of-the-evidence standard. See also United States v. Mantecon-Zayas, 949 F.2d 548, 550 (1st Cir. 1991) (“once a court finds itself in this situation—insisting on terms in a ‘release’ order that will cause the defendant to be detained pending trial—it must satisfy the procedural requirements for a valid detention order”).

4. Bail under Texas Law

The Texas state appellate court practice is similar to federal court practice. Texas courts have imposed or confirmed high money bail after a judicial officer holds an adversarial hearing, with defense counsel present, and issues a reasoned opinion with written findings permitting secured money bail despite inability to pay in felony cases in which pretrial preventive detention without bail is available under Article I, § 11 of the Texas Constitution. See, e.g., Jobe v. State, 482 S.W.3d 300 (Tex. App.—Eastland 2016) (charge of capital murder); Ex parte Ragston, 422 S.W.3d 904 (Tex. App-Houston [14th Dist.] 2014, no pet.) (capital murder, first-degree murder, and aggravated robbery); Ex parte Vasquez, 558 S.W.2d 477 (Tex. Cr. App. 1977) (capital murder). In cases in which preventive detention is not available, Texas appellate courts have confirmed high money bail in felony cases when the evidence did not show the defendant’s inability to pay. See, e.g., Ex parte Dupuy, 498 S.W.3d 220, 233 (Tex. App.—Houston [14th Dist.] 2016, no pet.) (“Appellant offered no evidence, and we see none in the record, suggesting the trial court set his bail at $200,000 for each case in order to keep him incarcerated.”); Cooley v. State, 232 S.W.3d 228, 235 (Tex. App.—Houston [1st Dist.] 2007) (“Cooley owns half of a multi-million dollar air freight business and did not introduce evidence that revenues from it were unavailable to him.”); Ex parte Charlesworth, 600 S.W.2d 316, 317 (Tex. Cr. App. 1980) (evidence showed the defendants “live in a style inconsistent with poverty”); Ex parte Welch, 729 S.W.2d 306, 310 (Tex. App.—Dallas 1987, no pet.) (incomplete and conflicting evidence on ability to pay in a case charging solicitation of capital murder committed while the defendant was already on pretrial release on a secured money bail).

In a narrow set of felony cases, Texas courts have imposed or confirmed high money bail despite evidence of inability to pay the amount needed for pretrial release. “When the offense is serious, and involves aggravating factors that may result in a lengthy prison sentence,” a higher money bail than the defendant can pay is permissible, but only after satisfying the same due process requirements as an actual detention order. Dupuy, 498 S.W.3d at 230; see, e.g., Maldonado v. State, 999 S.W.2d 91 (Tex. App.—Houston [14th Dist.] 1999, pet. ref'd) (charges of possessing cocaine with a street value of $11-72 million with a possible sentence of 99 years); Ex parte Miller, 631 S.W.2d 825 (Tex. App.—Ft. Worth 1982) (charges of murder and rape carrying life sentence); Ex parte Runo, 535 S.W.2d 188 (Tex. Cr. App. 1976) (bail set at $125,000 on a charge carrying a life sentence was not excessive, but bail set at $75,000 on a charge carrying a minimum two-year sentence was excessive and had to be reduced to $5,000). Even so, Texas courts are careful to distinguish between transparent pretrial preventive detention orders and de facto pretrial detention orders imposed by setting bail higher than the defendant can pay. See, e.g., Dupuy, 498 S.W.3d at 230; Ex parte Harris, 733 S.W.2d 712, 714 (Tex. *1077App.—Austin 1987, no pet.) (setting bail “on the obvious assumption that appellant could not afford bail in that amount and for the express purpose of forcing [the defendant] to remain incarcerated” was overturned for abuse of discretion); Ex parte Nimnicht, 467 S.W.3d 64, 70 (Tex. App.—San Antonio 2015) (“There is no evidence the trial court set bail with the intent to prolong Nimnicht’s incarceration, especially in light of the fact the trial court reduced the bail amount.”).12

The defendants argue that, while Texas law forbids setting bail higher than a defendant can pay in order to impose a de facto pretrial detention order, if a judicial officer weighs all five factors of Article 17.15 of the Texas Code of Criminal Procedure and then imposes a bail amount that an indigent arrestee cannot pay, the bailas not a de facto detention order. (See, e.g., Docket Entry No. 164 at 8-10, 18; No. 263 at 3-4; No. 266 at 7-8). The defendants overstate the Article 17.15 factors and their role. The Texas cases make clear that a judge may arrive at a bail amount that a defendant cannot pay when the defendant is facing a felony charge carrying an extended prison sentence. Even then, the bail setting requires an adversarial, counseled hearing at which the defendant can put on evidence of indigence and likelihood of compliance with nonfinancial conditions of release, and reviewable findings, stated on the record, that the secured financial condition is reasonably necessary to assure the defendant’s appearance at trial or law-abiding conduct.13 In misdemeanor cases, *1078pretrial preventive detention is permitted only when a defendant is facing a family violence charge after previously violating a release condition in an earlier family violence case. In those cases, it is not necessary to use secured money bail to effect the detention of those who cannot pay. The Texas Constitution permits a transparent order of pretrial preventive detention.

5. Recent Distinctions Drawn Between Bail and Preventive Detention

a. Washington, D.C.

In 1994, Washington, D.C. amended its Code using language substantially similar to the federal Bail Reform Act. The amended Code permits a judicial officer to set “a financial condition to reasonably assure the defendant’s presence at all court proceedings that does not result in the preventive detention of the person, except as provided in” the Code’s regulations of preventive pretrial detention orders. D.C. Cope § 23-1321(c)(3). The Code permits preventive detention only in cases involving a charge of violent or dangerous crime, as Well as in cases presenting a “serious risk that the person will flee.” Id. § 23—1322(b)(1). To order preventive detention, a judge must: hold a hearing at the first appearance of the defendant' before a judicial officer; appoint counsel for. the defendant; permit the defendant to put on evidence, testify, and call witnesses; and make written findings “by clear and convincing evidence that no condition or combination of conditions will reasonably assure the appearance of the person as required,.and the safety of any other person and the community.” Id. § 23—1322(b) (2)—(d)(7)."

Judge Truman Morrison of the D.C. Superior Court credibly testified at the motion hearing that until the 1994 amendment, the D.C. courts did not order preventive detention outright. The statutory prohibition on using secured money bail to assure community safety was also “a dead letter)’ Hearing Tr. 2-2:137. “So in cases of any seriousness, judges made an effort nontransparently, never saying what they were doing out loud, to immobilize high-risk people—who they thought were high-risk people—with money bonds that they hoped would be beyond their reach.” Id. Judge Morrison testified that after the 1994 rule change, “[f]or' the high-risk people that we used to immobilize nontransparently, we turned to this preventive detention statute that was moldering on the bookshelf, and prosecutors and judges began using that for high-risk people.” Id. at 2-2:139. For- “somewhat serious misdemeanors who we had been keeping in the jail on lower levels of money bond,” judges began to order alternative nonfinancial conditions of release with supervision provided by D.C.’s pretrial services agency. Id. at 2-2:139-40. Based on a recent report by that agency, Judge Morrison testified that although secured money bail is still available under the D.C. Code, such bail is almost never imposed in misdemeanor cases. Transparent preventive detention orders are issued in only about 1.5 percent of misdemeanor cases, and then only after counseled, adversary hearings with findings on the record that there are no less restrictive conditions that will assure the defendant’s presence at trial or the safety of the community.14 Id. at 2-2:149, 154; D.C. Code § 23—1322(b)(2)—(d) (7).

*1079 b. New Mexico

In 2014, the New Mexico Supreme Court ruled that “[n]either the New'Mexico Constitution nor our rules''of criminal procedure permit a judge to set high bail for the purpose of’ preventing a defendant’s pretrial release.” Brown, 338 P.3d at 1292 (citing N.M. Const. art. II, § 13—substantially the same language as Tex. Const, art. I, § 11). The court explained that “[ijntentionally setting bail so high as to be unattainable is simply a less honest method of unlawfully denying bail altogether.” Id. The supreme court-held that the trial court had abused its discretion by requiring secured money bail “solely on the basis of an accusation of a sierious crime” and had failed to apply the New Mexico Code of Criminal Procedure requirement that trial -courts impose the least restrictive bail and release conditions to reasonably assure a defendant’s appearance and the public’s safety. Id. at 1291—92.

In 2016, New Mexico voters codified the holding of State v. Brown in a constitutional amendment that passed with 87.2 percent of the vote,15 The amendment provided that “[b]ail may be denied by a court of record pending trial for a defendant charged with a felony if the prosecuting authority requests a hearing and proves by clear and convincing evidence that no release conditions will reasonably protect the safety of any other person or the community. An appeal from an order denying bail shall be given preference over all other matters.” Constitutional Amendment 1, New Mexico Senate Joint Resolution 1, March 1, 2016.16 The amendment also required that “[a] person who is not detaina-ble on grounds of dangerousness nor a flight risk in the absence of bond and is otherwise eligible for bail shall not be detained solely because of financial inability to post a money or property bond.” Id. Under the amendment, courts cannot- order preventive detention for misdemeanor arrestees or accomplish the same effect by setting a secured -money bail that an indigent defendant cannot pay.

c. New Jersey

New Jersey recently amended its constitution and statutes to enact statewide bail reforms. The- changes went into effect on January 1, 2017. The New Jersey Constitution now provides that “[pjretrial release may be denied to a person if the court finds that .no amount of monetary bail, non-monetary conditions of pretrial release, or combination of monetary bail and non-monetary conditions would reasonably assure the person’s appearance in court when required, or protect-the safety of.any other person, or the community, or prevent the person from obstructing or attempting to obstruct the criminal justice process.” N.J. Const. art. 1, § 11.

In a detailed law-enforcement directive, the New Jersey Attorney General concluded that under New Jersey’s prior practice, “in most cases the critical determination whether a defendant [was] released pending trial or instead incarcerated in a county jail [was] not made by a judge issuing a well-reasoned court order. Rather, for all practical purposes, defendants [were] re*1080leased or detained based on whether they happened] to have the financial means to post bail.” N.J. Attorney General Law Enforcement Directive No. 2016-6 at 9. New Jersey changed. Its current system creates a presumption against the use of secured money bail unless the prosecutor can show that “no non-monetary release condition or combination of conditions would be sufficient to reasonably assure the defendant’s appearance in court when required”; “the defendant is reasonably believed to have financial assets that will allow him or her to post monetary bail in the amount requested by the prosecutor without having to purchase a bond from a surety company or to obtain a loan”; and “imposition of monetary bail set at the amount requested would ... -make it unnecessary for the prosecutor to seek pretrial detention.” Id. at 56. Secured money bail cannot be used to achieve'or to have the effect of a pretrial detention order. Out of 3,382 cases filed in the first month under the new law, judges imposed transparent orders of pretrial detention in 283 cases and denied pretrial detention when requested to do so in 223 cases. Secured money bail was set in only 3 cases. Pls. Ex. 7(k) at 1.

'.New Jersey does not distinguish between felony and misdemeanor cases but between numbered “categories” of offenses, making comparisons to the Texas misdemeanor bail system difficult. The New Jersey numbers are for all ease categories. Id. Because this approach clearly applies to more serious felony-level cases, the basis for applying it to misdemeanor eases is even stronger.

d. New Orleans

On January 12, 2017, the Council of the City of New Orleans, where the municipal courts have jurisdiction only over misdemeanor cases, passed a measure reforming its bail ordinance. Pls. Ex. 12(tt). The preamble states that “incarcerating people solely due to their inability to pay for their release through the payment of cash bond violates the Equal Protection Clause of the Fourteenth Amendment.” Id. at 1 (citing Barnett v. Hopper, 548 F.2d 550, 554 (5th Cir. 1977). The new ordinance requires that except for four enumerated offenses— battery, possession of weapons, impersonating a peace officer, and domestic violence—all misdemeanor arrestees are to be released on personal recognizance. Id. at 2-3. For those charged with one of the enumerated offenses, the municipal courts must “impose the least restrictive non-financial release conditions.” Id. at 3. “For any person who qualifies for indigent defense, or does not have the .present ability to pay, the Court may not set” any financial condition of release or a nonfinaneial condition of release “that requires fees or costs to be paid by the defendant.” Id. Other than the four specific exceptions for offenses that involve violence or other public safety threats, all defendants must be released with no financial conditions. If a financial condition is imposed, the defendant must have “the present ability to pay the amount set.” Id.

e. Maryland

On February 17, 2017, the Maryland Court of Appeals adopted detailed changes to its court rules, the main source of criminal procedural law in Maryland. The rule changes will take effect on July 1, 2017. Pls. Ex. 12(p)(i), Court of Appeals of Maryland, Rules Order, Feb. 17, 2017 at 3. The rule ’changes are “designed to promote the release of defendants on their own recognizance or, when necessary, unsecured bond” by establishing a “[p]reference” for “additional conditions [of release] without financial terms.” Id. at 33. All defendants—both felony and misdemean- or—must be released on personal recognizance or unsecured bond unless a judicial *1081officer makes written- findings on the record “that no permissible non-financial condition attached to a release will reasonably ensure (A) the appearance of the defendant, and (B) the. safety of each alleged victim, other persons, or the community.” Id. at 35. Even in those circumstances, the new rules require that “[a] judicial officer may not impose a special condition of release with financial terms in form or amount that results in the pretrial detention of the defendant solely because the defendant is financially incapable of meeting that condition.” Id. at 39. Pretrial detention must not be the intended use or the incidental effect of secured money bail.

Before the rule.change, the Maryland Attorney General wrote to the rules committee chairman that “[although Maryland law permits unconditional pretrial detention only where no conditions of release will reasonably protect the public or ensure the defendant’s appearance at trial, nearly every evaluation of Maryland’s pretrial system has found no relationship between a pretrial detainee’s perceived risk and the bond amount set,... Lower risk defendants are detained because they cannot afford the bail, while higher risk defendants who have access to financial resources are able to make bail and are often permitted to do so without imposition of other conditions to protect the public.” Pls. Ex. 12(p) at 3 (citing reports). An advisory memo from the then United States Attorney General stated that “[a]s a general proposition, Maryland’s judicial officials ... do not properly and consistently consider defendants’ individual circumstances, and particularly their financial resources, in making bail determinations. As a result, arrestees in Maryland habitually face extended periods of pretrial detention not as a result of their dangerousness to the community or because they pose a substantial risk of flight, but solely because they are unable to pay bail.” Pls. Ex. 12(p)(ii) at 7. The memo concluded that this system, and those like it, violated both state law and the federal Constitution. Id. at 4-11.

f. Alabama

Some of the same lawyers representing the plaintiffs in this case have brought similar actions challenging bail systems around the country. Several actions were resolved with an agreed final judgment.. These judgments typically state that “[i]f the government offers release from custody after arrest upon the deposit of money pursuant to a bail schedule, it cannot deny release from custody to a person, without a hearing regarding the person’s indigence and the sufficiency of the bail setting, because the person is unable to deposit the amount specified by the schedule.” Jones v. City of Clanton, Alabama, Civil No. 15-34, 2015 WL-5387219 at *4 (M.D. Ala. Sep. 14, 2015) (citing Pugh v. Rainwater, 572 F.2d 1053 (5th Cir. 1978); Bearden v. Georgia, 461 U.S. 660, 103 S.Ct. 2064, 76 L.Ed.2d 221 (1983); State v. Blake, 642 So.2d 959 (Ala. 1994)); see also Jenkins v. City of Jennings, Civil No. 15-252 (E.D. Mo. Dec. 14, 2016); Bell v. City of Jackson, Civil No. 15-252 (E.D. Mo. June 20, 2016); Thompson v. Moss Point, Civil No. 15-182, 2015 WL 10322003 (S.D. Miss. Nov. 6, 2015); Snow v. Lambert, Civil No, 15-567, 2015 WL 5071981 (M.D. La. Aug. 27, 2015); Cooper v. City of Dothan, Civil No. 15-425, 2015 WL 10013003 (M.D. Ala. June 18, 2015); Pierce v. City of Velda, Civil No. 15-570, 2015 WL 10013006 (E.D. Mo. June 3, 2015). In Jones, the court independently confirmed the need for relief, reasoning that “[b]ail schemes.such as the one formerly enforced in the municipal court result in the unnecessary pretrial detention of people whom our system of justice presumes to be innocent,” and that “[cjriminal-defendants, presumed innocent, must not be confined in jail merely because they are poor.” Jones, 2015 WL 5387219 at *3.

*1082The U.S. Department of Justice filed a statement of interest in Jones, stating that “[i]ncarcerating -individuals solely because of them inability to pay for their release, whether through the payment of fines, fees, or a cash bond,. violates the Equal Protection Clause of the Fourteenth Amendment.”17 See Varden v. City of Clanton, Alabama, Civil No. 15-34, Docket Entry No. 26 at 1 (M.D. Ala. Feb. 13, 2015). The Justice Department reasoned that because rigidly adhering to a secured money bail schedule “do[es] not account for individual circumstances of the accused, [it] essentially mandate[s] pretrial detention for anyone who is too poor to pay the predetermined fee. This amounts to mandating pretrial detention only for the indigent.” Id. at 9, After Jones, fifty of Alabama’s largest cities, accounting for 40 percent of the population, voluntarily reformed their bail systems to either release misdemeanor defendants on personal recognizance or, at a minimum, to set an early hearing to consider alternative methods’ of release to secured money bail. Pls. Ex. 12(l).

g. Calhoun, Georgia

Some of the plaintiffs’ counsel also represented the plaintiffs in Walker v. City of Calhoun, Georgia, Civil No. 15-170, 2016 WL 361612 (N.D. Ga. Jan. 28, 2016). A putative class of misdemeanor arrestees alleged that Calhoun detained them on prescheduled amounts of secured money bail that were not reviewed except at'court sessions held each Monday. Id. at *1. The trial court granted the plaintiffs’ motion for a preliminary injunction, finding that “keeping individuals in jail solely because they cannot pay for them release, whether via fines, fees, or a cash bond, is impermissible.” Id. at *10 (citations omitted). The court ordered Calhoun to- “implement post-arrest proceddres that comply with the Constitution,” and directed that “until Defendant implements lawful postarrest procedures, Defendant must release any other misdemeanor arrestees' in its custody, or who comé into its custody, on their own recognizance or on unsecured bond in a manner otherwise consistent with state and federal law and with standard booking procedures.” Id. at *14. The Eleventh Circuit vacated the injunction because requiring the defendant to implement constitutional procedures was “the archetypical and unenforceable ‘obey the law’ injunction” forbidden by Federal' Rule of Civil Procedure 65. Walker v. City of Calhoun, Georgia, 682 Fed.Appx. 721, 724, 2017 WL 929750 at *2 (11th Cir. Mar. 9, 2017). The panel did not consider the merits, instead remanding for the district court to enter a specific order consistent with Rule 65. See id.

Walker attracted significant attention. Ten amicus briefs were filed, including by the American Bar Association; the U.S. Department of Justice, the Pretrial Justice Institute and National Association of Pretrial Services Agencies, the Cato Institute, and various representatives of bail bonds associations, Georgia law-enforcement personnel, and other municipalities and their insurers. The relevant amicus briefs are includéd in the record here.

The American Bar Association’s amicus brief in Walker argued that “[m]onetary conditions of release should never be drawn from an inflexible schedule, should be imposed only after consideration of the *1083defendant’s individual circumstances, and should never prevent the defendant’s release solely because the defendant is unable to pay.” Pls. Ex. 12(ff) at 12: The Third Edition of the ABA Standards for Criminal Justice, Pretrial Release (3d ed. 2007), recommend-“procedures designed to promote the release of defendants on their own recognizance or, when necessary, unsecured bond.” Standard 10-1.4(a). Jurisdictions should impose financial conditions only “when no other conditions will ensure appearance,” and financial conditions “should not be employed to respond to concerns for public safety.” Standard 10-1.(4c)-(d). The Standards also emphasize that “[t]he judicial officer should not impose a financial condition, of release that results in the pretrial detention of a defendant solely due to the defendant’s inability to pay.” Standard 10-1.4(e).

The American Bar Association’s brief emphasizes that “[Unwarranted pretrial detention infringes on defendants’ constitutional rights, ‘making it difficult for the defendant to consult with counsel, locate witnesses, and gather evidence’and placing a particularly heavy burden on ‘poor defendants and on racial and cultural minorities.’” Pls. Ex. 12(ff) at 14 (quoting Standards at 32—33). The commentary to the Standards states that “[i]f the court finds that unsecured bond is not sufficient, it may require the defendant to post bail; however, the bail amount must be within the reach of the defendant and should not be at an amount greater than necessary to assure the defendant’s appearance in court.” Id. (quoting with emphasis Standards at 43-44). The brief concludes that detaining a defendant solely for failure to pay a secured financial condition of release is unwarranted and unconstitutional. Id.

The Justice Department’s brief expanded the statement of interest it submitted in the Jones Alabama bail case. The brief reasoned that, based on Supreme Court precedent, “[i]f a court finds that no other conditions may reasonably assure an individual’s appearance at trial, financial conditions may be constitutionally imposed—but ‘bail must .be set by, a court at a sum designed to ensure that goal, and no morel ” Pls. Ex. 12(dd) at 18 (quoting with emphasis Salerno, 481 U.S. at 764, 107 S.Ct. 2095). “Although the imposition of bail in such circumstances may result in a person’s incarceration,” the Department explained, “the deprivation of liberty in such circumstances is not based solely on inability to pay.” Id. But adhering to “fixed bail schedules that allow for the pretrial release of only those, who can pay, without accounting for ability to pay and alternative methods of.assuring future appearance, do not provide for such individualized determinations, and therefore unlawfully discriminate based on indigence.” Id.

The Justice Department’s argument is stated less strongly than the American Bar Association’s. While the American Bar Association argues that defendants must not be detained solely because of their inability to pay secured money, .bail, the Justice Department interprets “solely” to exclude those who cannot pay a, secured money bail because it has been set beyond then? reach due to their risk of flight. See id. Both arguments are consistent with the reforms surveyed above. Some jurisdictions, such as Washington, D.C., New Mexico, New Jersey, and New Orleans, do not permit secured- money bail settings to result in pretrial detention or operate as de facto pretrial preventive detention orders in misdemeanor cases, in line with the American Bar Association’s recommendations. Others, such as Maryland and Alabama, permit secured money bail to have the effect of detention 'only if the court follows the procedures required for pretrial preventive detention, in line with the Justice Department’s argument. In those cases, a judicial officer must make written findings *1084after an adversarial, counseled hearing that secured money bail in the amount set is the only, or the least restrictive, condition that can reasonably assure the defendant’s appearance at trial.

The Pretrial Justice Institute and the National Association of Pretrial Services Agencies submitted a brief in Walker using empirical data to argue that secured money bail, as opposed to an unsecured appearance bond; is never the only reasonable condition that will assure an individual’s appearance at trial or community safety. Pls. Ex. 12(hh). The brief presented data showing that those released on secured money bail do not appear at greater rates or commit new crimes at lower rates than those released on unsecured bonds. Id. Secured money bail schedules can effectively increase rates of appearance when they operate as detention orders, but “the use of such schedules inevitably leads to the detention of some, persons who pose little threat to public safety, but are too poor to afford release, while releasing others that pose a higher safety risk (but can afford to post bond).” Id. at 25.

In addition to the policy changes that a number of jurisdictions have already implemented or are in the process of implementing, even more jurisdictions have announced that they are examining or are about to reform their bail systems.18 A common theme among these reformed and reforming jurisdictions is that; before recent rule changes, each jurisdiction as a matter of routine practice either intentionally used or indifferently permitted the use of secured money bail as de facto detention orders against those financially unable to pay. See, e.g., Brown, 338 P.3d at 1292 (“We understand that this case may not be an isolated instance and that other judges, may be imposing bonds based solely on the nature of the charged offense without regard to individual determinations of flight risk or continued danger to the community.”); Pls. Ex. 12(p); Hearing Tr. 2-2:137. The other theme is that this practice did not hold up to historical, empirical, political, or legal scrutiny.

h. Conclusion

Whether by legislative enactment, judicial rulemaking, or court order, there is a clear and growing movement against using secured money bail to achieve a misdemeanor arrestee’s continued detention. Of course, it is not a federal court’s role in any way to make policy judgments. See, e.g., Brown v. Plata, 563 U.S. 493 at 537-38, 131 S.Ct. 1910, 179 L.Ed.2d 969 (2011). The question this case presents is not what is the best or even a good bail policy. The question is what bail system the Constitution requires and what system it prohibits. The Constitution sets minimum standards of due process and protects basic rights such as the presumption of innocence and the ability to prepare for trial. State and local governments may add to, but may not detract from, these basic protections. See, e.g., Gerstein, 420 U.S. at 124, 95 S.Ct. 854. The question is whether Harris County meets the constitutionally minimum standards and procedures.

D. The Use of Bail in Harris County Misdemeanor Pretrial Detention

1. The Statutory Framework

The Texas Code of Criminal Procedure defines “bail” as “the security given by the *1085accused that he will appear and answer before the proper court the accusation brought against him, and includes a bail bond or a personal bond.” Tex. Code Crim. Pro. art. 17.01. Except for certain types of felonies, “a magistrate may, in the magistrate’s discretion, release the defendant on his personal bond without sureties or other security.”19 Id. art. 17.03(a). A personal bond requires the defendant to swear an oath that if he or she fails to appear, the principal sum the court sets becomes due. Id. art. 17.04. The magistrate granting a personal bond may assess a nonrefundable bond fee “of $20 or three percent of the amount of the bail fixed for the accused, whichever is greater.” Id. art. 17.42, § 4(a). Magistrates may postpone, reduce, or waive the fee. Id. art. 17.03(g).

Texas law does not facially provide for release on no financial conditions. The “personal bond” defined in Texas law differs from what other jurisdictions call a personal bond or a personal recognizance bond by requiring a principal sum that becomes due if the defendant fails to appear. See, e.g., Goldfarb, supra, at 153-54 (personal recognizance in New York is release solely “on one’s honor pending trial”); Brown, 338 P.3d at 1289 (distinguishing release on “personal recognizance” from release “upon the execution of an unsecured bond,” which makes a sum due if the defendant fails to appear).20

The Texas Code of Criminal Procedure states that “[t]he amount of bail to be required in any case is to be regulated by the court, judges, magistrate or officer taking the bail; they are to be governed in the exercise of this discretion” by five rules:

1. The bail shall be sufficiently high to give reasonable assurance that the undertaking will be complied with.

2. The power to require bail is not to be so used as to make it an instrument of oppression.

3. The nature of the offense and the circumstances under which it vías committed are to be considered.

4. The ability to make bail is to be regarded, and proof may be taken upon this point.

5. The future safety of a victim of the alleged offense and the community shall be considered.

Tex. Code Crim. Pro. art. 17.15.

In Harris County, “magistrates” include Hearing Officers and County Judges. See id. art. 2.09. In addition to a magistrate’s discretion to issue a personal bond, the Texas Code of Criminal Procedure permits the arresting officer to release defendants *1086accused, of certain misdemeanors by citation only. Id. art. 14.06. The Code permits the arresting officer to cite-and-release those arrested for Class A or B misdemeanors for possessing small amounts of marijuana or certain other controlled substances, criminal mischief causing damage up to $2,500, graffiti, theft of property or service up to the value of $2,500, supplying contraband to prisoners, or driving without a license. Id. Major Patrick Dougherty testified that the Houston Police Department and Harris County follow the cite- and-release practice only for traffic-related Class C misdemeanor arrestees. Hearing Tr. 3-2:47-48 (“All Houston police officers basically book all their prisoners in the City Jail, regardless of whether it is a felony, misdemeanor or a Class C offense.”), 52. The Harris .County District Attorney has recently implemented a cite- and-release policy, as well as a diversionary program, for misdemeanor arrests for possessing small amounts of controlled substances. Hearing Tr. 3-2:178. Under the County’s diversionary program, the District Attorney’s office postpones charges for misdemeanor arrestees who agree to complete educational courses. These arrestees are not subjected to the booking and bail setting processes described below because the District Attorney declines charges at that time. Hearing Tr. 4-1:21-22.

The Texas Government Code permits the County Judges to “adopt rules consistent with the Code of Criminal Procedure ... for practice and procedure in the courts. A rule may be adopted by a two-thirds vote of the judges.” Tex. Gov’t Code Ann. § 75.403(f). At least three times since the beginning of 2016, the Harris County Criminal Courts at Law Judges, sitting en banc and voting by two-thirds majority, adopted or amended the Harris County Criminal Courts at Law Rules of Court.21 The current version is the Rules of Court as amended on February 9, 2017. The Rules of Court contain a misdemeanor bail schedule, id. Rule 9, and provide that “[t]he: initial bail amount may be changed on motion of the court, the hearing officer, or any party subject to the following criteria”:

4.2.3.1.1. the bail shall be sufficiently high to give reasonable assurance that the defendant will comply with the undertaking;

4.2.3.1.2. the nature of the offense for which probable cause has been found and the circumstances under which the offense was allegedly committed are to be considered, including both aggravating and mitigating factors for which there is reasonable ground to believe shown, if any;

4.2.3.1.3. the. ability to make bail is to be regarded, and proof may be taken upon this point;

4.2.3.1.4. the future safety of the victim and the community may be considered, and if this is a factor, release to a third person should also be considered; and . .

4.2.3.1.5. the criminal law hearing officer shall also consider the employment history, residency, family affiliations, prior criminal record, previous court appearance performance, and any outstanding bonds of the accused.

Id. Rule 4.2.3. The County Rules of Court state that “all law enforcement officials in Harris County shall cause the pretrial detainees in their respective custody, who have been charged with a class A or class B misdemeanor, to be delivered to the criminal law hearing officer not later than *108724 hours after arrest.” Id. Rule 4.2.1.1. Misdemeanor defendants arrested without a warrant who are not given a probable cause hearing within 24 hours after arrest must be released on a personal bond of no more than $5,000 when the 24 hours have expired.22 See Tex. Code Crim. Pro. art. 17.033; Sanders v. City of Houston, 543 F.Supp. 694, 705-06 (S.D. Tex. 1982); Roberson v. Richardson, Agreed Final Judgment, Civil No. 84-2974 (S.D. Tex. Nov. 25, 1987).23

At the 24-hour hearing, commonly referred to as the probable cause hearing, in addition to finding probable cause for the arrest, Hearing Officers are to “set the amount of bail required of the accused for release and shall determine the eligibility of the accused for release on personal bond, cash bond, surety bond, or other alternative to scheduled bail amounts, and shall issue a signed order remanding the defendant to the custody of the sheriff.” Rules of Court 4.2.2.1.11. On August 12, 2016, the County Judges amended the County Rules of Court to provide that “personal bonds”—unsecured appearance bonds—“are favored” in twélve specific misdemeanor categories.24 Id. Rule 12. Rule 12 lists five circumstances in which personal bonds “are disfavored,” including when “the defendant has demonstrated a risk to reoffend or harm society” or “has previously failed to appear in court as instructed.” Id.

The next step in «the process is scheduling cases for arraignment, referred to as the “first appearance settings.” Arrestees released on secured money bail., before booking are scheduled for arraignment one week from the day of their arrests (or on a Friday if the arrest was over a weekend). Id. Rule 4.1.2. Those released on a personal bond are scheduled for arraignment the same day, or the next business day if released , after 9:00 a.m. Id.. Rule 4.1.4. Those booked into the County Jail who request counsel are scheduled for arraignment the next business day, when counsel may be appointed. Id. Rule 24.9.1.

On February 9, 2017, the County Judges amended the County Rules of. Court to provide first appearance settings for all misdemeanor arrestees booked into the County Jail- the next business day after booking, “regardless of whether the defendant has been released from custody.” Id. Rule 4.1.2. At this first appearance, the County Judge must “review conditions of release, bail amount set, and personal bond decision and modify if good cause 'exists to do so.” Id.

2. Arrest and Booking

According to the 2015 annual report of Harris County Pretrial Services, 50,947 people were arrested in Harris County on only Class A or Class B misdemeanor charges in 2015. Pls. 10(c), 2015 Pretrial Services Annual Report at 8. In that year, 27.9 percent were arrested by the Harris County Sheriffs Office. The rest were ar*1088rested by other law-enforcement agencies, principally the Houston Police Department.25 Id.

Major Dougherty testified based on his thirty-five years of service with the Houston Police Department that those arrested without a warrant by the City of Houston are taken to the City Jail, tested for drugs or alcohol if applicable, and fingerprinted. Hearing Tr. 3-2:48-49. Either at the site of the arrest or at the City Jail, the arresting office calls a District Attorney hotline that is staffed 24 hours a day by an Assistant District Attorney, who decides whether to accept the charge. Id. at 3-2:50-51. If the Assistant District Attorney declines the charge, the arrestee is promptly released. Id. at 3-2:50. If the charge is accepted, an officer prepares a District Attorney Intake Management System (DIMS) report and electronically forwards it to the District Attorney’s office, where the formal charge is prepared. Id. at 3-2:48-49. Major Dougherty testified that by this point in the process, 1 to 3 hours have elapsed, depending on the time spent transporting arrestees from outlying areas of Houston to the City Jail. Id. at 3-2:49.

The District Attorney’s office prepares a formal charging document and applies a secured money bail using the County Judges’ bail schedule to set the amount. Id. at 3-2:51; see Rules of Court 2.3. The scheduled bail amount is set based on the charge and the defendant’s criminal history. See Rules of Court 9.1. The document is forwarded to the District Clerk’s office, which assigns the case to a County Judge’s court and sets a first-appearance date. Major Dougherty testified that the Clerk’s assignment makes the case “paper-ready” and is completed about 12 to 16 hours after arrest. Hearing Tr. 3-2:52. Once the case is paper-ready, a misdemeanor defendant with access to enough money may pay the amount necessary for release and be promptly released from custody.26 Id. at 3-2:54. A defendant may pay the entire bond amount into the registry of the court, to be refunded at case disposition if the defendant makes all scheduled appearances—a “cash bond.” Or, and most often, the defendant pays a nonrefundable premium to a commercial bondsman, who posts the principal with the court—a “surety bond.”27 See Pls. 10(c), 2015 Pretrial Services Annual Report at 9; 2-1:55-58.

Harris County Pretrial Services personnel have offices at the City Jail. Hearing Tr. 3-2:57, 59. Arrestees who have not bonded out at the earliest opportunity and *1089who have not already been taken from the City to the County Jail have their interview with a Pretrial Services officer while at the City Jail. Id. at 3-2:56-58. Harris County Director of Pretrial Services Kelvin Banks testified that interviews typically take 15 to 20 minutes. Id. at 3-2:170. In the interview, Pretrial Services asks for the defendant’s background information, including residence, employment, education level, and past criminal history. See, e.g., Pis. Ex. 8(d). After the interview,. Pretrial--Services tries to verify the information by calling references and running internet searches. Hearing Tr. 3-2:171.

Before this suit was filed, an unwritten policy required Pretrial Services to obtain two verified references before a defendant could be released on a personal bond. (See Docket Entry No. 162 at 5; No. 166 at 10 n.13); see also Def. Ex. 52. In August 2016, the County Judges sent a Tetter to the Hearing Officers changing the policy to permit release on personal bond with only one verified reference. Def. Ex. 52. The verification requirement is not codified in the County Rules of Court or in State law. Until the August 2016 letter, the requirement appears to have been an unwritten policy promulgated by County Judges and enforced as a practice or custom by Hearing Officers and by Pretrial Services personnel.28

Pretrial Services officers complete a validated risk-assessment form, which uses a point-weighting system to itemize and evaluate the defendant’s risk of flight or risk of new criminal activity during pretrial release. Hearing Tr. 3-2:172; Pls. Ex. 8(d). A risk-assessment tool is “validated” when its risk indicators have- been empirically shown to reliably predict outcomes such as nonappearance or new criminal activity. Hearing Tr. 3-2:172.

The current risk-assessment tool that Harris County Pretrial Services uses assigns points to seventeen different risk indicators. See, e.g., Pls. Ex. 8(d). Under “Criminal Risk Items,” arrestees are given a point if the current charge involves a crime of violence, a point if the defendant is on probation, a point if the defendant is on parole, a point for a prior misdemeanor conviction and another point for multiple prior convictions, a point for a prior felony conviction and another point for multiple prior convictions, a point for a past failure to appear, and a point if the .defendant has a formal “hold,” .such as an outstanding warrant from another jurisdiction. See id.; Hearing Tr. 2-1:134-37.

Under “Background Risk Items,” a defendant receives a point for being male, a point for lacking a high school diploma or GED, a point for not having a land line phone, a point for living with someone other than a spouse or family, a point for not'owning an automobile, a point for lacking full-time employment, ’and a point for being under 30 years of age.29 Hearing Tr. 2-1:134-37; see generally Pls. Ex. 8(d).

The point totals from both the Criminal Risk Items and Background Risk Items *1090are added to reach a single score which is set on a risk scale. Id. Defendants with three points or fewer are scored as low risk, four to five points are scored as “low moderate risk,” six to seven points are scored as moderate risk, and eight points or above are scored as high risk. Id. Criminal risk points are weighted the same way as background risk points. A 29-year-old man who works part-time and rents and apartment with a roommate, who does not own' a car or a land line phone, but who has no criminal history would receive the same risk score as an older woman on probation who has multiple felony convictions, a past failure to appear, am outstanding warrant and a current charge involving a crime of violence. Both cases would be assigned at least six points and be categorized' as “moderate” risk. See id. Mr. Banks testified that for defendants whose risk scores are increased because of the background factors that correlate with poverty rather than criminal activity, the standard Pretrial Services procedure is to recommend release on personal bond, notwithstanding the .higher risk score. Hearing Tr. 4-1:55.

The collected information, verified references, risk-assessment score sheet, and the Pretrial Services recommendation, for release are all gathered into a report and transmitted to a Hearing Officer,for the defendant’s probable cause hearing. See generally Pls. Ex. 8(d). Mr. Banks testified that currently, if Pretrial Services makes a recommendation, it recommends either that a Hearing Officer grant a personal (unsecured) bond with standard conditions (such as supervision by Pretrial Services), grant a personal (unsecured) bond with additional conditions (such as geographic restrictions), or “detain.” Id. at 3—2:173. Mr. Banks explained that Pretrial Services makes a recommendation to “detain” misdemeanor arrestees with immigration or other warrant holds on their record. Id. at 3-2:173-75; 4-1:41-44. A recommended high bail setting is intended to keep arres-tees detained to address the hold. Id. at 3-2:175; 4-1:41-42.-But, as explained below, secured money bail, if unpaid, prevents the defendant from addressing the hold or from being transferred to the agency imposing the hold, extending the overall time spent in custody.30 Mr. Banks testified that Pretrial Services also recommends “detain” for “high risk” arrestees. Id. at 3-2:173-75; 4-1:41-44. Because Texas law prohibits pretrial preventive detention inmost misdemeanor cases, a recommendation to “detain” is a recommendation to set a high secured bail in order to detain until a judicial officer considers the case or the case is terminated. Id. at 3-2:175; 4-1:41-42, A defendant who pays the bail is released, despite the “high- risk” .category and the recommendation to detain.

Pretrial Services does hot directly ask defendants whether they can pay the bail amount set or what amourit they could pay. Hearing Tr. 4-1:47. Mr. Banks testified that instead, Pretrial Services asks a “litany” of questions about a defendant’s assets, income, and expenses. Id. at 4-1:48. Defendants sometimes refuse to be interviewed, and the record indicates that they may do so because they do not understand that the interview is non-adversarial and that providing responsive answers is the only way they can be released on nonfinancial conditions. Hearing Tr. 2-2:29-30; 3-2:125. Other defendants are confused by the questions. For instance, “Do you have a place to stay?” may be taken to mean “Can you afford rent or housing?” But it could also mean “Are you likely to leave the jurisdiction because you do not have a place to live here?” Answering the first *1091question in the negative when the second question is the one asked can—and in Harris County does31—become the basis for detention rather than release on unsecured financial conditions. Hearing Tr. 2-2:12-13.

If Pretrial Services officers at the City Jail determines that an arrestee is a good candidate for release on personal bond, they may forward the interview papers to their counterparts at the County Jail for “early presentment” to a Hearing Officer. Hearing Tr. 3-2:182-83. The Hearing Officer may approve or deny release on personal bond using only the charging papers (the DIMS report) and interview papers prepared for the early presentment. Id. If the Hearing Officer approves release on personal bond, the arrestee can be released from the City Jail without being transported to or booked in the County Jail. Id. 3-2:57-58. Early presentment depends on the availability of Pretrial Services personnel. The record evidence clearly shows that early presentments are rare. In 2015, only 90 out of 21,748 Houston Police Department arrestees were released on personal bonds after early presentment.32 Pis. Ex. 4(d), Second Rebuttal Report at 15; Pis. Ex. 10(c), 2015 Pretrial Services Annual Report at 8.

Arrestees who do not pay for release or obtain release on personal bond by early presentment at the City Jail are taken to and booked in the Harris County Jail. Hearing Tr. 3-2:65. Transport buses run every two hours, but Major Dougherty testified that capacity limits at the County Jail Inmate Processing Center create significant delays. These limits prevent paper-ready misdemeanor arrestees from being transported to the County Jail on the next available bus. Id. at 3-2:67-68. Because the Inmate Processing Center is the only holding facility for County arrestees, they are given priority over arrestees waiting for transport from the City Jail, which adds to the delays. Id. at 3-2:69-70.

Major Dougherty testified that those arrested without a warrant by the County are taken either directly to the County Jail or to one of four outlying County detention centers, with transport to the County Jail within 4 hours. Id. at 3-2:54-55. Once at the County Jail, County arrestees go through the same process as City arres-tees—they are charged, fingerprinted, drug and alcohol tested, and interviewed by Pretrial Services in the . Inmate Processing Center next to the Jail. Id. at 3-2:84-85. The Inmate Processing Center runs 24 hours a day, 7 days a week. The booking process takes between 8 and 12 hours, Id. at 2-1:38. Booking at the County Jail relies on a paper, rather than an electronic, system. Id. at 3-2:85. Arrestees with the financial means .to .do so may pay their money bonds or a bondsman’s premium while still in the Inmate Processing Center and be released, usually within 12 to 15 hours of arrest. Id. While in the Processing Center, arrestees do not have access to counsel or family members. Id. at *10922-1:43. Those who do not pay their secured money bonds while in.the Processing Center are assigned and transferred to a housing unit in the County Jail. Id.

About 7 percent of misdemeanor arrests annually are arrested after a warrant has issued. Pls. Ex. 10(c), 2015 Pretrial Services Annual Report at 8; (see also Docket Entry No. 290, Ex. 1 at 8). When Harris County magistrates—Hearing Officers or County Judges—issue warrants, they affix the prescheduled secured money bail amount to the warrant. Hearing Tr. 2-1:216-17; 3-2:55; 8-2:108. Misdemeanor defendants with access to money can pay the secured amount or a bondsman’s premium and be processed without being arrested. Id. Misdemeanor defendants who áre arrested on a warrant but cannot pay are subject to the same procedures as warrantless arrests. Id. Although a magistrate has already found probable cause to issue the warrant, the Texas Code of Criminal Procedure requires each arrestee to appear before a magistrate to be informed of his or her rights and to request counsel. Tex. Code Crim. Pro. art. 15.17. What Harris County calls “probable cause hearings” fulfill the function of informing arrestees of their rights, finding probable cause for warrantless arrests, and setting bail or, more often, confirming the pre-scheduled amount of bail on a secured basis.

3. The Probable Cause and Bail-Setting Hearing

The Hearing Officers hold probable cause hearings every 2 hours, 24 hours a day, 7 days a week. Hearing Tr. 4-1:160. Defendants arrested without a warrant who have completed processing at the Inmate Processing Center are put on the next available docket for a probable cause hearing. Id. at 2-1:93. Probable cause hearings are conducted by videolink connecting a Hearing Officer’s courtroom, an Assistant District Attorney’s office, and a large room in the County Jail. See generally Pls. Ex. 2. Up to forty-five arrestees may be adjudicated at a single probable cause hearing. Hearing Tr. 4-1:161. When an arrestee’s case is called, the arrestee stands on a marked square in the center of the room and faces, a screen showing the Hearing Officer and Assistant District Attorney. See Pls. Ex. 2. The hearings are recorded. Id.

Hearings typically last about one to two minutes per arrestee.33 See generally Pls. Ex. 2. During this brief period, the Assistant District Attorney reads the charge, and the Hearing Officer determines probable cause and sets bail. Id. Hearing Officers have discretion to release arrestees on personal bond, to impose additional conditions of release (such as geographical restrictions), or to raise or lower the bail amount from the scheduled amount. See Tex. Code Crim. Pro. art. 17.03, 17.15, 17.40-44. For those misdemeanor arres-tees who have not had their Pretrial Services papers given to a Hearing Officer for early presentment—the vast majority—the first setting is their earliest opportunity to be considered for release on a personal bond. Hearing Tr. 2-1:45-46.

As noted, under Texas law, misdemean- or defendants arrested without a warrant must be released on an unsecured personal *1093bond if a magistrate does not find probable cause within 24 hours of arrest. Tex. Code Crim. Pro. art. 17.033. The Houston City-Jail is not equipped to provide videolink hearings and does not provide opportunities for live presentment to the Hearing Officers. Hearing Tr. 3-2:61-63. When the Inmate Processing Center at the County Jail is at capacity and arrestees cannot be transported promptly from the City Jail, those who have not paid and been released may wait at the City Jail more than 24 hours before they are transported to the County Jail and can have their probable cause and bail-setting hearing before a Hearing Officer. Id. at 4-1:9,136. To avoid releasing these arrestees on unsecured personal bonds at the. 24-hour time limit, the customary unwritten practice is to hold in absentia “paper hearings.” Id. at 2-1:92-93. At a paper hearing, the Hearing Officer finds probable cause based on the DIMS report that the arresting officer prepared and that the Assistant District Attorney used to draw up the charge. Id. Pretrial Services forms are not made available at paper hearings. The DIMS report does not provide any of the defendant’s financial information. Hearing Officers do not set bail or consider eligibility for unsecured personal bonds at paper hearings. Id. at 4-1:133-35.

Defendants almost never have counsel at the probable cause and bail-setting hearing. See Def. Ex. 23. Those who are indigent have not yet had counsel appointed. Those who can afford counsel have either paid their bonds and been released or have not been able to arrange their counsel’s presence. See id. Both the Sheriffs deputies and the Hearing Officers instruct the defendants not to speak except to answer specific questions, lest they incriminate themselves. Hearing Tr. 4-1:178. Because the Hearing Officers are not judges of courts of record, they do not make written findings or issue reasoned opinions explaining why they set bail on a secured or unsecured basis, or why they select the bail amount imposed. (Docket Entry No. 138 ¶ 72); Hearing Tr. 4-1:145.- The video recordings show that Hearing Officers occasionally state that bail is set at a certain level or that a personal bond is denied “based on your .priors” (see below). Hearing Officers occasionally make notes on the Pretrial Services forms, such as “Criminal History”; “Safety of Community”; or “Safety.” See Pls. Ex. 9, e.g., (M. W.), (A. G.), (H. P.). These cryptic, one-to-three word notations are just that. They do not show that Hearing Officers weighed the statutory factors in setting bail, much less how they did so.

Chief Hearing Officer Blanca Villagomez testified that before granting an unsecured personal bond, she “look[s] at the five factors obviously that are set out in Article 17.15. I listen to the prosecutor and whatever allegations that led to their charge, secondly. I will look at all of the information that is available to me that is provided by Pretrial Services and reach a conclusion on that.” Hearing Tr. 4-1:117.-She testified that on occasion, based on the circumstances and the evidence presented, she has denied release on an unsecured personal bond to defendants who score low on the risk scale because she • perceived a threat to public safety. On other occasions, she disregards a high risk score based on background resource factors, such as not owning a land line phone or a car. Id. at 4-1:126. Judge Villagomez testified that she does not reach a conclusion on whether secured money bail will operate as a condition of detention, but that she does realize that detention, ,-rather than release, will be the outcome of setting secured money bail for indigent defendants more than “rarefiy].” Id. at 4-1:140-42. She nevertheless sets bail on a secured -basis at the scheduled amounts in those-cases. Id. She *1094testified that she believes it is lawful under Texas law to require a secured money bail she knows a defendant cannot pay “if I have taken in all of the factors in 17.15 into consideration because [ability to pay] is not the only one.” Id. at 4-1:144-45.

Hearing Officer Eric Hagstette testified that' he discounts high risk scores when they are based on background factors showing poverty rather than a history of nonappearance or criminal activity. Id. at 4-1:163-65. He did not disagree with Judge Villagomez’s approach. He testified that the Hearing Officers “all go about -our job pretty much the same way, do what we are statutorily required to do during these hearings and then make the decision with the information that is available and is presented at the hearing.” Id. at 4-1:168-69. He explained that he does not impose secured money bail with' an intent to detain but that “[t]he intent is'to set a bond that is sufficiently high based on the factors I’m obligated to consider.” Id. at 4-1:171. When asked how he would approach a defendant with no job, no income, no assets, and a history of failing to appear, for whom the scheduled bond' amount would be $4,000, he testified that he would not release that defendant on an unsecured $4,000 bond because “[i]t depends again on the other factors being balanced.” Id. at 4-1:172.

Judge Villagomez testified that she does not and cannot keep track of how many times she raises or lowers a bond, how often she rejects a Pretrial Services recommendation, or whether, and how often, defendants she releases on unsecured personal bonds fail to appear at hearings. Id. 4-1:126-27, 132, 150. Judge Hagstette testified that he raises and lowers bail amounts in roughly equal numbers—“I knock them down and I raise, them up”— but he does not know how often those he releases on unsecured personal bonds fail to appear. Id. at 4—1:163, 167-68,173.

' The Pretrial Services Annual Report provides system-wide statistics on how often Hearing Officers implement or' reject Pretrial Services recommendations. Pis. Ex. 10(c), 2015 Pretrial Services Annual Report at 14. In 2015, for the 9,388 defendants for whom Pretrial Services recommended release on unsecured personal bond with standard conditions of supervision, Hearing Officers denied a personal bond 56.3 percent of the time.34 Id. In 1,831 cases, Hearing Officers granted release on unsecured personal bonds on the condition that Pretrial Services could verify the references. Id. The data do not show in how many cases that did or did not happen.35 For the 4,816 defendants for *1095whom Pretrial Services recommended release on personal bond with enhanced supervisory conditions, Hearing Officers denied a personal bond 84.8 percent of the time.36 Id. For the 11,935 defendants for whom Pretrial Services made no recommendation, Hearing Officers denied a personal bond 96.9 percent of the time.37 Id. For the 4,716 defendants for whom Pretrial Services recommended “detain” (15.3 percent of all defendants interviewed by Pretrial Services), Hearing Officers denied a personal bond 97.1 percent of the time.38 Id. Overall, Hearing Officers reject Pretrial Services recommendations for release on a personal bond 66,3 percent of the time. Id. Pretrial Services acknowledges the wide discrepancy between what they recommend based on the County’s validated risk-assessment tool and what the Hearing Officers order based on the preset bail schedule. The Frequently Asked Questions page on the Pretrial Services public website asks, “Why aren’t there more Personal Bonds approved?” The answer: “Good question!”39 Hearing Tr. 4-1:57.

Among all cases in which Pretrial Services interviewed the misdemeanor defendant, whether Hearing Officers granted release on secured or on unsecured financial conditions, the Hearing Officers lowered the bail amount from what was-stated on the charging document in 7.2 percent of cases and raised the- bail amount in 10.7 percent of cases.40 Pls. Ex. 10(c), 2015 Pretrial Services Annual Report at 14. In 2015, Hearing Officers lowered the-amount below $500—the' minimum amount on the bail schedule—in 4 cases, out of nearly 51,000 arrests with bail set.41 Id. at 8. The plaintiffs’ expert, Dr. Stephen Demuth, credibly testified that from the beginning of 2015 to the end of January 2017, Hearing Officers adhered -to the prescheduled bail amount stated on the charging documents in 88.9 percent of all misdemeanor cases.42 Pls. Ex. 4(d), Second Rebuttal Re*1096port at 10; Hearing Tr. 6-2:119-21. When they do change the amount, they raise it about 67 percent of the time. Id.

Dr. Demuth presented credible evidence based on Harris County’s administrative data that from January 2015 through January 2017, only 9.7 percent of all misdemeanor arrestees were granted release on an unsecured personal bond, with or without additional nonfinancial conditions. Pis. Ex. 8(d), Second Rebuttal Report at 9. That figure is consistent with the Pretrial Services annual reports, which show that 8.5 percent of misdemeanor arrestees were granted an unsecured personal bond in 2015, and 10.8 percent in 2016.43 Pis. Ex. 10(c) at 9; (Docket Entry No. 290, Ex. 1 at 9). In 2015, 46.1 percent of arrestees were released on a surety bond, 5.1 percent on a cash bond, and the remaining 40.3 percent were detained until case disposition. In 2016, the figures were nearly identical: 43.4 percent released on a surety bond, 5.6 percent on a cash bond, and 40.1 percent detained until case disposition. Id. Virtually all misdemeanor arrestees detained until disposition have a secured bail amount set that, if paid, would result in the prompt release of the arrestee. See Pls. Ex. 10(c) at 8; (Docket Entry No. 290, Ex. 1 at 8).

The court credits the Hearing Officers’ testimony that they consider the Article *109717.15 factors in some way. But their impressions about how frequently certain case outcomes occur is not reliable and not worthy of greater weight than the data presented in the Pretrial Services Annual Report. The Hearing Officers’ testimony that they do not “know” whether imposing secured money bail will have the effect of detention in any given case, e.g., Hearing Tr. 4-1:141, 4-2:16, and their testimony that they do not intend that secured money bail have that effect, is not credible. Other record evidence, including the Pretrial Services public reports; the high number and percentage of misdemeanor defendants detained rather than released because they are subject to secured money bail at the scheduled amount; the high number and percentage whose bail is set by the schedule rather than by an individualized inquiry; the infrequency of deviations from imposing the scheduled bail amount on a secured basis; and the video recordings of probable cause hearings, which consistently show an indifference as to whether pretrial detention will result from setting secured bail, all weigh heavily in favor of finding little to no credibility in the Hearing Officers’ claims of careful case-by-case consideration under the Roberson order and the Article 17.15 factors.

This is not a personal criticism of any one or all of the Hearing Officers. To say that their job is difficult is a dramatic understatement. The sheer numbers of defendants the Hearing .Officers confront on a daily basis makes individual consideration extraordinarily difficult. The absence of counsel adds to the difficulty. The Hearing Officers clearly work steadily and hard. They see a difficult population—including both misdemeanor and felony defendants—every day and all day. It is unsurprising that a. system of virtually automatic adherence to a bail schedule has developed, given the large number of defendants, the small number of Hearing Officers, and the. limited time for hearings.

The record contains 2,300 recordings of misdemeanor probable cause hearings before the Hearing Officers. The recordings begin in March 2016—before the lawsuit was filed—and continue through early November 2016. Pis. Ex. 2. The court has reviewed many hours of foQtage. The results are consistent and support this court’s findings and conclusions. Two hearings are illustrative. The court chooses them not because they are extreme examples of any particular feature, but because they appear pretty ordinary. Neither hearing is procedurally unusual. The parties did not cite or play either one at the motion hearing.44

*1098D. M. was arrested early in the morning of August 24, 2016 and charged with possessing less than two ounces of marijuana. See Pis. Ex. 4(b), Working Database, His probable cause hearing was at 4:00 p.m. the same day. Id. The recording shows the following:

• The Hearing Officer finds probable cause and tells the defendant* ‘Tour bond is incorrect based on” his five prior felony and nine misdemeanor convictions. Pis. Ex. 3, August 24, 2016, 15.22 at 37:25.

• The defendant responds that he has only one prior felony conviction. The Hearing Officer spends the bulk of the unusually long four-and-a-half minute hearing thumbing through the défendant’s record arid counting convictions. The Hearing Officer counts as prior felony convictions two felony charges that were reduced to misdemeanor convictions but still does not arrive at five felony convictions. He tells the defendant, “Either way your bond was incorrectly set, so it’s now set at $5,000, which is what it should have been set at. [I’m] going to deny your personal bond based on all your priors.” Id.

• The defendant requests a personal bond because his fiancée is pregnant and he is the only income earner in the household. The Hearing Officer responds, “I take all that into consideration” but again points to the defendant’s prior convictions. The defendant points out that he .has never missed a court appearance for any of those prior arrests and convictions. The Hearing Officer cuts him off, stating, “That is one factor, the other factor is everything else.... Based on the nature of the offenses for which you were charged, I’m not going to consider you” for a personal bond. Id.

• The defendant confirms he will need a court-appointed lawyer. The Hearing Officer concludes that if the defendant would like a personal bond he can ask the County Judge for one in the morning at his first appearance. Id.

If the defendant had been able to pay a bondsmanV premium, he would have been released notwithstanding his criminal history. D.M. appeared before a County Judge the next day and pleaded guilty. He was released later that day. See Pls. Ex. 4(b), Working Database.

A. G. was arrested on October 1, 2016 at 9:30 p.m. for unlawfully wielding a five-inch knife. See id. His probable cause hearing was held the next afternoon. It is one of the more recent recordings in evidence. Id. The recording shows the following:

• The Hearing Officer finds probable cause and confirms the scheduled secured money bail amount of $2,500. Pls. Ex. 3, October 2, 2016, 12.16 at 27:39.

• The defendant confirms that he will need a court-appointed lawyer and tries to ask a question. The Hearing Officer cuts him off, stating, “Nobody who’s got the criminal history you have out of Flqrida is'going 'to get a pretrial [bond] from me, for fear of what would happen to the safety of the community.” The defendant again tries to speak. The Hearing Officer again cuts him off: “I have more people to consider than you in this, and the safety of the public is one of them.” The defendant tries a third time to speak, and again the Hearing Officer shouts over him, saying “You’re not going to be able to talk to me because I’m not letting you talk, because I’m. going by what I feel is best for the community,” Id.

• After a pause, the defendant quietly asks if he may speak. The Hearing *1099Officer shouts “No!” The defendant pauses again and then states that his only criminal history is a 25-year-old matter in Florida and that he is nearly finished with his exams to become a medical professional. The Hearing Officer responds that “your 25 year ago tendencies seem to be revisiting me, and I am afeared for the people in the State of Texas.” Id.

• The Hearing Officer again confirms that the defendant will need a court-appointed attorney, then dismisses him. As the defendant leaves the room, the Hearing Officer quips to the Assistant District Attorney that it “makes me feel better” that the defendant is returning to detention. The Assistant District Attorney laughs. Id.

The defendant’s first appearance before a County Judge was held the next day but then reset for October 7, 2016. Pls.’ Ex. 4(b), Working Database. At the rescheduled hearing, after seven continuous days in detention, A.G. was released on an-unsecured personal bond. See id. His case remained pending at the time of the most recent data production from the County. There is no indication’that he has failed to appear or has been re-arrested since October 2016. See id.

The two recordings illustrate what many other recordings confirm. Hearing Officers treat the bail schedule, if not as binding, then as a nearly irrebuttable presumption in favor of applying secured money bail at the prescheduled amount.45 Amounts that deviate from the schedule are.treated as “incorrect,”46 and requests for a personal bond, if not denied outright, are deferred until the County Judge holds a later hearing.47 Hearing Officers routinely adjust initial bail settings to conform to, not to deviate from, the bail schedule. Defendants who try to speak are commanded not to, shouted down, or ignored.48

The Hearing Officers testified that they cannot let one factor—the inability to *1100pay—control their bail determination. Healing Tr. 4-1:124, 171. But they frequently cite only one factor—criminal history—as controlling them decision to set secured money bail that the defendant clearly cannot pay.49 And although the Hearing Officers testified that they do not “know” in any given case whether a defendant can pay secured money bail, they routinely set secured scheduled money bail amounts despite: (1) being informed of a defendant’s indigence on the Pretrial Services report; (2) being told of a defendant’s indigence by the defendant; (3) being aware that a defendant’s charge clearly relates to poverty (such as begging or sleeping at a bus stop); and (4) recording that a defendant needs court-appointed counsel because of indigence.50 The evidence that this occurs is overwhelming. The Hearing Officer’s wisecrack that setting a $2,500 bond for reasons of community safety “makes me feel better” clearly shows intent to use secure money bail to detain that defendant indefinitely.51

The court finds and concludes that in the typical case, Hearing Officers set se*1101cured money bail as a condition, of detention operating only against those who are indigent and cannot pay the bail, rather than as a mechanism for pretrial release. In the vast majority of cases, the Hearing Officers use their discretion to consider the five Article 17.15 factors to almost automatically impose the prescheduled secured bail amounts, notwithstanding Pretrial Services recommendations to release defendants on unsecured personal bonds and notwithstanding clear evidence of indigence. Hearing Officers make these decisions in brief, uncounseled hearings at which the defendants are actively discouraged from speaking, and no reviewable findings are made on the record.

4. The First Appearance Before a County Judge

Before the most recent change to the County Rules of Court in February 2017, any “incarcerated person” who remained in detention after the probable cause hearing would be scheduled to appear before a County Judge “the next business day” after the probable cause hearing. At this first appearance before a County Judge, counsel was appointed if requested. Rules of Court, Rule 24.9. The plaintiffs offered unrebutted testimony that, although misdemeanor defendants were taken to the County Courthouse on the scheduled day, they usually did not appear in the courtroom before the County Judge unless they offered to plead guilty at that time. Hearing Tr. 2-1:59-61, 63; 3-1:8-9. Bail review was at the County Judge’s initiative, and done only in a minority of the eases. In some cases, the review was prompted by a Pretrial Services recommendation or by defense counsel. See Pls Ex. 10(c), 2015 Pretrial Services Annual Report at 15; Hearing Tr. 5:108; 7-2:56-61. One County Judge testified that in his experience as a former criminal defense attorney, seeking a bail reduction before a County Judge was formally available, but practically futile. Hearing Tr. 2-1:10. Defendants who did not plead guilty but wanted to contest their bail settings depended on court-appointed counsel filing a formal motion for bail review. That motion would not be considered until a later hearing, usually held one or two weeks later. Id. at 3-1:10-11. The only way to gain release earlier was to pay the bail or to plead guilty.

'The February 9, 2017 amendment took effect oh March 9, 2017. The amended County Rules of Court require “any arres-tee that is booked into the Harris County Jail” to be presented at a “Next Business Day Setting,” even if that arrestee is released from custody between booking and the next business day. Rules of Court, Rule 4.3.1. If the probable cause hearing has not been held by the Next Business Day Setting, the County Judge rather than the Hearing Officer will determine probable cause and set bail. Id. The amended Rules state that “[ajbsent a waiver by the defendant and defense counsel, the court will review conditions of release; bail amount set, and personal bond decision and modify if good cause exists to do so.” Id.

Judge Darrell Jordan, the presiding judge of County Criminal Court at Law No. 16, testified that in e-mail exchanges, some County Judges have objected that *1102because the new Rule 4.3.1 is not based on the Texas Code of Criminal Procedure, County Judges do not have to review bail at the Next-Business Day Setting. Hearing Tr, 3-1:98, 117. Whether County Judges do or do not review the bail that the Hearing Officers set, Judge Jordan testified that in his experience as a criminal defense attorney on many misdemeanor cases, seeking a bail review at the first appearance was futile because .County Judges “stick to. the bond schedule. That would be the answer. What does the bond schedule say?” Id. at 3-1:10.

Judge Jordan testified that he takes’ a different approach to bail from his fifteen County Judge colleagues. Id. at 3-1:16-17. The Hearing Officers and other County Judges who. testified agreed. Id. at 3-2:176-77; 4-1:113-16, 180-81; 5:62, 120. Judge Jordan testified that in his interpretation of the Roberson order and Article 17.15 factors, the first factor—that “the bail shall be sufficiently high to give reasonable assurance that the undertaking will be complied with”—means that a person who has the funds available for a secured money bail should post, security within his or, her financial means to assure appearance at trial. Id. at 3-1:62-63. The fourth factor—requiring consideration of ability to pay—means that bail should be set ’at an amount and on terms the defendant can meet to be released. That may mean a $2,000 secured bail if the defendant can afford the $200 commercial surety premium, and an unsecured personal bond if the defendant cannot pay the premium. Id. at 3-1:65-66. “Otherwise, I[ would be] keeping them in jail because they can’t afford the bond.” Id. at 3-1:66.

As for the nature of the offense and consideration of community safety, Judge Jordan testified that he reviews charging documents with the assigned Assistant District Attorney before the first appearance hearings. Id. at 3-1:66-67; For- cases that present troubling charges or circumstances, Judge Jordan has the defendants appear in his courtroom and engages them in a colloquy. Id. “I want to talk -to them and fully understand what is going on so then I can make a decision on what we should do with their bond.” Id. at 3-1:66. “But at-no time in my analysis do I say setting a money bond is going to make them a better person or make the victim safer because the person had $500.... Money does not make somebody safe.” Id. at 3-1:71. Instead, Judge Jordan testified that he orders additional, nonfinancial conditions of release on personal bond, such as GPS monitoring for those at risk of violating a protective order. Id. at 3-1:73-74.

Judge Jordan has experienced the Harris County misdemeanor pretrial justice system both as a lawyer representing defendants and as a County Judge ruling on defendants’ cases. He does not believe that the Texas Code or County Rules of Court are unconstitutional as written. He testified that judges can apply the rules in a constitutional manner, and that the way he applies them is constitutional. Id. at 3-1:87-88. But he also believes that outside of his jurisdiction over those assigned to County Court No. 16, the County engages in a widespread practice of detaining misdemeanor defendants before trial on secured bail amounts County personnel know the defendants cannot pay because they are indigent. Id. at 3-1:61. Judge Jordan testified that without an injunction from this 'court, that practice will continue. Id.

Testifying on behalf of- herself and County Judge Margaret Harris, Judge Paula Goodhart, the presiding judge of County Criminal Court at Law No. 1, testified that she believes. Judge Jordan “consider[s] one factor and one factor only, *1103which is the ability to pay.” Id at 5:115, 120. She interprets the Roberson order and Article 17.15 to require her to “look at a person’s individual liberty and'weigh that with the risk to the community and the risk that they are not going to appear and consider it altogether with all of those factors and set a reasonable' and rational bond that we believe is going to secure their reappearance and it is going to minimize their risk to reoffend,” Id. at 5:116. She testified that “[a]fter going through the whole process, I have set a bond that I did not think it was likely that the person could make, not as an .instrument of oppression or with the intent to-detain, but because after considering all of the factors, that was the reasonable and rational non-excessive thing to do.” Id. at 3-1:121.

Judge Goodhart disagreed with Judge Jordan’s conclusions about the incentives resulting from the secured money bail system. She testified to her understanding that under Texas law, having a bond revoked for new criminal activity creates a financial incentive for those who post secured money bail to comply with the conditions of their release. Id. at 3-1:123. Judge Goodhart apparently did not know that Texas does not permit a financial forfeiture when a defendant released on bond commits a new offense. See Tex. Code Crim. Pro. art. 22.01-02 (permitting forfeiture with a right to collect the financial security only in cases of failure to appear). In fact, the re-arrest of a defendant during pretrial release guarantees that the bond will not be forfeited. Id. art. 22.13(5).

Judge Goodhart testified that no Harris County policymaker, so far as she is aware, has examined Harris County data to compare pretrial failure-to-appear rates or bond forfeiture rates between those released on secured or unsecured financial conditions. Hearing Tr. at 5:137-38. Her impression was confirmed by Director of Pretrial Services Kelvin Banks and the Hearing Officers. See id. at 3-2:146; 4-1:149-50, 162-63. Dr. Marie VanNostrand, the County’s consultant on pretrial reform, testified that Harris County may collect the data that would allow this study but has never undertaken such a study or compiled the data to do so. Id. at 6-1:111. Judge Goodhart testified on behalf of herself and another County Judge that even if she learned from such a study that secured money bail provides no financial incentive to comply with the conditions’ of release, it would not change her subjective belief that secured money bail is better for community safety than unsecured bail. Id. at 5:131.

The court finds and. concludes that the Harris County policymakers with final authority over the County’s bail system have no adequate or reasonable basis for their belief that for misdemeanor defendants, release on secured- money bail provides incentives for, or produces, better pretrial behavior than release on unsecured or nonfinancial conditions; The policymakers are apparently unaware of important facts about the bail-bond system in- Harris County, yet they have devised and implemented bail practices and customs, having the force of policy, with no inquiry into whether the bail policy is a reasonable way to achieve the goals of assuring appearance at trial' or law-abiding behavior before trial. In addition to the absence of any information about the relative performance of secured and unsecured conditions of release to achieve thesé goals, the policymakers have testified under oath that their policy would not change despite evidence showing that release on unsecured personal bonds or with no financial conditions is no less effective .than release on- secured money bail at achieving the goals of appearance at trial or avoidance of new criminal activity during pretrial release.

*1104Dr. Demuth presented uneontroverted and reliable evidence that in 2015 and 2016, the County Judges changed the bond amount and type from that set by the Hearing Officers in fewer than 1 percent of misdemeanor cases. Pls. Ex. 4(d), Second Rebuttal Report at 10; see also Pls. Ex. 10(c), 2015 Pretrial Services Annual Report at 15. That is compelling evidence that, like the Hearing Officers, County Judges presiding over Court Nos. 1 through 15 are not making individualized bail assessments under either the Roberson Order or the Article 17.15 factors.

. The County’s rule change to require a bail review at a defendant’s first appearance within one business day of booking, rather than within one business day of the probable cause and bail-setting hearing, has been in effect only since March 9, 2017. See Rules of Court 4.3.1. The record evidence does not show whether this earlier bail review has had any effect. Mr. Bob Wessels, who served as court administrator for the County Criminal Courts at Law for decades, testified that the bail review is not a new change to the rules but a codification of prior consistent practice. Hearing Tr. 5:29. Judge Jordan credibly testified that some of his colleagues have refused to conduct bail reviews, even under the new rule. Id. at 3-1:98. Judge Jordan also testified that bail reviews are usually futile because the County Judges adhere to the bail schedule on a secured basis. Id. at 3-1:10. Assistant District Attorney JoAnne Musick testified that, before the rule change, County Judges typically presumed that a misdemeanor defendant was indigent and appointed counsel if the defendant was still detained at the first appearance. Id. at 2-1:60-61.. Only those who had posted bond to be released were made to submit an affidavit of indigence before counsel could be appointed. Id. That means that for years, under a consistent practice now codified in the County Rules of Court, the County Judges have presumed misdemeanor defendants to be indigent because they remained detained by their inability to pay a secured financial condition of release, yet in about 99 percent of the cases, the County Judges have neither adjusted the bail nor granted release on unsecured or nonfmancial conditions. There is no basis in the record to find or conclude that the rule change requiring bail review at the Next Business Day Setting has altered or will alter these practices.

5. Disposition of Misdemeanor Cases

Unless a district attorney declines a charge or a Hearing Officer finds no probable cause, the earliest opportunity to dispose of a misdemeanor case is at the defendant’s first appearance before a County Judge, if the defendant pleads guilty and is sentenced.52 Ms. Musick testified based on her lengthy experience as a criminal defense attorney that many misdemeanor defendants “don’t really want to plead guilty, but sometimes they want to get out of jail, return to family, return to work, what have you. So they will inquire about a plea so that they can get out.” Hearing Tr. 2-1:65. Judge Jordan testified that in his experience, Assistant District Attorneys would make a plea offer in 85 or 90 percent of the misdemeanor cases at a defen*1105dant’s first appearance. Id. at 3-1:14. Both testified that the typical sentence for those pleading guilty at a first appearance is either the time already served in pretrial detention, or some number of days that with a two-for-one or three-for-one credit for the time served would allow release within a day of the first appearance. Id. at 2-1:67; 3-1:12. Judge Goodhart testified that prosecutors sometimes threaten to seek sentencing enhancements for certain offenses to convince misdemeanor defendants to plead and receive a time-served sentence at their first appearances. Id. at. 6:114.

Another indication that misdemeanor defendants abandon valid defenses and plead guilty to obtain faster release than if they contested their charges is a report from the National Registry of Exonera-tions showing that Harris County has led the United States in the total number of criminal exonerations each of the last two years. Def. Ex. 110. Most of Harris County’s exonerations come from misdemeanor drug offenses that evidence samples conclusively prove the defendant did not commit. See id. But rather than wait for lab tests that may exonerate them, misdemeanor arrestees who cannot pay for release before their first appearances plead guilty in order to end their pretrial detention and be released. See id.; Pls. Ex. 13(a); 7(h) at 2.

Defendants who do not plead guilty at the first appearance have a hearing set, generally two or three weeks, later. Id. at 2-1:10. Defendants who cannot pay their bail during this time remain in pretrial detention. Id. at 2-1:10-11. Defendants released on bond typically have their hearings set much later. Dr. Demuth presented uncontroverted and reliable testimony that from 2015 to early 2017, for misdemeanor arrestees who did not bond out—40 percent of all misdemeanor arrestees—the median time between arrest and case disposition was 3.2 days. Of those, 72 percent resolved their cases within 7 days; 90 percent resolved their cases within 30 days. Pls. Ex. 4(d), Second Rebuttal Report at 4. Over the same period, for misdemeanor arrestees released on bond (either secured or unsecured)—60 percent of misdemeanor arrestees—the median time to disposition was 120 days. Of those, 5 percent resolved their cases within 7 days; 13 percent resolved their cases within 30 days. Id.

Dr. Demuth presented uncontroverted and reliable testimony, based on the County’s own data, that the likelihood of a conviction differs dramatically depending on whether a defendant is detained before trial. In 2016 and 2016, 84 percent of misdemeanor arrestees detained at case disposition pleaded guilty, while 49 percent of those released before disposition pleaded guilty. Id. at 4. Only 13 percent of those still detained at case disposition had their cases dismissed, and 2 percent received deferred adjudications. Id. For those released before case disposition, 32 percent had their cases dismissed and 12 percent received deferred adjudications. Id.; Pls. Ex. 4(b), First Rebuttal Report at 16-18. These figures are consistent with, and support, the plaintiffs’ theory that for misdemeanor defendants unable to pay secured money bail, Harris County maintains a “sentence first, conviction after” system that pressures misdemeanor defendants to plead guilty at or near their first appearances because that is the only way to secure timely release from detention. Hearing Tr. 6-2:172-73.

In one of the most sophisticated and rigorous studies of bail and pretrial detention in misdemeanor cases to date,53 re*1106searchers at the University of Pennsylvania examined hundreds of thousands of Harris County misdemeanor arrest cases. The results are presented in a peer-reviewed study forthcoming in the Stanford Law Review. Pls. Ex. 12(d), Paul Heaton et al, The Downstream Consequences of Misdemeanor Pretrial Detention 69 Stan. L. Rev, (forthcoming) (July 2016) (“Heaton-Study”); Pls. Ex. 12(d)(i) (peer-review policy for empirical research). The Heaton Study analyzed the differences in case outcomes between misdemeanor defendants, who did not post bond within the seven days following the probable cause hearing, and those who did post bond within that period and were released. The researchers found that the still-detained defendants were 25 percent (14 percentage points) more likely to be convicted, and 43 percent (17 percentage points) more likely to be sentenced to jail than those who bonded out earlier. Id. at 4. Detained defendant? received sentences nine days longer on average, more than double the average sentence of similar, released defendants. Id. The researchers concluded that the fact of detention itself, rather than the. defendant’s charge, criminal - history, or other variables, causally affects these outcomes. Id. at 3-4.

The" findings of Dr. Demuth and of the Heaton Study are'supported by the record, casé law, and commentary. The case law and commentary recognizes that those released from pretrial detention are better able to consult with counsel and prepare a defense without hazarding- their' employment, housing, or family obligations. See, e.g., Brown, 338 P.3d at 1287 (“Congress attempted to remediate' the array of negative impacts experienced by defendants who were unable to pay for their pretrial release,. including the adverse effect on defendants’ ability to consult.with counsel and prepare a defense, the financial impacts on their families, a statistically less-favorable outcome at trial and sentencing, and the fiscal burden .that pretrial incarceration imposes on society at large.”) (citing 1966 U.S.C.C.A.N. at 2293, 2299); (see also Docket Entry No. 182 at 7; No. 272 at 9). Above all, they are free from the pressure to plead guilty as the only way to be released from detention in a reasonably short period, Ms. Musick credibly testified *1107that many of her misdemeanor clients chose to abandon valid defenses by pleading guilty at the first appearance so they could get out of jail instead of remaining detained for the two or three weeks it would take even to raise those defenses— or their inability to pay secured bail—in court. Hearing Tr. 2-1:68-69. Judge Jordan credibly testified that it was common to have misdemeanor clients who professed their innocence and had valid defenses to nevertheless plead guilty in order to be released much earlier than if they sought an unsecured bond based on indigence or challenged the prosecution’s ease. Id. at 3-1:11-12.

The defendants note that every misdemeanor defendant who pleads guilty affirms under oath that he or she does so voluntarily. That is true. It is also true that the County Judges engage the misdemeanor defendant in a counseled colloquy to affirm that the.plea is made voluntarily. Id. at 2-1:83-84; 3-1:98-99; 5:112-13. But these arguments miss the point. The credible, reliable, and well-supported testimony of the witnesses and the statistical studies in the record overwhelmingly prove that thousands 'of misdemeanor defendants each year are voluntarily pleading guilty knowing that they are choosing a conviction with fast release over exercising their right to trial at the cost of prolonged detention. This Hobson’s choice is, the evidence shows, the predictable effect of imposing secured money bail on indigent misdemeanor defendants.

6.. The Use of Bail to Detain

The consistent testimony of the plaintiffs’ witnesses, including Assistant District Attorney Musick, Sheriff Gonzalez, and Judge Jordan, is that Harris County routinely detains misdemeanor defendants, who would be released if they could pay secured money bail, because they are unable to pay the amount needed for release. Hearing Tr. 2-1:68-69; 3-1:51-52, 61; 3-2:9; see also Pls. Ex. 7(h) ¶ 5; (Docket Entry No. 206). The consistent testimony of the Hearing Officers and other County Judges is that they do not detain misdemeanor defendants solely because they cannot pay but because the balance of state-law factors,' including the need to ensure future appearances and to protect community safety, require money bail that is secured and generally (in 90 percent or more of the cases) set at the scheduled bail amount, calculated básed on the charge and the defendant’s criminal history and no other factors. These requirements, the defendants testified, can and frequently do outweigh the misdemeanor defendant’s inability to pay the bail on a secured basis. Id. at 4-1:117-18, 123-25, 144-45, 168-69; 5:34, 58, 71-72; Def. Ex. 23. At the motion hearing, the parties agreed with the court’s characterization of this conflict as one between a but-for cause and' a proximate cause view of detention. Id. at 1:99— 100; 4-2:15-16. In the plaintiffs’ theory, thousands of defendants are detained but for their ability to pay secured money bail, (See, e.g., Docket Entry No. 143 at 15-17; No. 188 at 4-7). In the defendants’ view, secured money bail for many defendants is out of reach because of the defendants’ problematic criminal history, the serious nature of the charges, the need for mental health evaluations, or other factors. (See, e.g., Docket Entry No. 162 at 15-16; No. 164 at 8-9).

Closer examination of the record evidence and the hearing testimony undermines the defendants’ proximate-cause explanation for detention. First, there is the overwhelming credible evidence that, with the exception of Judge Jordan, Harris County. Hearing Officers and County Judges do not make individualized determinations of bail based on each defendant’s circumstances, but instead consis*1108tently adhere to the predetermined bail schedule. Second, the facts established by-other overwhelming evidence undermines the judicial defendants’ position that in many cases, their individualized review shows that the public interest in the misdemeanor defendant’s appearance in court and law-abiding behavior before trial requires secured money bail at the scheduled amount, notwithstanding the misdemeanor defendant’s apparent indigence and the state-law prohibition on preventive detention orders in misdemeanor cases. See, e.g., Hearing Tr. 5:71-72.

The defendants argue that secured money bail provides incentives not delivered by unsecured personal bonds to induce appearance at trial. See, e.g., Hearing Tr. 5:127-28. The defendants cite what they call the “indemnitor effect”: commercial sureties and acquaintances of a defendant who put up the money for the defendant’s pretrial release on secured bail have an incentive to ensure that defendant’s return to court. See, e.g., id. at 4-2:79-80; 6-2:68. Under Texas law and the County Rules of Court, however, unsecured personal bonds provide similar incentives, or lack thereof. Texas law requires those released on unsecured personal bonds to swear to appear or forfeit the principal bond amount. See Tex. Code Crim. Pro. art 17.04. Harris County Pretrial Services is required by its policies to supervise misdemeanor defendants released on unsecured personal bond, keep them informed of court dates, administer drug tests and other appropriate monitoring services, and send out an investigator when a defendant fails to appear. Hearing Tr. 3-2:148; 4-1:18-19; Pls. Ex. 10(c), 2015 Pretrial Services Annual Report at 5. Pretrial Services is required by its policies to supply the “indemnitor effect” for those released on unsecured bonds. At most, commercial sureties and a defendant’s social network can prompt the defendant to appear at hearings, or, in the case of sureties, petition the courts to revoke the release on bond. But these are the same actions that Pretrial Services may—and under its policies, must—take for those released on unsecured personal bonds. Compare Hearing Tr. 5:127-28, with id. at 4-1:18-19; Pls. Ex. 10(c), 2015 Pretrial Services Annual Report at 5.

Formally, the financial incentives are the same across bond types. Those who are released and fail to appear either forfeit a cash bond, become civilly liable to Harris County for the principal bail amount, or become civilly liable to a bondsman for the principal bail amount.54 In each case and for each category of bond, nonfinancial incentives provide more powerful reasons to appear. These reasons include fear of a warrant for re-arrest and the possibility of being charged with, and convicted of, an additional misdemeanor for failure to appear. See Tex. Penal Code § 38.10.

At bottom, even if there were á difference between the indemnitor effects of having a commercial bondsman paid by the defendant’s friends or family monitor and encourage the defendant’s appearance,55 versus having Harris County Pretrial Services provide the monitoring and encouragement, that difference cannot be the basis for imposing secured, rather *1109than unsecured, bail without making indigence at least the proximate cause of the differential treatment. The defendants essentially argue that co-indemnitors—family and friends with access to money— makes secured bail a better assurance of appearance than unsecured bail. See, e.g., Hearing Tr. 3-1:16, 129. On that basis, the homeless and the friendless are denied release on personal bond because they lack co-indemnitors.56 See id. at 3-1:43-44; Pis. Ex. 1, Appendix E at 6. An indigent homeless individual’s lack of co-indemnitors is, however, both a cause and a consequence of indigence. The rigid demand for secured, rather than unsecured, money bail from a homeless individual is indistinguishable from an order that a misdemeanor defendant so indigent as to be homeless be detained because of that indigence.

Other than a fully-paid-up-front cash bond, the unsecured personal bond and the secured surety bond provide an equivalent lack of financial incentives to.appear during pretrial release. Harris County personnel testified that the County does not try to collect unsecured bonds forfeited for nonappearance. Hearing Tr. 3-1:36; 3-2:148; 6:24. Even if commercial bondsmen file civil suits to collect forfeited bond amounts, for misdemeanor defendants who lack assets—who are judgment-proof— that civil liability does not create a meaningful incentive. See id. at 1:190; 3-2:148-49; 4-1:170. The up-front payment of the bondsman’s premium is a sunk cost, and is not recoverable even if the defendant appears for every court date. Id. at 2-1:63; 4-2:13-14. Neither secured nor unsecured bonds provide meaningfully different financial incentives.

The incentive argument fares no better with respect to deterring new criminal activity during pretrial release. The evidence is that neither a secured nor unsecured bond is subject to forfeiture for new criminal activity. See Tex. Code Crim. Pro. art. 22.01-02; 22.13(5); Hearing Tr. 4-1:58-59. The record establishes that requiring secured money bail provides no incentive to -law-abiding behavior during pretrial release that is not equally provided by unsecured personal bonds—the main incentive, of course, being the threat of re-arrest and extended sentences for new criminal activity, incentives that apply equally across all classes of released defendants. See Tex. Penal Code § 38.10.

Secured money bail ensures better results than unsecured appearance bonds only when the secured money bail operates as an order of detention because the defendant cannot pay. Those who are detained because they cannot pay secured money bail necessarily make their court appearances and do not re-offend. But that success is because of the detention, not because of the financial security. And it applies only to those who cannot, pay the secured financial conditions of release.57

*1110The defendants argue that even if judges gave greater consideration to a misdemeanor defendant’s inability to pay, the defendants argue, some indigent defendants would still be detained under other state-law factors, such as a history of prior failures to appear or criminal convictions. The defendants cite Ms. ODonnell and Mr. Ford as examples. (Docket Entry No. 164 at 18). The problem is that although there is no meaningful difference in the financial or other incentives provided between secured and unsecured money bail, those with “priors” will be detained on secured bail, only if they are too poor to pay it. See also Hearing Tr. 5:33-34. The defendants repeatedly argue that because Texas law does not permit pretrial preventive detention in most misdemeanor cases, the only way to address serious concerns about nonappearance or new criminal activity is with a secured money bail too high for the defendant to pay. Hearing Tr. 1:115-16; 3-1:72-73; 5:43-44, 70; (Docket Entry No. 166 at 13-14). But it is the fact of pretrial detention, not the secured money amount,- that addresses these concerns, and only for those too poor to pay. An arrestee with access to money but with similar present charges, similar prior failures to appear, and similar criminal history could pay the secured bond and be released, despite the risks to public safety or of nonappearance. That arrestee would face no meaningfully different incentives than if released on an unsecured bond for the same amount.

Both Judge Goodhart and Judge Villago-mez testified that one reason they reject Pretrial Services recommendations to release defendants on unsecured financial con&tions in some cases is that the judges are able to access and consider the charging documents and other information that make the misdemeanor offense worse than the charge makes it appear, while Pretrial Services is limited to resource and criminal-history information obtained in the interview with the defendant. Hearing Tr. 4-1:125-26; 5:69-70; see also id. at 5:6-7. This is not a credible explanation for why the Hearing Officers and County Judges adhere to the bail schedule nearly 90 percent of the time. That aside, the judges’ reasoning assumes at the least that if the circumstances surrounding the crime appear graver than the misdemeanor charge on its own indicates, imposing secured money bail at the scheduled amount will induce better pretrial behavior from the defendant. That assumption has no basis in evidence or experience in misdemeanor cases when the defendant is released. In effect, the defendants’ position is that misdemeanor defendants should be incarcerated for the risks they pose, but only if a secured financial condition beyond the their ability to pay accomplishes the incarceration.

The fact that the defendants consistently interpret the Roberson order and Article 17.15 of the Texas Code of Criminal Procedure to refer only to secured bail is telling. The order and the Code provision refer only to “the amount of bail.” The Code defines “bail” as both secured and unsecured bonds. See Tex. Code Crim. Pro. art. 17.01; 17.15. While the defendants may increase the bail amount based on a misdemeanor defendant’s past, conduct, neither the Code nor the Roberson order require the higher level of bail to be imposed only on a secured basis. Judge Hagstette acknowledged that on occasion he has, con*1111sistent with the order and Code, set misdemeanor bail at the maximum scheduled amount of $5,000 but on an unsecured basis, so that the defendant could be released on a personal bond. Hearing Tr. 4-1:169; see also Ex parte Gentry, 615 S.W.2d 228, 231 (Tex. Cr. App. 1981) (confirming bail at $2,500 but ordering release on “the security of a personal bond in the amount fixed”). The fact that Hearing Officers and County Judges rarely engage in this practice shows they set'secured money bail not with an eye to the incentives provided by higher bail amounts, but with the understanding and expectation that secured bail will detain outright. Their shorthand for personal bonds- as “PR bonds,” meaning “pretrial release bonds,” betrays the same understanding. Id. at 3-2:86-87; 4-1:75, 169; see also Pls. Ex. 3, February 8, 2017, 6.41 at 37:36; November 2, 2016, 6.06 at 1:00:02; May 12, 2016, 9.49 at 24:41.

Although the Texas Code consistently states that the purpose of thé probable cause hearing is to “determine[ ] whether probable cause exists to believe that the person committed the offense,” see, e.g., Tex. Code Crim. Pro. art. 17.033, the orders the Hearing Officers issue are titled “probable cause for further detention,” see generally Pls. Ex. 9. On these orders, Hearing Officers check a box stating that. “[t]he Court FINDS PROBABLE CAUSE for further detention EXISTS” and requiring that the “Defendant shall remain in the Sheriffs custody until he posts [secured] bail in this cause.” Id. In Harris County, secured money bail is not just a de. facto pretrial detention order; it is literally a pretrial detention order.

The plaintiffs’ understanding of those detained “solely” because they are financially unable to pay secured money bail at the scheduled amount more accurately describes the current reality in Harris County. While Texas law guides the judicial officers? discretion in setting bail amounts, it does not require bail to be set on -a secured basis. Judicial offic’ers in Harris County follow a custom and practice, without sufficient basis in data or experience, of setting bail on a secured basis to address concerns about a defendant’s risk of failing to appear or of committing new criminal activity. The only way that secured bail addresses those concerns is by effectively ordering pretrial preventive detention. This occurs only when, and because, the defendant is too poor to pay the amount of bail imposed. In Harris County, secured financial conditions of release in misdemeanor cases effectively function as detention orders only against the indigent.

E. The Population Statistics of Misdemeanor Detainees . at Each Stage in the Post-arrest Process

In mid-February 2017, Harris County produced data drawn from its administrative, records purporting to account for all adults booked into the Harris County Jail from January 1, 2015 to February 14, 2017. Def. Ex. 28 at 2. The data set included 106,055 case entries. Pls. Ex: 4(b), Second Supplemental Report at 1. Id. at *2-3. Both parties’ experts relied principally on this data set to reach their conclusions about the misdemeanor population in the Harris County Jail.

1. Arrestees Detained More than 24 Hours Before the Probable Cause Hearing

From 2015 to early 2017, nearly 67 percent of misdemeanor arrestees were detained from arrest until the probable cause hearing. Pls. Ex. 4(b), Expert Report at *2, Second Supplemental Report at 2. Almost all of the' remaining 33 percent paid a secured money bond to' be released before the probable cause hearing. Only 90 people were released on personal bond through early presentment to a Hearing Officer in *11122015, and 240 in 2016. That is around 1 percent of arrestees held in custody by the City of Houston Police Department. Id; (Docket Entry No. 207-1 at 15; No. 290, Ex. 1 at 8).

Of those still detained at the probable cause and bail-setting hearing, more than 14,000 misdemeanor defendants—a little over 20 percent of those detained at that point58—waited more than 24 hours after arrest for the hearing. Pls. Ex. 4(d), Second Rebuttal Report at 1. Over 600 people—1.0 percent of those detained—waited more than 72 hours after arrest for the hearing.59 Id. The plaintiff Robert Ryan Ford was detained 32 hours after his arrest before he appeared before a Hearing Officer. See Pls. Ex. 8(c)(iii), Ford Docket Sheet.

Under Texas law, Harris County is required to release misdemeanor defendants if they have not had a probable cause hearing within 24 hours of arrest. Tex. Code Crim. Pro. art. 17.033. Release must be on an unsecured personal bond if the defendant cannot pay secured money bail. Id. at 17.033(b). Probable cause hearings for those arrested by the City of Houston Police Department may be delayed because of crowded conditions at the County Jail, causing backups in transporting arrestees from the City to the County Jail and booking them there.60 For some whose probable cause hearings are delayed more than 24 hours after their arrests, the Hearing Officers may hold hearings in absentia or “on the papers.” Hearing Tr. 2-1:92-93. The Hearing Officers find probable cause based on the DIMS report provided in-the charging documents. Id. That situation rarely occurs. The parties’ experts agreed that only 3 to 4 percent of the entire arrest population has probable cause determined on the papers. See Def. Ex. 28A; Hearing Tr. 6-2:31-32, 121-22. That means Harris County has over the last two years detained more than 10,000 misdemeanor arrestees more than 24 hours after arrest without either a probable cause hearing or a probable cause determination on the papers. And on the relatively few occasions when Hearing Officers make probable cause determinations on the papers,, they testified that they do not consider the amount of bail or eligibility for release on unsecured personal bond at that time. Id. at 4-1:133-35.

The court finds and concludes that Harris County is not providing a bail-setting hearing within 24 hours in thousands of cases.

2. Arrestees Detained More than 48 Hours Before a Bail Review

From 2015 to early 2017, nearly 50 percent of misdemeanor arrestees were de-*1113tamed from arrest until their first appearance before a County Judge. Pls. Ex. 4(d), Second Rebuttal Report at 2. In April 2016, one month before the plaintiffs filed suit, only 7-5 percent of all misdemeanor arrestees were released on personal bond, almost all of them by Hearing Officers at the probable cause hearing. Def. Ex. 47; Pls. Ex. 4(d), Second Rebuttal Report at 10. By the end of 2016, seven months after the plaintiffs filed suit and three months after the County Judges changed the Rules of Court to instruct the Hearing Officers to presume that unsecured personal bonds for twelve offense categories,61 16 percent of all misdemeanor arrestees were released on personal bond. Def. Ex. 47; Pls. Ex. 10(b), December 2016 Pretrial Services Monthly Report. The overall rate of release of misdemeanor defendants on unsecured personal bonds from 2015 to early 2017 was 9.7 percent—10.8 percent in 2016 alone. Pls. Ex. 4(d), Second Rebuttal Report at 9; (Docket Entry No. 290, Ex. 1 at 9).

Over the last two years, around 52,000 misdemeanor arrestees were still detained after their probable cause hearings before the Hearing Officers. Pls. Ex: 4(d), Second Rebuttal Report at 2. The next hearing, before a County Judge, is generally within one business day after the probable cause hearing. But more than 26,000 misdemean- or arrestees—over 51 pércent of those still detained—waited more than 48 hours after their arrests before their first appearances before a County Judge. Id. Over 6,800 people—-just over 13 percent of the detained population—were held longer than 96 hours after arrest before their first appearance. Pls. Ex. 4(b), Second Supplemental Report at 1. The plaintiff Loetha McGruder was detained 87 hours after her arrest before her first appearance before a County Judge. Pls. Ex. 8(c)(ii), McGruder Docket Sheet.

The defendants dispute these numbers, but their expert, Dr. Morris, provided no alternative figures on the length of detention between arrest and first appearances. See Def. Ex. 28A at 1-3. He argued that Harris County’s data contains too -many gaps, clerical errors, and problematic distributions to provide a basis for reliable calculations or conclusions. Id.; Hearing Tr. 4-2:201-03. Dr. Morris specifically cited a distribution chart showing hours-to-release as containing too many sharp peaks and valleys, indicating that the data did not accurately reflect the length of detention. Def. Ex. 28A at 1-3; Hearing Tr. 6-2:79.

The plaintiffs’ expert, Dr. Demuth, accounted for the problems Dr. Morris identified. Dr. Demuth excluded arrestees who had holds, had prior failures to appear, were on probation, faced multiple charges, faced concurrent felony charges, had prior convictions, were admitted for mental health or medical evaluations, or had high-risk designations. Pls. Ex. 4(b), Rebuttal Report at 3. Dr. Demuth found the same rates and distribution of delays across the remaining population. Id. Dr. Demuth testified that the peaks and valleys in the distribution are likely caused by the fact that Harris County does not record the time of first appearance. Hearing Tr. 6-2:112-16. Dr. Demuth adjusted for this by assuming that first appearances occur for all defendants at 9:0t) a.m., when the County Courts open their sessions for the day. Id. at 6-2:115-16. This is a realistic estimate and a conservative approach. Rather than the smoother distribution that actually occurs as. arrestees make their first appearances throughout the day, the 9:00 a.m. assumption makes the distribution reflect and exaggerate the rhythms of the *1114County’s arrest cycle. The relatively more numerous misdemeanor defendants arrested in the afternoon and evening appear to have their first appearances all at once at 24-hour intervals of 9:00 a.m. on the days after their arrest. The relatively smaller number arrested late at night make then-assumed 9:00 a.m. appearances seem relatively scarcer. Id. at 6-2:112-16. Dr, Demuth’s calculations and conclusions are reliable and helpful, even with the gaps’and flaws in the Harris County records and data. Of course, Harris County is welcome to provide more accurate information at the merits trial. On the present record, Dr. Demuth has sufficiently addressed Dr. Morris’s concerns by basing his calculations on realistic and conservative assumptions.

The court finds and concludes that at least half of the detained misdemeanor population in Harris County wait 48 hours or longer after arrest before seeing á County Judge, and at least 13 percent wait 96 hours or longer.

3. Arrestees Detained Until Case Disposition

Harris County’s annual and monthly Pretrial Services reports show that a remarkably stable 40 percent of misdemean- or arrestees remained detained until case disposition.- See generally Pls. Ex. 10(b), 10(c). In both the 2014 and 2015 annual reports, the rate is identical: 40.3 percent. Pls. Ex. 10(c). The 2016 Pretrial Services Annual Report, released after the motion hearing, shows that 40.1 percent of misdemeanor arrestees were detained until case disposition in 2016. (See Docket Entry No. 290, Ex. 1 at 8). The 2016 change in the County Rules of Court to presume release on personal bond in twelve offense categories has apparently had little impact.62

Of the 84 percent of detained arrestees who plead guilty at their first appearance,63 67 percent are released within a day. Pls. Ex. 4(d), Second Rebuttal Report at 3. About 83 percent are released within five days of their first appearance. Id. Those who do not plead guilty typically wait for one to three weeks or more before a second hearing before a County Judge. See Hearing Tr. 2-1:68; 6-2:168-69.

' 'Dr. Demuth testified that the likelihood a misdemeanor defendant will be detained at disposition correlates strongly with the indicators of poverty Pretrial Services uses to assess risk. Those Who had one point for criminal risk on the assessment but no points' for background, or resource, factors were detained at disposition 14 percent of the time. Id. at 7-1:9-10. Those with one point for criminal risk and seven points for background risk—meaning young males who did not own a home, an automobile, or a land line and who were-Unemployed or underemployed, or poorly educated—were detained at disposition 53 percent of the time. Id. At two points of criminal risk, those witb-no background risk points were detained until case disposition 33 percent of the time; those with seven background risk points were detained until case disposition 74 percent of the time. Id. at 7—1:10—13.

. 4. Arrestees Detained “Because of’ Indigence

The defendants argue that the plaintiffs’ statistical reports do not prove that large numbers of misdemeanor arrestees are detained solely because of indigence and' that the plaintiffs are assuming that if those detained could pay for release, they would. (Docket Entry No. 162 at 15-16); Hearing Tr. 8-2:53. Both parties’ experts tried to discern from Harris County data whether *1115and to what extent misdemeanor defendants are detained because they cannot pay a secured money bail. Dr. Demuth relied on a computer program the plaintiffs developed that took “snapshots” of the data on the Harris County Jail’s misdemeanor population at particular times on particular dates, pulled each defendant’s public records from the County’s public-facing online interface, and excluded those with nonfinancial reasons for detention on misdemeanor charges, such as concurrent pending felony charges. Hearing Tr. 2-2:6-7; 7-1:38-39. The most recent series of snapshots showed that on average, between February 15, 2017 and March 14, 2017, every day in the Harris County Jail there were:

• 328 people charged only with misdemeanors.

• 240 people charged only with misdemeanors and not subject to formal holds, such as warrants from another jurisdiction.

• 154 people charged only with misdemeanors, not subject to holds, who had been in jail for 3 or more days.

• 126 people charged only with misdemeanors, not subject to holds, who had been in jail for 5 or more days.

• 84 people charged only with misdemeanors, not subject to holds, who had been in jail for 10 or more days.

Pls. Ex. 4(d), Second Rebuttal Report at 9. The plaintiffs contend that at the very least, the 154 people in the County Jail every day who have been detained for three days or more on misdemeanor charges and are not subject to other holds have been found eligible for pretrial release and would be released if they paid the secured money bail.64 (Docket Entry No. 145 at 6; 146 at 13; No. 188 at 11). Dr. Demuth credibly testified that only the arrestees’ inability to pay keeps them detained. 6-2:168, 180-81; 7-1:39.

The defendants’' cross-examination of Dr. Demuth demonstrated that in a handful of entries for February 15, 2017, the plaintiffs’ computer program had failed to capture 'the fact that a misdemeanor arrestee was also charged with a felony or was about to be released on bond. Id at 7-1:41-58. In some instances, these additional docket activities took place the same day as the snapshot and may have occurred hours after the snapshot captured the data. This would indicate that the program worked as designed, including that it captured data only for a particular point in time and did not track cases over time. Id. at 7-1:33-34, 43. In a few other instances, the program did not work as designed in that an entry was miscoded. Id. at 7-1:43, 45.

The defendants also demonstrated that certain entries in the “snapshot” included misdemeanor- arrestees who were detained for mental-health evaluations or had formal “holds,” such as flags indicating that the arrestee was subject to extradition to another jurisdiction. Id. at 7-1:67-74. The court finds, that the defendants’ focus on mental-health status and other holds is *1116misplaced. Article 16.22 of the Texas Code of Criminal Procedure permits magistrates—including the Harris County Hearing. Officers and County Judges—to collect information about, and order the assessment of, an arrestee’s mental-health status. But Article 16.22(d) clearly states that “[t]his article does not prevent the applicable court from, before, during, or after the collection of information regarding the defendant as described by this article: (1) releasing a mentally ill or mentally retarded defendant from custody on personal bond or surety bond....” Tex. Code Crim. Pro. art 16.22. Article 16.22 is the only legal basis the defendants identified to detain misdemeanor arrestees for mental-health evaluations. Hearing Tr. 8-2:16. Judge Jordan testified that while misdemeanor arrestees who are ordered to have a mental-health evaluation ordinarily are detained pending the evaluation, the only way to ensure detention is to order secured money bail and refuse to grant a personal bond, knowing that the arrestee cannot pay the secured bail. Id. at 3-1:46-48. If the court ordered the arrestee evaluated but the arrestee had access to money, he or she could pay for prompt release, despite the evaluation order. Id. Misdemeanor- arrestees waiting for mental-health information to be collected or -evaluated are detained by secured money bail because they cannot pay.

As for “holds,” the plaintiffs offered unrebutted testimony that misdemeanor arrestees subject to holds, such as immigration detainers or pending warrants in other counties, are released “to their holds” only when they have either posted bond or disposed of the misdemeanor case. Id. at 2-2:30-32; 4-1:154-57. For instance, if an arrestee has a warrant pending in a neighboring county,, that county has ten days to take custody of the arrestee. But' the ten days do not begin to run until the arrestee has either paid the secured money bail set in the misdemeanor case, been granted a personal bond, or resolved that case by pleading guilty, being convicted, or having the charges dismissed. Id. Misdemeanor arrestees who have secured money bail imposed for their misdemeanor charges are detained in Harris- County not because of the hold, which they are legally unable to address, but because they are unable to pay the secured money bail.

Excluding bail-as-detention-orders for mental-health evaluations and holds that are irrelevant to this case (because they do not prevent release for a defendant who can pay the secured money bail), the defendants have shown that Dr. Demuth’s estimated average of those detained because they are unable to pay is inflated at most by a dozen entries in each category of the “snapshot.” On the present record, the court finds and concludes that more than 100 individuals are detained in the Harris County Jail each day, who have judicially been found eligible for release and who would be released but for their inability to pay secured money bail.

The defendant’s expert, Dr. Morris, attempted a different method of counting who was detained in the Harris County Jail solely due to indigence. Def. Ex. 28. Dr. Morris drew on the Pretrial Services risk assessments for all interviewed misdemeanor defendants from January 1, 2015 to February 14,2017, a total of 92,941 risk-assessment reports. Id. at 10. He excluded those with prior arrests or higher risk scores, because “[tjhose who have more of a criminal history are of a higher risk to have some unmeasured legal factor delaying release.” Id. Dr. Morris concluded that over the nearly 26-month period, no defendants who had all five indicators- of indigence tracked by Pretrial Services—no employment, no car, no land line phone, no *1117high school education, and no family residence—were detained solely by inability to pay. Id. He found only 65 detained individuals who had one Pretrial Services resource factor of indigence who were low risk, had no other reasons for detention, were eligible for release on a secured bond, but had not paid the bond and been released. Id.

Dr. Morris’s study is critically flawed in at least two ways. It first adopts the defendants’ mistaken outlook that .Texas law allows misdemeanor-only defendants to be detained before trial. See id. (“some unmeasured legal factor delaying release”). With a narrow exception for certain family violence cases, Harris County uses no other mechanism to detain misdemeanor defendants before trial than by imposing secured money bail. By excluding defendants with prior arrests or high-risk scores from consideration, Dr. Morris excluded a significant population of misdemeanor arres-tees who were judicially deemed eligible for pretrial release and would have been released if they could have paid the upfront amount needed under the secured money bail set.65

An even more basic flaw in Dr. Morris’s study was his exclusion of all misdemeanor defendants who had “moderate” or “high” risk scores from the population he considered. As explained above, Pretrial Services current risk-assessment tool counts resource factors such as the lack of a land line phone or an automobile as the same type of risk points as prior convictions or failures to appear.66 A misdemeanor defendant with no criminal history who met all of the poverty indicators would have at least five risk points—for not having a car, a family residence, a land line phone, a high school diploma, and for being unemployed or underemployed—and up to seven points if the defendant were a young male. See generally Pls. Ex. 8(d). But Dr. Morris excluded these defendants from his survey.

In sum, Dr. Morris excluded indigent defendants' from his survey to conclude that, of the misdemeanor defendants surveyed, none was detained because of indigence. Dr. Morris’s conclusion is not entitled to any weight. These critical flaws undermine his credibility and diminishes the court’s confidence in the reliability of the opinions he expressed, whether deriving from his own research or criticizing the analytic methods and conclusions of others.

. In his supplemental report, Dr. Morris ran his calculations including those with low-moderate and moderate risk scores.' Def. Ex. 28, Supplemental Report at 3. Dr. Morris’s attempt to salvage his report is not successful. He again excluded “high risk” defendants, which automatically excludes many young misdemeanor defendants who have all five poverty indicators on the Pretrial Services current risk-assessment form. - His exclusion of defendants with prior arrests, mental-health evaluations, - or assault charges again assumes that people are being detained for those reasons when the only mechanism under Texas law to detain them is to impose secured money bail that they are unable to pay. Even with all of these exclusions, Dr. Morris found that 1,623 people with- at least one poverty, indicator were detained in Harris County solely because of their inability to pay the secured bail imposed. Id.

5. Bond Forfeitures and Re-Arrests for New Criminal Activity

Harris County does not track’ the comparative failure-to-appear or new-criminal-activity rates of misdemeanor defendants *1118released on different types of bonds. Pis. Ex. 4(d), Second Rebuttal Report at 11; Hearing Tr. 3-2:146; 4-1:88; 5:138; 6-1:127. Harris County has not coded, collected, or analyzed data on the' different types of pretrial misconduct. It cannot, as other jurisdictions have, determine whether new misconduct by those released on surety bond ,or on personal bond is violent or is the type of nonviolent offense for which release on unsecured personal bond is presumed. See'id, The defendants’ expert, Dr. Morris, agreed that “it’s a shame we don’t have good data on court appearance.” Hearing Tr. 6-2:50. Dr. VanNostrand noted that Harris County does not currently compile the data to know how many defendants fail to appear for hear'ings when released on different types of bonds. The County will have to compile data on failures to appear as part of the Arnold Tool’s risk assessment. Id. at 6-1:110-12, 127-28. But for now, the County is imposing secured money bail, usually at prescheduled amounts, for almost all misdemeanor defendants, with no ability to tell how effective this, type of bond is to prevent failures to appear or new criminal activity compared to release on unsecured or nonfinancial conditions.

Harris County does keep, and was able to produce, data coded as “bond forfeiture,” “bond revocation,” and “bond surrender.” But this data is not consistently kept or recorded. See Pls. Ex. 4(d), Second Rebuttal Report at 11. Some 'County Judges “forfeit” a bond after a single failure to appear. Others reset hearings and do not record a bond as forfeited until after multiple failures to appear. A single entry in the “forfeiture” data may mean one failure to appear or many. Hearing Tr. 3-1:105, A bond may be revoked because a defendant failed a drug test, even if the defendant appeared at every court setting and is never arrested or charged with another offense, or revoked because the defendant failed to appear; Id. at 3-1:105; 3-2:148, 154. Similarly, one “revocation” entry may indicate one failure to appear, many, or none at all, and may or may not indicate new criminal activity. Commercial sureties can ask for bond surrender for a variety of reasons. Judges may rely on a variety of factors to grant or deny the request. Hearing Tr. 5:132-33.

The parties’ experts nonetheless tried to compare the “failure” rates of misdemean- or defendants released on different kinds of bonds. Dr. Demuth treated all coded forfeitures, revocations, and Surrenders as a general proxy for pretrial misconduct, without distinguishing between failures to appear or new criminal activity. Pls. Ex. 4(d), Second Rebuttal Report at 12; Hearing Tr. 6-2:157-58. Dr. Morris apparently examined coded forfeitures as a straightforward proxy for failures to appear. Def. Ex. 28A at 11. Using this approach, Dr. Demuth calculated that those released at any stage in the pretrial process on a surety bond have a failure rate of 11.1 percent; those released on an unsecured personal bond have a failure rate of 13.7 percent; and those released on a cash bond have a failure rate of 5.9 percent. Pls. Ex. 4(d), Second Rebuttal Report at 12.

Dr. Demuth credibly explained that this comparison of these general populations is misleading, because it does not control for the fact that many released early in the arrest process on surety bonds are, because of their relatively greater access to money or credit, likely to be an inherently less risky population than those released later in the arrest process, whether on a surety or a personal bond. See id.; Hearing Tr. 6-2:157-58; see also id. at 6-1:113. Dr. Demuth tried to account for this difference by comparing misdemeanor defendants released on different types of bonds only after a probable cause hearing. That is, he considered and compared those who could *1119not afford to bond out right away on secured money bail, but who were-able to come up with the money to post bond at a later stage in the process. Id. Dr. Demuth found that among these populations with a more similar risk profile,-those released on surety bond have a failure rate of 14.4 percent, while those released on unsecured personal bond have a failure rate of 13.6 percent. Pls. Ex. 4(d), Second Rebuttal Report at 12, That is, even with Harris County’s incomplete data, those released on unsecured personal bond have slightly better pretrial success rates than those released on a commercial surety bond.

Dr. Morris also tried to control for the different risk profiles by rejecting a general comparison of populations and using a propensity score matching algorithm that “pairs” criminal defendants. who share background characteristics but who are released under .different conditions. See Def. Ex. 28A at 11-12. Using this method, Dr. Morris concluded that for female, misdemeanor arrestees, there was no difference in the pretrial performance between those released on surety bonds and those released on unsecured personal bonds. Id. For male defendants, those released on surety bonds had a failure rate of 14,0 percent, while those released on personal bonds had a failure rate of 16.2 percent.67 Def. Ex. 28A at 12; Hearing Tr. 6-2:43-44.

.Dr. Morris’s decision to disaggregate his findings by gender and provide no overall failure rates is puzzling, to say . the least. Dr. Morris’s earlier reports did not disaggregate by gender. His broader past work in the field of pretrial studies did not disaggregate by gender. None of the. studies that Dr. Morris seeks to rebut disaggregate by', gender. Harris County’s forthcoming reforms specifically aim to be gender-blind in their risk assessments and prescriptions. See Hearing Tr. 6-2:40-41; 6-1:78-79. Dr. Morris found identical failure rates among women. His decision to disaggregate his findings had the effect of inflating the slight difference in failure rates between secured and unsecured bonds among men and made it appear greater than the overall.rate of failure, which Dr, Morris did not provide.

On the credible, reliable evidence in the present record, the court finds and concludes that: (1) Harris County has not compiled the data it has to compare failure-to-appear or - new-criminal-activity rates by bond type among misdemeanor defendants during pretrial release; and (2) to the extent the information is available, it -shows that those released on personal bond have substantially similar—or even somewhat better—pretrial failure rates as those released on surety bonds.68 Secured money bail in Harris County does not meaningfully add to assuring misdemeanor *1120defendants’ appearance at healings or. absence of new criminal activity during pretrial release.

This finding is consistent with recent empirical work in other jurisdictions. According to the most recent and credible evidence, secured financial conditions of pretrial release do not outperform alternative nonfinancial or unsecured conditions of pretrial release in ensuring the appearance of misdemeanor defendants at hearings. See, e.g., Pls. Ex. 12(h), Arpit Gupta et al., The Heavy Costs of High Bail: Evidence from Judge Randomization, 45 J. Leg. Studies 471, 475 (2016) (“We find no evidence that money bail increases the probability of appearance.”). One landmark study examined appearance rates in Colorado, where courts presume that misdemeanor defendants should be released on unsecured bonds and, unlike Harris County, track comparative rates of pretrial failures to appear. This study found that unsecured appearance bonds are equally effective as secured money bail, at both assuring appearance at trial as well as law-abiding behavior before trial. Pls. Ex. 7(q), Ex. 2, Claire M.B. Brooker et al., The Jefferson County Bail Project: Impact Study Found Better Cost Effectiveness for Unsecured Recognizance Bonds Over Cash and Surety Bonds (Pretrial Justice Institute, June 2014).

In New York City, which holds bail-setting hearings every day from 9:00 a.m. to 1:00 a.m., see Pls. Ex. 17, two large charitable bail-fund programs have paid the secured money bail amounts in misdemeanor cases for thousands of defendants for years. See Pls. Ex. 7(u). None of those defendants has a financial incentive to return to court. Only the bail funds lose money if the arrestee fails to appear. But the bail funds have consistently achieved 95 to 96 percent appearance rates. Id. The bail funds achieve these rates of appearance through simple and relatively inexpensive supervision methods, like sending text message reminders of hearings to the misdemeanor defendants. Id.; Pls. Ex. 12(ss).

These studies are consistent with Harris County’s own data. Although Harris County does not track pretrial failures-to-appear or new criminal activity by secured versus unsecured conditions of release, the parties’ experts found only slight, if any, differences in pretrial failure rates between those released on secured money bail and those released on unsecured personal bonds.

The defendants rely on a single study comparing rates of failures to appear and new criminal activity for misdemeanor defendants: that of their expert, Dr. Morris.69 In a study that has not yet been published or completed the peer-review process, Dr. Morris compared failure-to-appear rates and rates of new criminal activity for misdemeanor defendants released on different *1121categories of bond in Dallas County, Texas in 2008. Def. Ex. 30; see also Def. Ex. 163 (2012 update). Dr. Morris found that those released on a commercial surety bond failed to appear 26.7 percent of the time and were charged with a new offense 26 percent of the time within 12 months of their initial arrest. Def. Ex. 30 at 7-8. Those released on personal bond failed to appear 39.6 percent of the time and were charged with a new offense 29.1 percent of the time within 12 months of arrest. Id. Those released on cash bond failed to appear 30.2 percent of the time and were charged with new offenses 13.7 percent of the time within the 12-month period. Id. Dr. Morris testified that the difference in recidivism was not statistically significant. Hearing Tr. 6-2:53.

The court finds that Dr. Morris’s study is entitled to substantially less weight than the published, peer-reviewed articles in the record that rigorously compare pretrial failure rates among misdemeanor arres-tees released on different categories of bond. Dr. Morris testified that, as is true of Harris County, Dallas County does not compile comparative data on failures to appear. Instead, Dallas County tracks “forfeitures,” which may include multiple failures per entry. Id. at 6-2:50, 55. Dr. Morris did not provide the court or opposing counsel with access to the underlying data tables his calculations generated. The reason he gave—to protect the peer-review process the article is still undergoing—does not take into account the availability of a confidentiality or protective order, or a partially sealed filing, to achieve this same protection. Id. at 6-2:67-69.

The plaintiffs offered reliable evidence that in Dallas County, only those posting commercial surety bonds may be released within the first 24 hours after arrest on misdemeanor charges. Commercial bondsmen use the time to offer secured bonds to the least risky defendants. Pls. Ex. 7(i). Although Dr. Morris used his proximity score matching algorithm to attempt to control for background risk factors, it js unclear without the underlying data whether or to what extent it is possible to control for the significant dissimilarities between the two populations created by the commercial sureties’ “head start” on selecting the least risky and most financially secure misdemeanor defendants for surety bonds. See Pls. Ex. 4(d), Rebuttal Report at 13-14. And unlike Harris County’s extensive Pretrial Services program, Dallas County provides almost no supervision and therefore no incentives or reminders to those released on personal bond. Hearing Tr. 6-2:53-54. In sum, the court finds that Dallas County’s procedures, and Dr. Morris’s study of them, do not offer an effective or reliable comparison to Harris County.

F. The Effects of Pretrial Detention on Misdemeanor Defendants Who Cannot Pay Secured Money Bail

Recent studies of bail systems in the United States have concluded that even brief pretrial detention because of inability to pay a financial condition of release increases the likelihood that misdemeanor defendants will commit future crimes or fail to appear at future court hearings. See, e.g., Pls. Ex. 12(c), Christopher T. Lowenkamp et al., The Hidden Costs of Pretrial Detention (Laura and John Arnold Foundation, Nov. 2013). A study co-authored by Dr. VanNostrand, who is helping Harris County reform its bail system, found that for misdemeanor defendants, even two to three days of pretrial detention correlated at statistically significant levels with recidivism. See id. at 26. Pretrial detention made it more likely that misdemeanor defendants would fail to appear at future hearings. See id. at 14. Other studies have *1122confirmed these findings and shown that the likelihood of recidivism and failure to appear correlates with the imposition of secured money bail, not with a particular bail amount. See Pls. Ex. 12(h), Gupta et al., supra, at 473; see also Heaton Study at 19-24; Pls; Ex; 12(g), Megan Stevenson, Distortion of Justice: How the Inability to Pay Bail Affects Case Outcomes (Working Paper, University of Pennsylvania, Nov. 2016).

The Heaton Study found that if, during the six years between 2008 to 2013, Harris County had given early release on unsecured personal bonds to the lowest-risk misdemeanor defendants—those .receiving secured bail amounts of $500 or less— 40,000 more people would have been released pretrial; nearly 6,00.0 convictions and 400,000 days in jail at County expense would have been avoided; those released would have committed 1,600 fewer felonies and 2,400 fewer misdemeanors in the eighteen months following pretrial release; and the County would have saved $20 million in supervision costs alone. See Heaton Study at 45-46. Sheriff Gonzalez credibly testified that the research showing the “crimi-nogenic” effects of even a short period of pretrial detention and the high public costs of extended detention is consistent with his own experience as a Harris County law-enforcement officer. Hearing Tr. 3-2:11, 14.

A growing literature examines empirical data on “cumulative disadvantage” in pretrial detention. “Cumulative disadvantage” is -a “sequence of undesirable events whereby the occurrence of earlier negative events increases .the odds of subsequent negative events." Hearing Tr, 6-1:66 (referencing Stephen Demuth, Racial and Ethnic Differences in Pretrial Release Decisions and Outcomes: A Comparison of Hispanic, Blade, and White Felony Arrestees, 41 Criminology 873 (2003)). “Bail exacerbates and perpetuates poverty because of course only people who cannot afford the bail assessed or to post a bond—people who are already poor—are held in custody pretrial. As a consequence, they often lose their jobs, may.- lose their housing, be forced to abandon their education, and likely are unable to make their child support payments,” Pls. Ex. 12(f), Lisa Foster, Office for Access to Justice, Remarks al the American Bar Association’s- 11th Annual Summit on Public Defense (Feb. 6, 2016); see also Pls. Ex. 12(a)(iii), Marie VanNostrand, Legal and Evidence-Based Practices: Applications of Legal Principles, Laws, and Research to the Field of Pretrial Services at 15 (National Institute of Corrections, Apr. 2007); (Docket Entry No. 182 at 7; No. 272 at 9).

Money-based pretrial systems exacerbate the racial disparities in pretrial detention and posttrial outcomes. See Pls. Ex. 12(mm), Cynthia E. Jones, “Give Us Free”: Addressing Racial Disparities in Bail Determinations, 16 Legislation & Pub. Pol’y 919 (2013). An ‘amicus filing by Harris County Commissioner Rodney Ellis and the NAACP Legal Defense and Educational Fund notes that African-Americans make up 18 percent' of Harris County’s adult population but 48 percent of the Harris County Jail’s adult population. (Docket Entry No. 272 at 8). A 2011 study found that in Harris County, 70 percent of white misdemeanor defendants obtain early pretrial release from detention, but only 52 percent of Latino misdemeanor defendants and 45 percent of African-American misdemeanor defendants do so. (Id.). The defendants did not dispute this data.

G. Comparisons to Other Jurisdictions

The parties supplied additional briefing comparing, when possible, Harris County’s pretrial misdemeanor system to systems *1123used in other jurisdictions. (Docket Entry Nos. 233, 237, 255). The parties agree that “relatively little attention has been paid to analyzing bail determinations in misdemeanor cases” across the country. (Docket Entry No. 233 at 1; see also No. 255 at *8-9). Most large urban centers appear to hold daily bail hearings, at which arrestees appear before judicial officers within 24 hours of arrest. These occur in New York City; Cook .County, Illinois; Maricopa County, Arizona; and Miami-Dade County, Florida. (See Docket Entry No. 237). Washington, D.C., holds daily bail hearings except on Sunday, when the courts are closed. Hearing Tr. 2-2:194. Some bail courts run 24 hours a day, (see Docket Entry No. 237 (Maricopa County)); others operate during business hours, (see id. (Cook County; Miami-Dade County)). New York City holds bail hearings between 9:00 a.m. and 1:00 a.m. every day. (Docket Entry No. 233 at 1).

The defendants note that all of these jurisdictions permit secured money bail. Some broadly permit preventive detention in misdemeanor cases, and all, according to the defendants, permit secured money bail to result in detaining defendants “who pose a risk to the community that cannot be mitigated by conditions or are likely to fail to appear.” (Docket Entry No. 255 at *2). The defendants claim that the plaintiffs seek to hold Harris County to an anomalous standard by eliminating secured money bail/ while even those jurisdictions that no longer impose secured money bail as a matter of.practice still, permit it as a matter of law. Hearing Tr. 5:41; (Docket Entry No. 166 at 13-14; No. 255).

As the court surveyed above, some jurisdictions permit secured money bail to result in pretrial detention,' but only when that satisfies the due process, required of actual detention orders.70 Although other jurisdictions have timetables and procedures similar to Harris County’s, the practical effect of the bail hearings and resulting orders is dramatically different. In Washington, D.C., only 1.5 percent of misdemeanor arrestees are. detained until case disposition and virtually none on secured money bail. Hearing Tr. 2-2:149, 154. In New York City, only 3 percent of misdemeanor .arrestees are detained until case disposition. (Docket Entry No. 233 at 2). Under New Jersey’s recent reforms, statewide, only 8.4 percent of arrestees across all charge categories—-including felonies— were detained until case disposition. Pis. Ex. 7(k) at 1. In Kentucky, statewide, 25 percent of both felony and misdemeanor arrestees are detained until case disposition. Pls. Ex. 12(m)(i). The defendants do not identify a jurisdiction that, like Harris County, detains over 40 percent of those charged only with misdemeanor offenses until their cases are resolved. See Pls. Ex. 10(c),. 2015 Pretrial Services Annual Report at 8. If there is an anomalous standard here, it is set by Harris .County.

The defendants argue, and presented witnesses who testified, that Harris County leads most other jurisdictions in the timeliness of its proceedings because of how quickly district attorneys file charges and make release on secured money bail available. (See Docket Entry No. 286 at 3-5); Hearing Tr. 2-1:34-36; 3-2:50; 5:12-13. Arresting officers consult with Assistant District Attorneys over a 24-hour hotline. The Assistant. District Attorneys decide whether to accept charges before an arrestee is even booked. Hearing Tr. 2-1:34-36. In other jurisdictions, no arrestees, whether they can make bail or not, are even given the opportunity of release until a next-day or subsequent arraignment hear*1124ing, at which district attorneys decide whether to accept charges. Id. at 5:12-13; (Docket Entry No. 255 at *3-4).

Harris County’s speed at processing charges is commendable. When paired with the automatic imposition of secured money bail, however, it exacerbates the wealth-based differential treatment between those able to pay a bondsman to purchase early release and those who cannot. Those who can pay secured bonds are released within hours of arrest. Those who cannot are detained for days or weeks and face intense pressures to accept a guilty plea to end their pretrial detentions.

H. Proposed Bail Reforms

Dr. Marie VanNostrand, who the County has retained as a consultant on reforming its pretrial processes, testified about the policy changes the County expects to implement between July 1, 2017 and March 2018. For ease of analysis, those changes can be divided into three groups: policy changes affecting the County’s risk assessment of misdemeanor defendants; policy changes to enhance the efficiency of the Harris County pretrial system; and policy changes that will affect the combined probable cause and bail-setting hearings.

1. Changes to Risk Assessment

The centerpiece of Harris County’s proposed bail reforms is the adoption of the Arnold Tool, a nationally validated risk-assessment tool that will replace the County’s current validated risk-assessment tool. While the current risk-assessment tool relies on seventeen indicators of risk, including “background risk factors” such as home and automobile ownership,71 the Arnold Tool uses only nine indicators of risk. Almost all relate to either past criminal history, past failure to appear, or the severity of the current charge. The only “background factor” is the defendant’s age at time of arrest.72 Def. Ex. 157. Instead of simply adding all indicators up into a single risk score, the Arnold Tool scales and weights the indicators and provides three different scores on a 1- to 6-point scale. The scores are for risk of failure to appear, risk of new criminal activity, and risk of violence. Id.; Hearing Tr. 6-1:86-87.

Unlike the County’s current risk-assessment tool, the Arnold Tool would not score a misdemeanor defendant as a “moderate” risk based on poverty indicators such as the defendant’s educational level, or the lack of a car or land line phone. Def. Ex. 157. Dr. VanNostrand characterized the County’s current risk-assessment tool as “resource-based.” By contrast, the Arnold Tool is a “risk-based” assessment. Hearing Tr. 6-1:17-19. The substantial research behind the Arnold Tool shows that relying on resource-based factors does not predict failure-to-appear rates or new criminal activity better than excluding those factors and relying instead on the Arnold Tool’s nine risk-based indicators. Id. at 6-1:77-78.

As a condition of using the Arnold Tool, local jurisdictions must agree that they will not change it or put it to unintended uses. Def. Ex. 62. Local jurisdictions cannot change the risk indicators or how they are scored. Id.; Hearing Tr. 6-1:51. What local jurisdictions can control is the consequences that attach to each risk level. *1125Whether a defendant with a low risk score is released without any supervision or with some supervisory conditions, or released on secured money bail, unsecured bail, or no financial conditions, is a matter of local policy. Whether to detain a defendant who has a high risk score, release that defendant on secured monejr bail, or release that defendant with unsecured or nonfi-nancial conditions but with demanding- supervisory conditions, such as GPS monitoring, is also a matter of local policy. Hearing Tr. 6-1:51-52.

At the time of the motion hearing, Harris County had not yet decided what outcomes Pretrial Services would recommend based on various risk scores. Id. In general, under the new system, low-risk defendants will be recommended for release on unsecured personal bonds well in advance of their probable cause hearings. High-risk defendants will not have bail set at all until the probable cause hearing. Id. at 6-1:52-53. What constitutes low or high risk is not yet defined. Id. at 6-1:52, 134. For moderate-risk misdemeanor arrestees, the -County plans to continue the current system of setting secured money bail at a scheduled amount when the arrestee is charged. Id. at 6-1:141-42. Moderate-risk defendants with access to funds will be able to pay the secured money bail and be released before the probable cause hearing. Moderate-risk defendants without the means to pay a secured money bail will be detained until the hearing. If no changes are made to the bail setting, that misdemeanor arrestee will be detained until case disposition. Id. County policymakers have also stated their intention to continue to set secured money bail as a condition of release for high-risk misdemeanor defendants. Id. at 6-1:136. After the probable cause hearing, even the highest-risk defendants with the means to do so will be able to purchase release. For those without .the means to pay, the secured financial condition will keep them detained, operating as a detention order in all but name and process. Id. at 6-1:71.

Like Harris County’s current risk-assessment tool, the Arnold Tool is designed only to inform Pretrial Services recommendations. At most, recommendations such as release on personal bonds for low-risk defendants will be presumed in certain cases but required in none. Id. at 6-1:53. If judicial officers decide to reject the Arnold Tool’s recommendations for release on personal bond—as they currently do in nearly 67 percent of all misdémeanor cases using the County’s current risk-assessment tool73—they will not be acting contrary to Harris County policy. The wealth-based disparities will continue, with no empirical basis to conclude that imposing secured money bail promotes better rates of appearance or of law-abiding behavior for those on pretrial release.

2. Changes to the System’s Efficiency

Unlike the County’s current risk-assessment tool, the Arnold Tool will not require a Pretrial Services interview for a bail recommendation. Id. at 6-1:123-24. Once an arrestee has been identified, Pretrial Services can pull any prior criminal record, assess the risk, score, and generate a release recommendation, in some cases even before the arrestee has been transported from the scene of the arrest to the City or County Jail. Id. at 6-1:124. Early presentment of low-risk defendants for a personal bond will no longer depend on the availability of Pretrial Services personnel. Instead, it will be automatic. Id. at 6—1:136— 37. Dr. VanNostrand estimated that early presentments on paper will allow Hearing Officers to release low-risk defendants on personal bonds within 4 hours of arrest. *1126This would be among the fastest processing speeds in the nation. Id. at 6—1:138—39.

Dr. VanNostrand testified that increasing the number of arrestees released at early presentment will reduce jail crowding and speed pretrial release determinations and hearings for other misdemeanor defendants.74 Id. at 6-1:145-46. The County has hired two additional Hearing Officers and will be hiring more Pretrial Services personnel to handle the expedited processing. Def. Ex. 58; Hearing Tr. 3-2:128-30, 6-1:130. For the past year, Harris County has been redesigning its technology infrastructure to support the more streamlined system and to integrate information sources that diverse County agencies rely on. Hearing Tr. 6-1:119, 125, 130. Dr. VanNostrand testified that the current Harris County system “is very paper transport heavy.” Information is entered and reentered by hand as cases pass from arresting officers to district attorneys, to court clerks, and to Pretrial Services officers. Hearing Tr. 6-1:139. Infrastructure changes expected by July 1 are designed to cut down on paper transport and centralize more processes online.

Because the .new system will aim to make early-release determinations based on a defendant’s record rather than on an interview, the current Pretrial Services requirement of having to contact references to verify a defendant’s self-reported financial, employment, or other circumstances will apparently be eliminated, at least for defendants determined to be low-risk.75 Id. at 6-1:139-40. Dr. VanNostrand testified that Pretrial Services will continue to interview misdemeanor defendants who are not released at early presentment. Hearing Tr. 6-1: 155-56. Instead of verifying references, however, Pretrial Services will obtain an -affidavit of indigence similar to the affidavit used now to determine eligibility for appointment of counsel at a defendant’s first appearance before a County Judge. Id. at 4-1:48-49; 6-1:136, 161. Under the new system, -the same affidavit of indigence will be used for both appointing counsel as well as setting, release conditions, including bail. Id._

The County is also building.a new-inmate processing center, scheduled to open in March 2018. (Docket Entry No. 166 at 19 n.23); Hearing Tr. 3-2:74. The new center will process arrestees both from the City of Houston and from Harris County. It is designed to avoid the current bottlenecks that -occur in transporting defendants from one Jail to the other. Hearing Tr. 3-2:71-75.

3. Changes to the Probable Cause Hearings

With the adoption of the Arnold Tool, the Harris County policymakers intend to *1127amend the County Rules of Court to issue a new money bail schedule. Def. Ex. 67. While the current bail schedule is calibrad ed to the defendant’s current charge and criminal history, the new schedule will likely be calibrated to the predictive risk scores generated by the Arnold Tool. Hearing Tr. 6-1:151-52. The amended bail schedule will call for release on unsecured personal bonds for defendants with low-risk scores. Id. Those with moderate-risk scores will have a secured money bail set pending the probable cause hearing. Pretrial Services will recommend conditions of supervision during release, which the Hearing Officers may impose at the hearing. Id. at 6-1:142-43, 151-55. Those with high-risk scores will have no bond .set pending the probable cause hearing, with more stringent conditions of supervised release and the possible imposition of secured money bail to be considered at the hearing. Id. at 6-1:143, 153-54.

On February 28, 2017, the Harris County Commissioners approved a pilot program to provide public defenders at the probable cause hearings before the Hearing . Officers. Def. Ex. 59. As presently conceived, the public defender will help misdemeanor arrestees raise objections and provide relevant information about their inability to pay money bail, without risking incriminating statements.76 Hearing Tr. 1:103, 145-46. The pilot program is expected to launch by July 1, 2017, the same date the Arnold Tool and new -bail schedule are planned to go into effect. Def. Ex. 59.

It is unclear whether having defense counsel at the probable cause and bail--setting hearing will significantly -change the procedures or outcomes. Hearing Officers do not conduct proceedings of record under Texas law. Hearing Tr. 1:145-46. The Hearing Officers do not have to state findings or conclusions on the record for review by another judicial officer. Sheriff Gonzalez and Judge Jordan testified that the practical sustainability and impact of having counsel at probable cause hearings is doubtful. Id. at 3-1:113-14, 3-2:22. Judge Jordan opined that until Hearing Officers stop treating the County Judges’ bail, schedule as a.requirement to be applied in almost all cases on a secured basis, having defense counsel at the bail-setting hearing is unlikely to make.a.meaningful difference in the availability of release on unsecured or nonfinancial conditions. Id. at 3-1:114. ,

4. Texas House Bill 3011 / Senate Bill 1338

Bills have been introduced in the Texas -Legislature proposing wide-ranging amendments to Article, 17 of the Texas Code of Criminal Procedure, which regulates bail. House Bill 3011, and its companion Senate Bill 1338, if enacted as proposed, would permit pretrial preventive detention if a magistrate .“determines by clear and convincing evidence that requiring bail and conditions of pretrial release are insufficient to ensure” the defendant’s appearance in court or the safety of the community; Pls. Ex. 16 at 1.

Proposed Article" 17.028 -would set a statewide standard- of 48 hours after arrest, within which a magistrate would have to make an initial release decision, considering “any credible information provided by the defendant.” Id. at 3. Magistrates would be required to impose “the least restrictive conditions - and the minimum amount or type of bail necessary to rea*1128sonably ensure” the defendant’s appearance and the safety of the community. Id. at 4. The legislation would forbid “re-quirting] a defendant to provide a monetary bail bond for the sole purpose of preventing the defendant’s pretrial release.” Id. A magistrate who denied pretrial release would have to issue a reasoned opinion with written findings within 24 hours of the decision, subject to review on appeal. Id.

Under proposed Article 17.034, defendants who are released and fail to appear must again be released on personal bond if they can show good cause for their failure to appear. Id. at 8. Even if they cannot show good cause, magistrates “must set the amount of bail at the minimum amount” need to assure reappearance. Id.

“As soon as practicable” after a magistrate denies release, but no later than ten days after the decision, all defendants still detained would be entitled to an adversarial, counseled bail-review hearing, at which they would be able to put on evidence, testify, and call witnesses. Id. at 8-10. The reviewing judge must find by clear and convincing evidence “that monetary bail and conditions of release are insufficient to reasonably ensure the defendant’s appearance in court as required or the safety of the community” for the pretrial detention to continue. Id. at 11. The judge would have to issue that finding in a reasoned opinion, subject to appellate review. Id. at 12.

Proposed Article 17.20 provides that in misdemeanor cases, “Notwithstanding a bail schedule or any standing order entered by a judge,” a sheriff or other jailer “after considering the defendant’s pretrial risk assessment, may ... take the bail of the defendant in accordance with [proposed] Article 17.028.” Id. at 14. Proposed Article 17.028 is the article that requires release on the least restrictive condition. See id. at 4. Article 17.20 would confirm that sheriffs have independent judgment and authority to release misdemeanor ar-restees on less restrictive conditions than provided by a secured money bail schedule. Cfi id. at 14-15 (providing that in felony cases, a sheriff must deliver an ar-restee to a court to make the pretrial-release decision under Article 17.028 and permitting the sheriff to take the defendant’s bail only if the court so orders).

If enacted as proposed, the legislation would likely address many of the plaintiffs’ concerns. Under the Code as revised, Hearing Officers would not be authorized to exclaim “based on your priors!" at a one-to-two-minute hearing before imposing secured money bail that the defendant cannot pay, resulting in that defendant’s detention. The Hearing Officers would have to find and state the clear and convincing evidence supporting the specific reasons why paying a nonrefundable premium to a bail bondsman is reasonably necessary to ensure that the misdemeanor defendant will appear and refrain from new criminal activity while on pretrial release, and why no less restrictive condition is reasonable. The written findings would be reviewed in an adversarial, counseled hearing before County Judges, who must also apply the exacting clear-and-convincing evidence standard. Notwithstanding the County’s bail schedule, the Sheriff would be authorized to assess a misdemeanor defendant’s risk and release the defendant on an unsecured personal bond, with or without enhanced conditions for supervision. Bail could not be set to achieve pretrial detention without following the procedures required for a valid detention order under State law.

The defendants argue that because the Texas legislature has proposed these changes, it is the only body that can make them. (Docket Entry No. 266 at 9); see *1129 also Hearing Tr. at 3-1:72-73. But nothing in the current Code of Criminal Procedure prevents Harris County from imposing the least restrictive conditions on pretrial release, deciding those conditions by clear and convincing evidence, or making written findings and issuing reasoned opinions at bail reviews. Nothing in Texas state law permits, much less requires, Harris County judges to impose secured money bail for the purpose of detaining those who cannot pay it. A sheriffs role in releasing misdemeanor defendants on bail is already provided for under the current version of Article 17.20.77 The proposed legislation represents an acknowledgment that at least some jurisdictions in Texas are imposing secured money bail to detain misdemeanor arrestees because they cannot pay it, without the process a detention order requires. The proposed legislation makes explicit the due process requirements for setting bail under Texas law and sets boundaries on the procedures and time frames to meet those requirements.

I. Conclusions on Findings of Fact

Historically, bail has served as a mechanism of release from pretrial detention. Recently, many jurisdictions have acknowledged and repudiated long-standing practices of imposing, whether by intent or indifference, secured money bail that misdemeanor defendants are clearly unable to pay, resulting in pretrial detention of defendants otherwise eligible for release. Encouraged in their reforms by the American Bar Association and the U.S. Department of Justice, among others, these jurisdictions have followed two approaches to reforming the use of secured money bail for misdemeanor defendants. Some take the approach that a secured financial condition cannot result in the pretrial detention of misdemeanor defendants who cannot pay it, and who are otherwise eligible to be released. Other jurisdictions permit secured financial conditions of release to result in detention only when the process due before imposing a pretrial preventive detention order is provided. This includes timely, counseled, adversarial hearings at which the defendant may present evidence and the judge must issue a reasoned opinion with written findings explaining why the secured financial condition of release is the only reasonable way to assure the defendant’s appearance at hearings and law-abiding behavior before trial. The first approach recognizes that releasing those who can pay while detaining those who cannot pay would violate the Equal Protection Clause. The second approach recognizes that when secured money bail functions as a detention order against an indigent de*1130fendant, procedural protections are required under the Due Process Clause.

Texas law does not provide for pretrial release on no financial 'conditions. Texas law permits Harris County’s Hearing Officers and County Judges to choose between making financial release conditions secured—requiring a misdemeanor defendant or a surety to pay the amount up front to be released from jail—or unsecured—allowing release with the bond coming due only if the defendant fails to appear at hearings and a magistrate' orders the bond forfeited. In setting the bail amount, whether secured or unsecured, Texas law requires Hearing Officers to consider five factors, including the defendant’s ability to pay, the charge, and community. safety. A federal court consent decree requires. Hearing Officers to make individualized assessments of each misdemeanor defendant’s case and adjust the scheduled bail amount or.release the defendant on unsecured or nonfinancial conditions.

Harris County Hearing Officers and County Judges follow a custom and practice of interpreting Texas law to use secured money bail set at prescheduled amounts to achieve pretrial detention of misdemeanor defendants who are too poor-to pay, when those defendants, would promptly be released if they could pay. Complying with the County Judges’ policy in the bail schedule and the County Rules of Court, Harris County Assistant District Attorneys apply secured bail amounts to the charging documents. The schedule is a mechanical calculation based on the charge and the defendant’s criminal history. Although Texas and federal law require the Hearing Officers and County Judges to make individualized adjustments to the scheduled bail amount and assess nonfi-nancial conditions of release based on each defendant’s circumstances, including inability to pay, the Harris County Hearing Officers and County Judges impose the scheduled bail amounts on a secured basis about 90 percent of the time. When the Hearing Officers do change the bail amount, it is often to conform the amount to what is in the bail schedule, if the Assistant District Attorneys have set it “incorrectly.” The Hearing Officers and County Judges deny release on unsecured bonds 90 percent of the time, including’in a high majority of cases in which Harris County Pretrial Services recommends release on unsecured or nonfinancial conditions based on a validated risk-assessment tool. When Hearing Officers and County Judges do - grant release on unsecured bonds, they do sb for reasons other than the defendant’s inability to pay the bail on a secured basis.

The Hearing Officers and County Judges follow this custom and practice despite their knowledge of, or deliberate indifference to, a misdemeanor defendant’s inability to pay bail on a secured basis and the fact that secured money bail functions as a pretrial detention order. The Hearing Officers follow , an unwritten custom and practice of denying release on unsecured bonds to all homeless defendants. Those arrested for crimes relating to poverty, such as petty theft, trespassing, and begging, as well as those whose risk scores are inflated by poverty indicators, such as the lack of a car, are denied release on unsecured financial conditions in the vast majority of cases, when it is obvious that pretrial detention will result.- Hearing Officers style their orders as findings of “probable cause for further detention,” when the only condition of further detention is . the misdemeanor defendant’s inability to pay secured money bail. Pls. Ex. 9.

As a result of this custom and practice, 40 percent of all Harris County misdemeanor arrestees every year áre detained *1131until case disposition. Most of those detained—around 85 percent—plead guilty at their first appearance before a County Judge. Reliable and ample record evidence shows that many abandon valid defense's and plead guilty in order to be released from detention by accepting a sentence of time served before trial. Those detained seven days following a bail-setting hearing are 26 percent more likely to be convicted, 43 percent more likely to be sentenced to jail, and, on average, have sentences twice as long as those released before trial.

Harris County is required by Texas and federal law to provide a probable cause and bail-setting hearing for those arrested on misdemeanor charges without a warrant within 24 hours of arrest. At the hearing, Hearing Officers are supposed to provide “a meaningful review of alternatives to pre-scheduled bail amounts.” Roberson Order at 1. Although Texas law requires Harris County to release misdemeanor defendants who have not had a hearing within 24 hours, over 20 percent of detained misdemeanor defendants wait longer than 24 hours for a hearing. In some, but not all, of these cases, the Hearing Officers determine probable 'cause in the defendant’s absence, but the Hearing Officers admit that they do not provide a meaningful bail setting in absentia. For those misdemeanor arrestees who are detained for significant periods by the City of Houston Police Department before they are transported to the Harris County Jail, or for those booked into the Harris County Jail on a Friday, the Next Business Day Setting before a County Judge will not occur, until after three or four days in pretrial detention.

The record shows that County Judges adjust bail amounts or grant unsecured personal bonds in fewer than 1 percent of the cases. Prosecutors routinely offer, and County Judges routinely accept, guilty pleas at first setting and sentence the misdemeanor defendants to time served, releasing them from detention within a day of-pleading guilty. Those who do not plead guilty remain detained until they have a lawyer who can file a motion to contest the charge or the bail setting and request a motion hearing. These hearings are generally held one or two weeks later. The record shows that the motion hearing is the first opportunity a misdemeanor defendant has to present evidence of inability to pay and to receive a reasoned opinion explaining the bail setting. Testimony from the defendants’ expert on Harris County court administration establishes that the Next Business Day Setting rule codifies, rather than alters, these customs and practices.

The court finds and concludes that Harris County has a custom and practice of using secured money bail to operate as de facto orders of detention in misdemeanor cases. Misdemeanor arrestees who can pay cash bail up front or .pay the up-front premium to a commercial surety are promptly released. Indigent arrestees who cannot afford to do so are detained, most of them until, case disposition. Because the County Judges know and acquiesce in this custom and practice in their legislative capacity as rulemakers, this consistent custom and practice, amounts to an official Harris County policy.

Harris County does not compile comparative data on failures to appear by release on different bond types. No Harris County policymaker or judicial officer has attempted to examine the relative pretrial success or failure rates of misdemeanor defendants released on secured money bail versus those released on unsecured bail. The reliable, credible evidence, in the record from other jurisdictions shows that release on secured financial conditions does not assure better rates of. appearance or of law-*1132abiding conduct before , trial compared to release on unsecured bonds or nonfinancial conditions of supervision. Harris County’s proxy data for failure-to-appear is consistent with these studies. The information Harris County does keep shows no significant difference in appearance rates between those released on secured money bail and those released on unsecured appearance bonds, when properly controlling for the differences in risk profiles of the population.

The reliable evidence in the present record shows no meaningful difference in pretrial failures to appear or arrests on new criminal activity between misdemeanor defendants released on secured bond and on unsecured financial conditions. But even a few days in pretrial detention on misdemeanor charges correlates with—and is causally related to—higher rates of failure to appear and new criminal activity during pretrial release and beyond. Misdemeanor pretrial detention is causally related to the snowballing effects of cumulative disadvantage that are especially pronounced and pervasive for those who are indigent and African-American or Latino.

Harris County commendably plans to revise its pretrial processes and bail schedule by July 1, 2017. The County proposes to provide early release on unsecured bonds to “low-risk” misdemeanor defendants and to hold “high-risk” defendants— regardless of ability to pay money bail— until the probable cause hearing. “Moderate-risk” defendants will be granted release on a secured money bail, if they can pay the scheduled amount. The County plans to implement the Arnold Risk-Assessment Tool and integrate its information technology systems to avoid the delays that booking procedures and Pretrial Services interviews create. But Harris County’s policymakers and judicial officers have made clear their intent to continue imposing secured money bail on “high-risk” and “moderate-risk” defendants, categories as yet undefined. Those who can, pay the secured money bail, no matter their level of risk, will be released. Those who cannot will remain detained.

Except in the narrow case of defendants charged with a crime of family violence after violating a previously imposed condition of release, Texas law does not permit orders of pretrial preventive detention. Proposed legislation would permit magistrates to order preventive detention in certain cases, but only with procedural safeguards, and would forbid the use of secured money bail to accomplish preventive detention based on inability to pay. But for now, Harris County effectively gets around the Texas prohibition on pretrial detention by imposing secured money bail against indigent misdemeanor defendants knowing that they cannot pay. Harris County has its own extra-legal system of pretrial preventive detention through secured money bail that operates on the basis of wealth, fit accomplishes this without providing the procedural safeguards typically required of pretrial preventive detention orders.

II. Conclusions of Law

A. The Legal Standards

To obtain a preliminary injunction, the plaintiffs must establish “(1) a substantial likelihood of success on the merits, (2) a substantial threat of irreparable injury if the injunction is not issued, (8) that the threatened injury if the injunction is denied outweighs any harm that will result if the injunction is granted, and (4) that the grant of an injunction will not disserve the public interest.” Janvey v. Alguire, 647 F.3d 585, 595 (5th Cir. 2011); Nichols v. Alcatel USA Inc., 532 F.3d 364, 372 (5th Cir. 2008). “[A]t the preliminary injunction stage, the procedures in the dis*1133trict court are less formal, and the district court may rely on otherwise inadmissible evidence, including hearsay evidence.” Sierra Club, Lone Star Chapter v. F.D.I.C., 992 F.2d 545, 551 (5th Cir. 1993).

Summary judgment is appropriate if “no genuine issue of material fact exists and the moving party is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a). Rule 56 “mandates the entry of summary judgment, after adequate time for discovery and upon motion, against a party who fails to make a showing sufficient to establish the existence of an element essential to that party’s case, and on which the party will bear the burden of proof at trial.” Little v. Liquid Air Corp., 37 F.3d 1069, 1075 (5th Cir. 1994) (en banc). “The movant bears the burden .of identifying those portions of the record it believes demonstrates the absence of a genuine issue of material fact.” Triple Tee Golf, Inc. v. Nike, Inc., 485 F.3d 253, 261 (5th Cir. 2007) (citing Celotex Corp. v. Catrett, 477 U.S. 317, 322-25, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986)). If the burden of proof at trial lies with the nonmoving party, the movant may satisfy its initial burden by “ ‘showing’—that is, pointing out to the district court—that there is an absence of evidence to support the nonmoving party’s case.” Celotex, 477 U.S. at 325, 106 S.Ct. 2548. While the party moving for summary judgment must demonstrate the absence of a genuine issue of material fact, it does not need to negate the elements of the nonmovant’s case. Boudreaux v. Swift Transp. Co., 402 F.3d 536, 540 (5th Cir. 2005) (citation omitted).

In deciding a motion for summary judgment, the court draws all reasonable inferences in the light most favorable to the nonmoving party. Connors v. Graves, 538 F.3d 373, 376 (5th Cir. 2008). When the moving party has met its Rule 56 burden, the nonmoving party must identify specific evidence in the record and articulate how that evidence supports that party’s claim. Baranowski v. Hart, 486 F.3d 112, 119 (5th Cir. 2007). “This burden is not satisfied with some metaphysical doubt as to the material facts, by unsubstantiated assertions, or by only a scintilla of evidence.” Little, 37 F.3d at 1075 (internal quotation marks and citations omitted). Factual controversies resolve in the nonmoving party’s favor, “but only when there is an actual controversy, that is, when both parties have submitted evidence of contradictory facts.” Id.

B. Likelihood of Success on the Merits

1. The Standard of Review

Federal courts “generally analyze the fairness of relations between the criminal defendant and the State under the Due Process Clause, while [they] approach the question whether the State has invidiously denied' one class of defendants a substantial benefit available to’ another class of defendants under the Equal Protection Clause.” Bearden v. Georgia, 461 U.S. 660, 665, 103 S.Ct. 2064, 76 L.Ed.2d 221 (1983). The Supreme Court has noted that in cases of detaining the indigent, “[d]ue process.and equal protection principles converge in the Court’s analysis.” Id. (citing Griffin v. Illinois, 351 U.S. 12, 17, 76 S.Ct. 585, 100 L.Ed. 891 (1956)).

Although the legal standards and analysis overlap, the plaintiffs present one claim that is more appropriately analyzed under equal protection principles, and another claim more appropriately analyzed under due process. First, the plaintiffs allege that Harris County maintains a “wealth-based detention system” by setting secured money bonds higher than indigent misdemean- or defendants can pay, creating de facto orders of detention. (Docket Entry No. 54 at 9; No. 92 at 19). These detention orders *1134operate only against the indigent, because defendants who receive the same or similar secured bail but who can pay the bond or bondsman’s premium can be promptly released, regardless of the risk of nonappearance or new criminal activity. (Docket Entry No. 54 ¶ 42). Second, the plaintiffs allege that Harris County delays or fails t'o provide procedural protections required for a meaningful bail review. They allege that misdemeanor arrestees are frequently detained for days or even weeks before they can obtain any meaningful review of their bail setting in a counseled, adversarial hearing with findings on the record. (Docket Entry No. 54 ¶¶ 51, 82, 104; No. 92 at 21). Many plead guilty to obtain release rather than wait for a bail review that may, but likely will not, result in release.

The threshold question is what standard of review applies under either equal protection or due process analysis—rational basis, strict scrutiny, or something in'between. •

a. Equal Protection

In-its Memorandum and Opinion on the defendants’motions to dismiss, the court reviewed the Supreme Court’s trilogy of cases, Williams v. Illinois, 399 U.S. 235, 90 S.Ct. 2018, 26 L.Ed.2d 586 (1970), Tate v. Short, 401 U.S. 395, 91 S.Ct. 668, 28 L.Ed.2d 130 (1971), and Bearden v. Georgia, 461 U.S. 660, 103 S.Ct. 2064, 76 L.Ed.2d 221 (1983), along with the Fifth Circuit’s panel and en banc decisions, in Pugh v. Rainwater, 557 F.2d 1189. (5th Cir. 1977), vacated at 572 F.2d 1053 (5th Cir. 1978) (en banc), .to conclude that “[t]he ‘careful inquiry’-the [Supreme] Court requires in this type of. case calls for a more demanding- review” than rational basis. ODonnell, 227 F.Supp.3d at 730, 2016 WL 7337549 at *15. The court invited briefing from'the parties on the standard of review. Id. at 759-60, 2016 WL 7337549 at *39.

The County argues that rational basis review is the appropriate standard because the Williams-Tate-Bearden line of cases is limited to detention for defendants who do not pay post-conviction fines, and does not extend to detention for those who do not pay secured pretrial-bail. (Docket Entry No. 162 at 12). The County relies on cases holding that wealth-based distinctions are subject only to rational basis review because “[generally speaking, an individual’s indigence does not make that individual a member of a suspect class for equal protection purposes.” Driggers v. Cruz, 740. F.3d 333, 337 (5th Cir. 2014) (citing Maher v. Roe, 432 U.S. 464, 97 S.Ct. 2376, 53 L.Ed.2d 484 (1977)); see also San Antonio Indep. Sch. Dist. v. Rodriguez, 411 U.S. 1, 93 S.Ct. 1278, 36 L.Ed.2d 16 (1973); Carson v. Johnson, 112 F.3d 818, 821-22 (5th Cir. 1997) (“neither prisoners nor indigents constitute a suspect class”).

The Supreme Court and Fifth Circuit cases reviewed at the dismissal stage make clear that detention based on wealth' is an exception to the general rule- that rational basis review applies to wealth-based classifications. In Williams, the Supreme Court ruled that “[o]nce the State' has defined the outer limits of incarceration necessary to satisfy its penological interests and policies, it may not then subjects certain class of convicted defendants to a period of imprisonment beyond the statutory maximum solely by reason of their indigency.” 399 U.S. at 241-42, 90 S.Ct. 2018. In Tate, the Court extended the rule, holding that

the -same constitutional defect condemned in Williams also inheres in jailing an indigent for failing to make immediate payment of any fine, whether or not the fine is accompanied by a jail -term and whether or not the jail term of *1135the indigent extends beyond the maximum term that may be imposed on a person-willing and able to pay a fine. In each case, the Constitution prohibits the State from imposing a fíne'as a sentence and then automatically converting it into a jail term solely because the defendant is indigent and cannot forthwith pay the fine in full.

401 U.S. at 398, 91 S.Ct. 668 (quoting Morris v. Schoonfield, 399 U.S. 508, 509, 90 S.Ct. 2232, 26 L.Ed.2d 773 (1970) (plurality)). The Bearden Court reaffirmed that “if the State determines a . fine or restitution to be the appropriate and adequate penalty for the crime, it may not thereafter imprison a person solely because he lacked the resources to pay it.” 461 U.S. at 667-68, 103 S.Ct. 2064. Be ar-den made finding the least-restrictive alternative a constitutional requirement in cases in which inability to pay a fine results in imprisonment “[T]he court must consider alternative measures of punishment other than imprisonment. Only if alternative measures ■ are not adequate to meet the State’s interests in punishment and deterrence may the court imprison a probationer who has made sufficient bona fide efforts to pay.” Id. at 672, 103 S.Ct. 2064; see also Frazier v. Jordan, 457 F.2d 726 (5th Cir. 1972) (invalidating a law requiring certain defendants to choose between “a $17 fine or 13 days in jail” because it created two disparately -treated classes defined by wealth without a compelling state interest justifying the practice).

When the Supreme Court ruled in San Antonio Indep. School District v. Rodriguez that wealth-based classifications^ ordinarily require rational basis review, the Court specifically excepted the wealth-based detentions at issue in Williams and Tate. 411 U.S. at 20, 93 S.Ct. 1278. The Court recognized that in Williams and Tate, “[t]he individuals,' or groups of individuals, who constituted the class discriminated against ... shared two distinguishing characteristics: because of their impecunity they were completely unable to pay for some desired benefit, and as a consequence, they sustained an absolute deprivation of a meaningful opportunity to enjoy that benefit.” Id. The Williams-Tate exception did not apply to a case in which some could afford better schooling than others, but no one was completely cut off from public education by poverty. Id. at 25, 93 S.Ct. 1278. In that case, wealth classifications did not create a suspect class and rational basis review applied. Id. at 28-29, 93 S.Ct. 1278. But here, the plaintiffs’ claim is not that some are able to afford better conditions of pretrial release than others. The claim is that misdemeanor defendants who can pay secured money bail are able to purchase pretrial liberty, while those who are indigent and cannot pay are absolutely denied pretrial liberty and detained by their indigence. Under Williams, Tate, and Bearden, an absolute deprivation of liberty based on wealth creates a suspect classification deserving of heightened scrutiny.78

The ' defendants’ argument' that Williams, Tate, and Bearden are limited to detention for failure to pay post-conviction fines is' unpersuasive. Although state and local governments have compelling ihter-*1136ests in punishing and deterring violations of court orders, including the failure to pay court-ordered fines, the Supreme Court limits post-conviction detention of indigent defendants who cannot pay fines. The Court held that detention may be imposed only as a last resort, after a court carefully reviews the alternatives and makes findings on the record that detention is the least restrictive option. Bearden, 461 U.S. at 671-72, 103 S.Ct. 2064. By contrast, pretrial bail is not intended to be punitive. See, e.g., Brown, 338 P.3d at 1291 (“Bail is not pretrial punishment and is not be set solely on the basis of an accusation of a serious crime.”). The defendants have argued that the government’s interest in setting bail is to ensure that misdemeanor arrestees return for court appearances, not to protect public safety or to deter crime. (See Docket Entry No. 101 at 7,13). In the absence of a greater penological interest, and given the presumption of innocence for those awaiting trial, a government policy of wealth-based classifications for pretrial detention for misdemeanor offenses deserves, if anything, less deference than post-conviction detention.79

That conclusion is supported by the Fifth Circuit’s Rainwater decisions. The class plaintiffs in Pugh v. Rainwater alleged that Florida’s imposition of secured money bail without regard for an arres-tee’s ability to pay violated equal protection. 557 F.2d at 1190. The panel decision explicitly applied strict scrutiny, reasoning that “the [Supreme] Court has been extremely sensitive to classifications based on wealth, in the context of criminal prosecutions” and concluding from Williams-Tate that “the wealth classification in the instant ease warrants close judicial scrutiny” by creating a suspect class. 557 F.2d at 1197. The panel also stated that “[s]trict scrutiny is appropriate also because the inability to raise money bail necessarily affects fundamental rights of the indigent defendant. Foremost among these rights is the presumption of innocence.” Id. Pretrial detention based on inability to pay money bail also implicated and threatened “an accused’s right to a fair trial,” because “the ‘right to freedom before conviction permits the unhampered preparation of a defense.’ ” Id. (quoting Stack, 342 U.S. at 4, 72 S.Ct. 1).

The en banc court vacated the panel decision as moot because, while the appeal was pending, Florida issued a new written bail policy. 572 F.2d 1053. Although the en banc court did not comment on the scrutiny standard to be applied, it cited Williams and Tate, stating that “[a]t the outset we accept the principle that imprisonment solely because of indigent status is invidious discrimination and not constitutionally permissible.” Id. at 1056. The court viewed pretrial confinement “of one who is accused but not convicted of a crime as presenting a question having broader effects and constitutional implications than would appear from a rule stated solely for the protection of indigents.” Id.

Williams, Tate, Bearden, and Rainwater remain good law, neither overruled nor limited. The Supreme Court in San Antonio Indep. School District v. Rodriguez specifically excepted Williams and Tate from the general rule that wealth-classifications are reviewed under a rational basis standard. 411 U.S. at 20, 93 S.Ct. 1278. Bearden, decided a decade after San Antonio Indep. School District, confirmed that *1137using wealth-based classifications that result in detention for inability to pay a fíne “requires a careful inquiry into such factors as ‘the nature of the individual interest affected, the extent to which it is affected, the rationality of the connection between legislative means and purpose, [and] the existence of alternative means for effectuating the purpose.’ ” 461 U.S. at 666-67, 103 S.Ct. 2064 (quoting Williams, 399 U.S. at 260, 90 S.Ct. 2018 (Harlan, J., concurring)). The en banc court in Rainwater—also decided well after San Antonio Indep. School District—applied Williams and Tate to the pretrial bail context. Rainwater supports applying a standard of review more exacting than rational basis.80

At a minimum, heightened scrutiny requires a court to evaluate the government’s legitimate interest in a challenged policy or practice and then inquire whether there is a sufficient “fit” between the government’s means and ends. Cf. Harris v. Hahn, 827 F.3d 359, 365 (5th Cir. 2016) (“Classifications survive rational basis review ‘even when there is an imperfect fit between means and ends.’ ” (quoting Heller v. Doe by Doe, 509 U.S. 312, 321, 113 S.Ct. 2637, 125 L.Ed.2d 257 (1993))). At a maximum, “[classifications created by state action which disadvantage a ‘suspect class’ or impinge upon the exercise of a ‘fundamental right’ are subject to strict scrutiny, and will be upheld only when they are precisely tailored to serve a compelling state interest.” Clark v. Prichard, 812 F.2d 991, 995 (5th Cir. 1987).

State and local governments have “a compelling interest in assuring the presence at trial of persons charged with a crime.”81 Rainwater, 572 F.2d at 1056 (cit*1138ing Stack, 342 U.S. at 1, 72 S.Ct. 1). As a matter of law, Harris County has met -its burden to show a compelling state interest. The question is what level of tailoring heightened scrutiny requires in this case, and whether the plaintiffs have demonstrated a likelihood of showing that Harris County does not meet that standard. In Bearden, after the parties extensively argued about whether rational basis review or strict scrutiny applied, the Supreme Court cautioned that “[wjhether analyzed in terms of equal protection or due process, the issue [of detention based on indigence] cannot be resolved by resort to easy slogans or pigeonhole analysis, but rather requires a careful inquiry into such facts as. the nature of the individual interest affected, the extent to which it is affected, the rationality of the connection between legislative means and purpose, [and] the existence of alternative means for effectuating the purpose.” 461 U.S. at 666-67, 103 S.Ct. 2064 (internal footnotes, quotation marks, and citation omitted). The Rainwater panel interpreted Williams and Tate to require strict scrutiny in a challenge to a pretrial system of detaining indigent defendants because they could not pay secured money bail. 557 F.2d at 1197. The en banc court vacated on other grounds, without commenting on the scrutiny standard. See 572 F.2d at 1053.

The plaintiffs have suggested that “intermediate” scrutiny is the most conserva-five application of these precedents that recognizes both the government’s and the individual arrestee’s weighty interests. Hearing Tr. 1:73-75; see also Salerno, at 750-51, 107 S.Ct. 2095 (“On the other side of the scale, of course, is the individual's strong interest in liberty. We do- not minimize the importance and fundamental nature of this right.- But, as our cases hold, this right may, in circumstances where the government’s interest is sufficiently weighty, be subordinated tb the greater needs of society,”) “Narrow tailoring under intermediate scrutiny is different from strict scrutiny’s narrow-tailoring requirement. Strict scrutiny requires the government to show that it has used the least restrictive means of advancing a compelling interest.” Lauder, Inc. v. City of Houston, Texas, 751 F.Supp.2d 920, 933 (S.D. Tex. 2010), As- applied in free expression First Amendment case law, “the requirement of narrow tailoring is satisfied ‘so long as the ... regulation promotes a substantial government interest that would be achieved less effectively absent the regulation,” Id. (quoting Ward v. Rock Against Racism, 491 U.S. 781, 799, 109 S.Ct. 2746, 105 L.Ed.2d 661 (1989)).

In light of the plaintiffs’ burden to show a likelihood of success on the merits, the court applies the tailoring requirement of intermediate scrutiny. This standard is appropriately deferential towards the County *1139and appropriately protective of the misdemeanor defendants.

b. Due Process

In reviewing facial challenges to statutes regulating pretrial-confinement conditions, the Supreme Court evaluates whether “conditions and restrictions of pretrial detainment” impermissibly “amount to punishment of the detainee.” Bell, 441 U.S. at 533, 535, 99 S.Ct. 1861. The Court focuses “on whether the restrictions were imposed for a punitive purpose and, if not, on whether the restrictions are excessive in relation to a legitimate regulatory purpose.” Lopez-Valenzuela v. Arpaio, 770 F.3d 772, 778 (9th Cir. 2014) . (en banc) (collecting cases).

In United States v. Salerno, 481 U.S. at 739, 107 S.Ct. 2095, the Court reviewed a challenge to a provision of the federal Bail Reform Act of 1984 that permitted, pretrial detention of arrestees charged with serious felonies if the government demonstrated by clear and convincing evidence, at an adversarial hearing, that no release conditions would “reasonably assure” the safety of the community. 18 U.S.C. § 3142(e). Under Bell’s first prong, the Court found no evidence that Congress had intended pretrial detention to operate for a punitive purpose. 481 U.S. at 747, 107 S.Ct. 2095. Under Bell’s second prong, the Court upheld the preventive detention portion of the Act because it “carefully limits the circumstances under which detention may be sought to the most serious of crimes.” Id.

Having concluded that the challenged provision of the Bail Reform Act was regulatory and not punitive, the Court evaluated whether the Act’s procedures sufficiently protected “the individual’s strong interest in liberty.” Id. at 750, 107 S.Ct. 2095. The Court upheld the provision under this standard because: (1) “[d]etainees have a right to counsel at their detention hearing; (2) “[t]hey may testify in their own behalf, present information by proffer or otherwise, and cross-examine witnesses who appear1 at the hearing”; (3) the judicial officer “is guided by statutorily enumerated factors”; (4) “[t]he Government must prove its case by clear and convincing evidence”; and (5) “the judicial officer must include written findings of fact and a written statement of reasons'for a decision to detain.” Id. at 751, 107 S.Ct. 2095.

The Ninth Circuit, sitting en banc, interpreted Salerno to require strict scrutiny of pretrial detention conditions. Lopez-Valenzuela, 770 F.3d at 781 & n.3 (Salerno’s “heightened scrutiny” standard requires that pretrial detention policies be “narrowly tailored to serve a compelling state interest”). The Salerno Court itself did not describe its analysis as strict scrutiny or invoke the narrow-tailoring standard. But the Court did make clear that under the Due Process Clause, “liberty is the norm, and detention prior to trial or without trial is the carefully limited exception.” 481 U.S. at 755, 107 S.Ct. 2095. Salerno involved a facial challenge to' a federal law under the Due Process Clause, requiring a more demanding burden for the'plaintiffs than that involved here.82 Here, the plaintiffs’ due process challenge is to Harris County’s bail system as applied.

*11402. The Constitutional Requirements

a. Equal Protection

“The rule of Williams and Tate, then, is that the State cannot ‘impos[e] a fíne as a sentence and then automatically conver[t] it into a jail term solely because the defendant is indigent and cannot-forthwith pay the fine in full.’ ” Bearden, 461 U.S. at 667, 103 S.Ct. 2064 (quoting Tate, 401 U.S. at 398, 91 S.Ct. 668) (alterations in original). The Bearden Court concluded that while a state has broad discretion to decide what penalties satisfy its clear interest to deter and punish crime, once the state “determines a fine or restitution to be the appropriate and adequate penalty for the crime, it may not thereafter imprison a person solely because he lacked the resources to pay it,” unless a court finds either that (1) the defendant was not actually indigent and was refusing to pay in bad faith, or (2) “alternative measures are not adequate to meet the State’s interests in punishment and deterrence.” Id. at 667-68, 674, 103 S.Ct. 2064.

Applying Williams and Tate to the pretrial bail context, as Rainwater did, (and by extension, the post-Rainwater Bearden decision), the court concludes that Harris County has broad discretion to impose pretrial release conditions that meet the compelling interest of assuring a misdemeanor defendant’s appearance at trial. But once the County has chosen to impose a financial condition of pretrial release, the County may not use that condition to imprison defendants before trial because they lack the means to pay it. Rainwater, 572 F.2d at 1056. To do so impermissibly conditions “an absolute deprivation of a meaningful opportunity to enjoy [the] benefit” of liberty before trial or conviction on the basis of a. defendant’s poverty. San Antonio Indep. School District, 411 U.S. at 20, 93 S.Ct. 1278.

Under the Equal Protection Clause as applied in the Fifth Circuit, pretrial detention of indigent defendants who cannot pay a financial condition of release is permissible only if a court finds, based on evidence and in a reasoned opinion, either that the defendant is not indigent and is refusing to pay in bad faith, or that no less restrictive alternative can reasonably meet the government’s compelling interest. Bearden, 461 U.S. at 674, 103 S.Ct. 2064. In this case, the plaintiffs bear the burden of meeting the preliminary injunction requirements, but at the trial on the merits, the County will have the burden under heightened scrutiny to show that there is no reasonable alternative to a policy, custom, and practice of setting money bail on a secured basis in misdemeanor cases. See, e.g., Lauder, 751 F.Supp.2d at 933. The judicial defendants bear the burden to show that they make a finding of no reasonable alternative to imposing money bail on a secured, presched-uled basis for indigent defendants.

b. Due Process

In Turner v. Rogers, 564 U.S. 431, 131 S.Ct. 2507, 180 L.Ed.2d 452 (2011), the Supreme Court held that a state court’s detention order for civil contempt violated the Due Process Clause. Id. at 449, 131 S.Ct. 2507. The Court reasoned that while a civil contempt proceeding exposing the defendant to detention for up to one year did not require the assistance of counsel, the state had to provide “alternative procedural safeguards” such as “adequate notice of the importance of ability to pay [as an element to prove at the hearing], fair opportunity to present, and to dispute relevant information, and court findings.” Id. at 448, 131 S.Ct. 2507. The Court made clear that these were examples, not a complete description of what was needed for due process. The state could provide dif*1141ferent procedures “equivalent” to those the Court listed. Id.

Turner is a helpful starting point for-examining the plaintiffs’ likelihood of succeeding on their due process claim. Although the Supreme Court has not defined with precision the federal due process requirements for pretrial detention of misdemeanor defendants, at a minimum, state or local governments must provide notice of the importance of ability to pay in the judicial determination of detention, a fair opportunity to be heard and to present evidence on inability to pay; and a judicial finding on the record of ability to pay or a reasoned explanation of why detention is imposed despite an inability to pay the financial condition. Turner clarified that these procedures are required by the Due Process Clause even when the Sixth Amendment does not guarantee a right to counsel. Courts are divided over whether an initial bail-setting is a “critical stage” in the criminal process requiring counsel. See, e.g., Ditch v. Grace, 479 F.3d 249 (3rd Cir. 2007); Gonzalez v. Comm’r of Corr., 308 Conn. 463, 68 A.3d 624 (2013); Hurrell-Harring v. State, 15 N.Y.3d 8, 904 N.Y.S.2d 296, 930 N.E.2d 217 (2010); State v. Fann, 239 N.J.Super. 507, 571 A.2d 1023 (N.J.Super.L. 1990). Harris County does not currently provide counsel at the probable cause and bail-setting hearing but is exploring a pilot program to do so in July 2017.83

The defendants cite many cases for the proposition that “a bail setting is not constitutionally excessive merely because a defendant is financially unable to satisfy the requirement.” See, e.g., McConnell, 842 F.2d at 107. These cases in fact support the plaintiffs’ ’ due process claims. The cases the defendants cite involve serious felony charges with potentially lengthy sentences. The appellate courts affirmed'the imposition of secured money bail that a defendant could not pay. But the bail was imposed only after at least one counseled adversarial hearing, at which the defendant had an opportunity to present evidence and to be heard, with the court stating its findings on the record that either the defendant had not presented evidence of indigence or that no other condition could reasonably assure the defendant’s appearance at future hearings or protect the community from additional felony crimes.84

The. defendants cite only one case relating to detention on a misdemeanor charge, Fields v. Henry County, Tennessee, 701 F.3d 180 (6th Cir. 2012), But in Fields, the defendant could afford to pay money bail, and he was not detained because he was unwilling or unable to pay. See id. at 183 (the defendant was released on a $5,000 *1142bail bond). Instead, the defendant objected to Tennessee’s policy of detaining all those charged with family-violence offenses for 12 liours and the county’s policy of using a bail schedule. Id, at 184-85. Because the misdemeanor defendant failed “to point to any inherent problem. with the dollar amount set in his case,” the Sixth Circuit held that the bail schedule was not per se unconstitutional. Id, at 184 (“That is not. to say that using a bond schedule can never violate the Excessive Bail Clause.”).

The plaintiffs here do not challenge the bail schedules .as per se unconstitutional. See also Terrell v. City of El Paso, 481 F.Supp.2d 757, 766-67 (W.D. Tex. 2007) (use of a bail schedule not-inherently unconstitutional). Nor do the plaintiffs challenge the Texas statute allowing transparent pretrial detention orders in certain family-violence cases. Aside from this one case involving a misdemeanor defendant but not involving the same issues, the defendants rely exclusively on serious felony cases that permitted detention for failure to pay a financial condition only after a counseled, adversarial hearing with findings on-the record that no alternative to secured money bail could reasonably assure the defendant’s appearance given the potential for -a prison sentence of ten years to life and the resulting risk of flight.

Most importantly, in almost every case the defendants cite, the trial court could have—and sometimes did—order preventive detention, but ultimately set a secured financial condition with- the possibility of release as a less restrictive alternative to preventive detention.85- In Texas, however, pretrial preventive detention is-not available in misdemeanor cases except for those arrested on charges of family violence who have already violated a condition of pretrial release. See Tex, Const, art. 1 §§ 11b-11c.

The defendants argue that the Texas ban on preventive pretrial detention in most misdemeanor cases is not relevant because the plaintiffs are alleging violations only of federal law and have not pleaded state-law claims. (Docket Entry No. 266 at 3-5); Hearing Tr. 8-2:69-71. But federal due process protects state-created liberty interests. Liberty interests protected by the Due Process Clause “may arise from two sources—the Due Process Clause itself and the laws -of the States.” Ky. Dep't of Corr. v. Thompson, 490 U.S. 454, 460, 109 S.Ct. 1904, 104 L.Ed.2d 506 (1989) (quoting Hewitt v. Helms, 459 U.S. 460, 466, 103 S.Ct. 864, 74 L.Ed.2d 675 (1983)). The Supreme Court recognizes “that states- may, under certain circumstances,' create liberty interests which are protected by the Due Process Clause” and which entitle prisoners “to those minimum procedures appropriate under the circumstances and required by the Due Process Clause to insure that this state-created right is not arbitrarily abrogated.” Madison v. Parker, 104 F.3d 765, 767 (5th Cir. 1997) (citing Sandin v. Conner, 515 U.S. 472, 115 S.Ct. 2293, 132 L.Ed.2d 418 (1995); Wolff v. McDonnell, 418 U.S. 539, 557, 94 S.Ct. 2963, 41 L.Ed.2d 935 (1974)).

*1143“The Supreme Court has adopted a two-step analysis to examine ’ whether an individual’s procedural due process rights have been violated. The first question ‘asks whether there exists a liberty or property interest which has been interfered with by the State; the second examines whether the procedures attendant upon that deprivation were constitutionally sufficient.’ ” Meza v. Livingston, 607 F.3d 392, 399 (5th Cir. 2010) (quoting Ky. Dep’t of Corr., 490 U.S. at 460, 109 S.Ct. 1904). “State-law creates protected liberty interests only when (1) the state places substantive limitations on official conduct by using explicitly mandatory language in connection with requiring specific substantive predicates,” and (2) the state law requires a specific outcome if those substantive predicates are met.” Fields, 701 F.3d at 186, A “narrowly limited modicum of discretion” permitted to judicial officers does not deprive prisoners of a constitutionally protected right to be released. Teague v. Quarterman, 482 F.3d 769, 776 (5th Cir. 2007).

The Texas Constitution prohibits pretrial preventative detention orders in most misdemeanor cases. Tex Const. art. 1 §§ 11, 11b—11c; Ex parte Davis, 574 S.W.2d at 169. Texas has created a liberty interest in misdemeanor defendants’ release from custody before trial. Undér Texas law, judicial officers, as all parties admit, have no authority or discretion to order pretrial preventive detention in misdemeanor cases with a narrow exception for certain family-violence cases.86

To determine whether the procedures used sufficiently protect state-created liberty interests under the Due Process Clause, the Fifth Circuit applies the balancing test articulated in Mathews v. Eldridge, 424 U.S. 319, 96 S.Ct. 893, 47 L.Ed.2d 18 (1976). See Meza, 607 F.3d at 402. A federal court.must consider:

First, the private interest that will be affected by the official action; second, the risk of an erroneous deprivation of such interest through ' the procedure used, and the probable value, if any, of additional or substitute procedural safeguards; and finally, the Government’s interest, including the function involved and the fiscal and administrative burdens that the additional or substitute procedural requirement would entail.

Id. (quoting Mathews, 424 U.S. at 335, 96 S.Ct. 893).

, In this case, the private interest affected by Harris County’s policy is the misdemeanor defendant’s interest in release from custody before trial. That interest implicates fundamental constitutional guarantees: the presumption of innocence and the right to prepare for trial. See Salerno, 481 U.S. at 749-51, 107 S.Ct. 2095; Stack, 342 U.S. at 4, 72 S.Ct. 1; Rainwater, 572 F.2d at 1056-57. The record evidence shows that misdemeanor defendants in Harris County who are detained until case disposition are convicted at higher rates and given sentences twice as long as those released before trial.87 They plead guilty at rates much higher than those who are able to secure early release from pretrial detention.88 Detained misdemeanor defendants experience the multiplying effects of “cumulative disadvantages” when they lose jobs, places to *1144live, or family visitation rights because of pretrial detention.89

The risk of an erroneous deprivation of this liberty interest through the imposition of secured money bail is high. For the indigent, the risk of pretrial liberty deprivation because of the inability to pay secured money bail is certain. That deprivation is erroneous because the record evidence shows that secured money bail is not more effective at increasing the likelihood of appearance or law-abiding behavior before trial than release on an unsecured or nonfínancial condition. The record evidence shows that nearly 85 percent of those released in Harris County on an unsecured personal bond or other nonfínancial conditions do not forfeit their bonds for failing to appear or for committing new criminal activity.90 The rate is substantially the same as those released on secured money bail.91

As for the third factor, the defendants argue that alternatives to their system of detaining misdemeanor arrestees on secured financial conditions would be prohibitively expensive for the County. The defendants argue that adopting the Washington, D.C. system of releasing almost all misdemeanor arrestees before trial would cost the County tens or hundreds of millions of dollars. Def. Ex. 26 at 13-14; (Docket Entry No. 166 at 13-14). Mr. Banks testified that implementing the relief the plaintiffs seek would cost Harris County Pretrial Services $30 million annually. The current Harris County Pretrial Services budget is $7.5 million. Def. Ex. 46; Hearing Tr. 4-1:29-38.

The testimony relating to this argument is far from credible. Mr. Banks’s calculations assumed that this court would order every misdemeanor defendant released on personal bond, even if a defendant could pay a secured money bond. See Def. Ex. 46; Hearing Tr. 4-1:36-37. Mr. Banks not only assumed that the County would absorb the costs of supervising every arres-tee before trial, he also made the unwarranted assumption that 58.7 percent of those released would require GPS monitoring—the most restrictive form of supervision—and that over 10,000 arrestees—18.4 percent of all arrestees—would require alcohol-intake ignition locks, even though only about 6,000 arrestees—15 percent— are charged with misdemeanor driving-while-intoxicated offenses each year. Def. Ex. 46; Hearing Tr. 4-1:37-38; Pis. Ex. 10(c), 2015 Pretrial Services Annual Report at 12.

Mr. Banks greatly overstated the costs of pretrial supervision of misdemeanor defendants. Dr. VanNostrand’s work has shown that low-risk defendants require little to no supervision. Indeed, oversuper-vising misdemeanor defendants on pretrial release by, for example, subjecting them to frequent check-ins and drug tests, increases nonappearance rates. Pis. Ex. 12(j), Marie VanNostrand, Pretrial Risk Assessment in the Federal Court at 5 (U.S. Department of Justice, Apr. 2009). Dr. VanNostrand criticized Harris County Pretrial Services because of unnecessary—and unnecessarily costly—oversu-pervision. Hearing Tr. 6-1:155. Mr. Banks testified that the County likely oversuper-vises by automatically requiring drug tests of every defendant released on personal bond, even if the defendant’s misdemeanor charge is unrelated to drugs and his background shows no prior drug offenses. Id. at 3-2:143-44; 4-1:32. Neither Texas law nor the County Rules of Court require this approach.

*1145The credible evidence shows that, while Pretrial Services might incur some additional costs in supervising those who are now detained on a secured money bail they cannot pay, those costs are far less than the costs of detention. The issue is not added costs, but, more precisely, shifted costs. See, e.g., Pis. Ex. 12(kk), 12 (jj), 12(ww); Pis. Ex. 13(k); Heaton Study at 45—46^ Hearing Tr. 3-2:19. Mr. Banks estimated costs only for Harris County Pretrial Services. He did not estimate the costs the County would save by detaining far fewer people and for shorter periods. The contrary testimony relating to this factor is credible. Judge Morrison testified that most of the costs of the D.C. system arise from running a state-of-the-art drug-testing lab and paying all D.C. pretrial services officers at the federal salary pays-cale. Hearing Tr. 2-2:165-68. Neither is required for Harris County. The court concludes that the defendants’ testimony and evidence on the County’s costs of releasing misdemeanor defendants on alternatives to secured financial conditions is unreliable.

In Meza, the Fifth Circuit ruled that a parolee who had not been convicted of a sex offense had a Texas-created liberty interest in being free from requirements to register as a sex offender and to participate in sex-offender therapy. 607 F.3d at 401. Applying the Mathews balancing test, the court concluded that the parolee was owed “at least the same [due] process of an inmate, but as a parolee, he should generally be entitled to more favorable treatment than inmates.” Id., at 409. Applying Wolff v. McDonnell, 418 U.S. at 539, 94 S.Ct. 2963, on the process required to protect an inmate’s state-created liberty interests, the Fifth Circuit held that the parolee was owed “at a minimum: (1) written notice that sex offender conditions may be imposed as a condition of his mandatory supervision, (2) disclosure of the evidence being presented against [him] to enable him to marshal the facts asserted against him and prepare a defense, (3) a hearing at. which [the parolee] is permitted to be heard in person, present documentary evidence, and call witnesses, (4) an impartial decision maker, and (5) a written statement by the factfinder as to the evidence relied on and the reasons it attached sex offender conditions to his mandatory supervision.” (citing Wolff, 418 U.S. at 560-62, 94 S.Ct. 2963).

Under the federal case law defining due process for detention orders in general, as well as case law defining due process for state-created liberty interests, the court concludes that Harris County, in order to detain misdemeanor defendants unable to pay a secured financial condition of pretrial release, must, at a minimum, provide: (1) notice that the financial and other resource information its officers collect is for the purpose of determining the misdemeanor arrestee’s eligibility for release or detention; (2) a hearing at which the arrestee has an opportunity to be heard and to present evidence; (3) an impartial decisionmaker'; and (4) a written’ statement by the factfinder as to the evidence relied on to find that a secured financial condition is the only reasonable way tó assure the arrestee’s appearance at hearings and law-abiding behavior before trial.

The due process required for pretrial detention orders based on an indigent misdemeanor defendant’s failure to pay a secured financial condition of release is similar to the equal protection standard that prevents the government from converting financial conditions or penalties into detention orders without the following:- a hearing with, notice that pretrial liberty is at stake; with the opportunity to present evidence and to be heard; before a judge who must make findings on the record that *1146either the arrestee has the ability to pay' the amount needed for release, or that the government has no reasonable alternative to imposing detention for the failure to pay.

Due process also requires timely proceedings. In the context of misdemean- or arrests, pretrial detention of even three or four days can significantly increase the rates of nonappearance, recidivism, and the cumulative disadvantages of lost employment, leases, and family custody rights.92 Due process protections are meaningless if they are provided only after defendants effectively serve their sentences.

Texas and federal law provide guidance that due process requires the necessary hearing to be within 24 hours of arrest in misdemeanor cases. Texas law requires that a misdemeanor defendant arrested without a warrant must be released “not later than the 24th hour after the person’s arrest” if a probable cause hearing has not been provided. Tex. Code Crim. Pro. art. 17.033(a). “If the person is unable to obtain a surety for the bond or unable to deposit money in the amount of the bond, the person must be released on personal bond.” Id. In Sanders, the federal district court applied the 24-hour standard to setting bail in the City of Houston. 543 F.Supp. at 704. In the Roberson order, the federal district court required “a meaningful review of alternatives to pre-scheduled bail amounts” to be held within 24 hours from arrest. Agreed Final Judgment, No. 84-2974 at 1.

The defendants argue that evidentiary hearings with findings on the record are generally not possible within 24 hours because the available “information is necessarily lirhited” when the bail-setting hearings occur. They also argue that evi-dentiary hearings are not required because Gerstein permits jurisdictions to meet a less demanding due process standard in finding probable cause and in setting bail. (Docket Entry No. 166 at 16; No. 286 at 15). These arguments are unpersuasive. Sanders and Roberson were issued thirty years ago, before networked computing and com&unications technologies made it relatively fast and easy to transmit information. Those orders nonetheless set a 24-hour boundary on the time to complete ' the administrative incidents to arrest in misdemeanor cases in the City of Houston and in Harris County, Under Roberson, the County Judges are supposed to direct Pretrial Services “to make every effort to insure that sufficient information is available .., to determine an accused’s eligibility for a personal bond or alternatives to pre-scheduled bail amounts” .for a hearing to be held within 24 hours of arrest. Agreed Final Judgment, No. 84-2974 at 4.. Thirty years later, this 24-hour period is enough for Harris County to gather information on a misdemeanor defendant’s ability to pay secured money bail, com-pilé his or her criminal history and any other pending charges or holds, and make a finding as to" whether secured money bail or a less restrictive alternative is needed to meet the government’s interests.

As for Gerstein, the defendants' conflate' two separate parts of the Supreme Court’s opinion. The Court reasoned that “[because of its limited function and its nonad-versary character, the" probable cause determination is not a ‘critical stage’ in the prosecution that would require appointed counsel.” 420 U.S. at 123, 95 S.Ct. 854. Elsewhere, the Court noted that states are free to develop different pretrial processes. *1147Some states may choose “to make the probable cause determination at the suspect’s first appearance before a judicial officer, or the determination may be incorporated into the procedure for setting bail or fixing other conditions of pretrial release.” Id. at 123-24, 95 S.Ct. 854 (internal citations omitted). That does not mean, as the defendants appear to assume, that the minimal procedural protections for finding probable cause under the Fourth Amendment become the maximum procedures required for arraignments, bail-settings, or other proceedings a state chooses to combine with probable cause determinations. See id. at 125, 95 S.Ct. 854 n.27 (explaining that the majority opinion addressed due process only under the Fourth Amendment and that the “probable cause determination is in fact only the first stage of an elaborate system, unique in jurisprudence, designed to safeguard the rights of those accused of criminal conduct”),

Harris County may combine probable cause and bail-setting determinations in the same hearing. But the County must provide the procedures necessary both under the Fourth Amendment for the probable cause determination and under the Due Process and Equal Protection Clauses for setting bail and for ordering detention for indigent misdemeanor defendants unable to pay secured money bail.

c. Excessive Bail

As they did at the dismissal stage, the parties dispute whether this case is properly analyzed under the. Eighth Amendment’s prohibition on excessive bail. (See Docket Entry No. 101 at 18; No. 263). For the same reasons stated in its Memorandum and Opinion on the motions to dismiss, the court concludes that this is not an Eighth Amendment case. See ODonnell, 227 F.Supp.3d at 728, 2016 WL 7337549 at *13. As explained above, Texas law does not facially provide for release on no financial conditions. See Tex. Code Crim, Pro. arts. 17.01, 17.03. The requirement that magistrates consider five factors in setting the bail amount applies equally to secured and unsecured financial conditions Of release. See id. arts. 17.01, 17,15. The plaintiffs do not challenge the existence of Harris County’s bail schedule, the scheduled amounts, or the amounts the Hearing Officers and County Judges arrive at in applying the Texas-law factors. The plaintiffs do object to Harris County’s customs, practices, and policies bf setting money bail amounts on a secured basis for all but a few misdemeanor defendants, effectively detaining without due process those who would be released if they could páy, but who cannot and so are deprived of their pretrial liberty. These claims are not about the scheduled bail amounts in themselves. The claims are about the necessary procedures for requiring those amounts on a secured basis, the fact that those who can pay are promptly released, and the fact that those who cannot pay the secured bail suffer pretrial detention for their misdemeanor charges as a result.

The County Judges argue that the plaintiffs’ claims must be analyzed under the Eighth Amendment because when “a particular Amendment provides an explicit textual source of constitutional protection against a particular sort of government behavior, that Amendment, not the more generalized notion of ‘substantive due process,’ must be the guide for analyzing these claims.” Albright v. Oliver, 510 U.S. 266, 273, 114 S.Ct. 807, 127 L.Ed.2d 114 (1994). But the plaintiffs’ claims and the court’s conclusions do not rely on substantive due process. Williams, San Antonio Indep. School District,- and Rainwater make clear that detention based on wealth classifications triggers heightened scrutiny for suspect class discrimination under the Equal Protection Clause. See 399 U.S. at *1148242, 90 S.Ct. 2018; 411 U.S. at 21-22, 93 S.Ct. 1278; 572 F.2d at 1056. Salerno, McConnell and the cases on state-created liberty interests require procedural, not substantive, due process analysis. See 481 U.S. at 746, 107 S.Ct. 2095; 842 F.2d at 109 n.5; see also Mathews, 418 U.S. at 560-62.

Even if the plaintiffs were bringing an excessive bail claim, the analysis and outcome remain the same. Salerno and McConnell applied due process principles to analyze an Eighth Amendment claim that bail was excessive when it resulted in the automatic detention of a defendant who could not afford to pay. See generally 481 U.S. at 739, 107 S.Ct. 2095; 842 F.2d at 105. Rainwater applied equal protection principles to scrutinize a pretrial bail system that allegedly resulted in the system-wide detention of indigent arrestees. 572 F.2d at 1056-57. The defendants assume that if this is an Eighth Amendment case, the plaintiffs’ claims are defeated by McConnell’s reasoning that “a bail setting is not constitutionally excessive merely because a defendant is financially unable to satisfy the requirement.” 842 F.2d at 107; (see Docket Entry No. 101 at 18; No. 262 at 3). But the Eighth Amendment cases consistently hold that detention for failure to pay a financial assessment is permissible: (1) for dangerous felonies, in which the potential sentence ranges from ten years to life in prison to capital punishment; (2) after a judicial officer provides due process, including a counseled, adversarial, ev-identiary hearing with findings on the record and a reasoned opinion; (3) with a finding that no alternative to the secured financial condition can reasonably meet the government’s interests.93 To the extent they apply, the Eighth Amendment cases support the plaintiffs’ arguments. Nonetheless, these cases are not the basis of the claims or of the court’s findings- and conclusions.

3. Harris County Policies that Violate Constitutional Requirements

a. Municipal Liability under § 1983

A local government may be sued under § 1983 “when execution of a government’s policy .or custom, whether made by its lawmakers or by those whose edicts or acts may fairly be said to represent official policy, inflicts the [plaintiffs’] injury....” Monell v. New York City Dept. of Soc. Servs., 436 U.S. 658, 691, 98 S.Ct. 2018, 56 L.Ed.2d 611 (1978). Relief under § 1983 against a municipality requires “a plaintiff [to] show that (1) an official policy (2). promulgated by the municipal policymaker (3) was the moving force behind the violation of a constitutional right.” Peterson v. City of Fort Worth, 588 F.3d 838, 847 (5th Cir. 2009).

An official policy can be “a policy statement, ordinance, regulation, or decision officially adopted and promulgated by [the municipality’s] officers,” or a “governmental ‘custom’ even though such a custom has not received formal, approval through the body’s official decisionmaking channels.” Monell, 436 U.S. at 690-91, 98 S.Ct. 2018. “Official.municipal policy includes the decisions of a government’s lawmakers, the acts of its policymaking officials, and practices so persistent and widespread as to practically have the force of law.” Connick v. Thompson, 563 U.S. 51, 61, 131 S.Ct. 1350, 179 L.Ed.2d 417 (2011). “[A] municipal judge acting in his or her judicial capacity to enforce state law does not act as a municipal official or. lawmaker” for purposes of § 1983 liability. Johnson v. Moore, 958 F.2d 92, 94 (5th Cir. 1992). A municipality may be held liable for “deprivations resulting from the decisions of its duly constituted legislative body.” Bd. of *1149 Cty. Comm’rs of Bryan Cty. v. Brown, 520 U.S. 397, 403, 117 S.Ct. 1382, 137 L.Ed.2d 626 (1997). A claim against a municipal defendant in her official capacity is the equivalent of a claim against the municipality itself. See Kentucky v. Graham, 473 U.S. 159, 165, 105 S.Ct. 3099, 87 L.Ed.2d 114 (1985).

“Authority to make municipal policy may be granted directly by a legislative enactment or may be delegated by an official who possesses such authority, and of course, whether an official had final policy-making authority is a question of state law.” City of St. Louis v. Praprotnik, 485 U.S. 112, 124, 108 S.Ct. 915, 99 L.Ed.2d 107 (1988) (quoting Pembaur v. City of Cincinnati, 475 U.S. 469, 483, 106 S.Ct. 1292, 89 L.Ed.2d 452 (1986)) (internal quotation marks omitted). “[S]tate and local positive law, as well as ‘custom or usage’ having the force of law” determine whether a person is final policymaker. Jett v. Dallas Indep. Sch. Dist., 491 U.S. 701, 737, 109 S.Ct. 2702, 105 L.Ed.2d 598 (1989) (quoting Praprotnik, 485 U.S. at 124 n.1, 108 S.Ct. 915). “[Municipal liability under § 1983 attaches where—and only where— a deliberate choice to follow a course of action is made from among various alternatives by the official or officials responsible for establishing the final policy with respect to the subject matter in question.” Pembaur, 475 U.S. at 483, 106 S.Ct. 1292.

b. The County Judges’ Policies and Customs: Equal Protection

Harris County is not liable for the actions of the Hearing Officers or County Judges taken in a judicial capacity in adjudicating individual cases. See Johnson, 958 F.2d at 94. Nor is the County liable for policies that are set by the State of Texas and do not allow County officials to choose among alternatives. Pembaur, 475 U.S. at 483, 106 S.Ct. 1292. The County argues that it has no liability because the policies at issue are created by judges acting in their judicial capacities or are required by Texas law. (Docket Entry No. 266 at 9-11).

The record evidence, however, shows customs and practices, amounting to policy, that are neither created by judges in them judicial capacity nor mandated by Texas state law. By an uncodified policy and practice, the County does not permit misdemeanor arrestees to be released on unsecured personal bonds until references are verified.94 By unwritten policy and practice, Pretrial Services asks misdemeanor arrestees for information on their ability to pay without informing them that the purpose is to determine their eligibility for release on nonfinancial or unsecured financial conditions.95 Pretrial Services does not ask what bond arrestees are able to pay.96 None of these unwritten rules, customs, or practices is required by State law.

In its Memorandum and Opinion on the County’s motion to dismiss, the court ruled that the County can be liable under § 1983 for the policy choices made by the County Judges in their capacity as legislators and as administrative rulemakers. ODonnell, 227 F.Supp.3d at 745-46, 753-56, 2016 WL 7337549 at *27, 34-35. The County’s August 2016 letter changing the unwritten rule from requiring two verified references to requiring one verified reference of financial resources shows that the County Judges are final policymakers over this rule and that the policy is promulgated in their legislative or administrative capacity, not their judicial capacity in adjudicating *1150specific cases in their courts.97 See Def Ex. 52; ODonnell, 227 F.Supp.3d at 753 n.33, 2016 WL 7337549 at *33 n.33. The Roberson order—which ran against the County Judges only—required the County Judges to “direct the Pretrial Services Agency to make every effort to insure that sufficient information is available at the time of the hearings required herein for the Judicial Officer to determine ari accused’s eligibility for a personal bond or alternatives to pre-scheduled bail amounts.” .Agreed Final Judgment, No. 84-2974 at 4; Def Ex. 159. The Roberson order shows that the County Judges are final policymakers who, in an administrative capacity, direct Pretrial Services to gather misdemeanor arrestees’ financial information and present it to the Hearing Officers.

The plaintiffs allege that the County Judges promulgate an unwritten policy by knowingly acquiescing in and ratifying the Hearing Officers’ systemic custom and practice of setting money bail on a secured basis, following the bail schedule, without considering the misdemeanor arrestee’s inability to pay. (Docket Entry No. 54 ¶¶ 19, 56, 84-85, 103). The defendants argue that the Hearing Officers do consider ability to pay as one of the state-law factors -in setting bail. (Docket Entry No. 166 at 9, 17-18). But because the record evidence showed that- the Hearing Officers almost automatically set secured money bail at unpayable amounts in cases clearly involving indigent misdemeanor defendants, the plaintiffs argue that the County Judges promulgate an unwritten policy permitting Hearing Officers to use secured money bail as de facto pretrial detention orders, without providing due process and contrary to Texas’s ban on pretrial detention in all but one category of misdemeanor cases.98 (Docket Entry .No. 161 at 7; No. 189 at 4-5; No. 190 at 11),

The court finds and concludes on the present record that the plaintiffs have demonstrated a clear likelihood of success on the merits of their allegations. Based on the Pretrial Services monthly and annual public reports, the court finds and concludes that the County Judges know that Harris County detains over 40 percent of all misdemeanor defendants until the disposition of their cases.99 The County *1151Judges know that Hearing Officers deny Pretrial Services recommendations for release on unsecured and nonfinancial conditions around 67 percent of the time.100 They know that Hearing Officers deviate from the bail schedule—up or down—only about 10 percent of the time.101 The County Judges understand—because all but one of them share the same view—that what Hearing Officers mean when they Isay they “consider” an arrestee’s ability to pay is that they disregard inability to pay if any other factor in the arrestee’s background provides a purported basis to confirm the prescheduled bail amount and set it on a secured basis.102 Harris County’s Director of Pretrial Services testified that there is an “[u]nwritten custom” to deny all homeless arrestees release on unsecured or non-financial conditions. The County Judges know that Pretrial Services and the Hearing Officers treat homeless defendants’ risk of nonappearance as a basis to detain them on a secured financial condition of release they cannot pay.103 Hearing Tr. 4-1:43-44. The County Judges testified that they could change these customs and practices legislatively in their Rules of Court, but that they choose not "to. Hearing Tr, 5:49-50,150-51.

These legislative rulemaking choices are not required by Texas law. The Texas Code of Criminal Procedure makes ability to pay one of five factors to consider in setting the bail amount, but the Code does not require bail to be set on a secured basis and does not require that the five factors be used to decide whether to set bail on a secured basis. See Tux. Code Crim, Pro. art. 17.01, 17.15. The parties agreed that in County Court No. 16, Judge Jordan follows a different practice.104 Judge Jordan does not set bail on a secured basis if it would operate to detain an indigent misdemeanor defendant. If a defendant has the means to pay some bail on a secured basis, Judge Jordan considers the five - factors to set bail within • an amount the defendant can pay. If a defendant cannot pay a financial condition up front, Judge Jordan considers the five factors, sets the bail amount on an unsecured basis, and orders nonfinancial conditions of pretrial supervision to release the defendant while addressing the defendant’s risk of nonappearance or of new criminal activity.105 Judge Jordan’s judicial practice is consistent with Texas law and, when done timely, is consistent .with equal protection and due process. But as a legislative body that votes to enact policy by a two-thirds majority, the County Judges knowingly acquiesce in and ratify customs .and practices so consistent and widespread as to have the force of a policy. That policy is to detain misdemeanor defendants before trial who are otherwise eligible for release, but whose indigence makes them unable to pay secured financial conditions of release.

This policy is . not narrowly tailored to meet the County’s compelling interest in having misdemeanor defendants appear for hearings or refrain from new criminal activity before trial. Even applying the less stringent standard of intermediate scrutiny, the present record does not show that rates of court appearance or of law-abiding behavior before trial, would be lower absent the use of secured money bail against misdemeanor defendants. See Lauder, 751 F.Supp.2d at 933 (under intermediate scrutiny, “the requirement of narrow tailoring *1152is satisfied so long as the ... regulation promotes a substantial government interest that would be achieved less effectively absent the regulation”) (internal quotations marks and citation omitted). Recent rigorous, peer-reviewed studies have found no link between financial conditions of release and appearance at trial or law-abiding behavior before trial.106 Harris County policymakers have not attempted to collect, much less review, the County’s own data to determine whether secured financial conditions of release work better in Harris County than unsecured or nonfi-nancial conditions.107 That lack of inquiry is one indication the policy is not narrowly tailored. The other indication is that both parties’ experts evaluated Harris County’s data and found no significant difference in appearances at hearings or in new arrests between misdemeanor defendants released on secured money bail and those released on unsecured personal bonds.108

To be sure, requiring secured money bail for misdemeanor defendants does not run' afoul of equal protection principles when those defendants are actually released. If two defendants take advantage of similarly timed opportunities for pretrial release on secured money bail, the fact that it may be harder for one to come up with the money than the other does not create a suspect classification between the two and does not trigger heightened scrutiny. See San Antonio Indep. Sch. Dist., 411 U.S. at 23-24, 93 S.Ct. 1278. But when a secured financial condition of release works an absolute deprivation of pretrial liberty because a defendant is indigent or so impecunious that he or she cannot pay even a bondsman’s premium required for release, the County must show that requiring a secured money bail is at least more effective than a less restrictive alternative at meeting the County’s interests, even if it is not the least restrictive means to do so. See id. at 20-22, 93 S.Ct. 1278; Bearden, 461 U.S. at 672, 103 S.Ct. 2064.

Based on the present record, the court finds and concludes that, as a matter of law, Harris County cannot make this showing. The cases in which the government is able to show no reasonable less restrictive alternative to detaining an indigent defendant by imposing a secured money bail all involve charges for serious felonies that carry lengthy potential sentences. The Harris County Criminal Courts at Law have jurisdiction only over misdemeanor cases. The plaintiffs were charged only with misdemeanor offenses and have no pending felony charges. Texas law forbids pretrial preventive detention of misdemeanor arrestees in all but one category of cases—those who are arrested on family violence charges and who have violated a prior family violence protective order while released before trial. In that narrow category, the State provides enhanced procedures to protect the defendant’s liberty interests. See Tex. Const, art. 1, §§ llb-llc; Tex. Code Ceim. Pro. art. 17.29-292. Outside that category, Texas law does not distinguish among misdemeanor arrestees in terms of their eligibility for pretrial release. Hearing Officers recognize . this approach whenever they permit release on secured money bail. A defendant who can pay is released regardless of risk. Once deemed eligible for release, indigent misdemeanor defendants who cannot pay the secured financial condition of release cannot be detained on that basis without a hearing and judicial findings on the record that no other reasonable alternative is available. In Harris *1153County misdemeanor cases, reasonable alternatives to continued detention are readily available for indigent defendants unable to pay a secured money, bail. Those alternatives include reducing the bail amount, as Judge Jordan does, imposing unsecured money bail, or releasing on non-fínancial conditions of pretrial supervision. Hearing Tr. 3-1:62-66.

Harris County is not liable for the individual adjudications of its Hearing Officers and County Judges in specific cases, even if those orders detain indigent arrestees because these cannot pay secured money bail. See Johnson, 958 F.2d at 94. But the County is hable for the legislative and administrative policies of its County Judges who knowingly or with reckless indifference acquiesce in and ratify a custom and practice that achieves pretrial preventive detention on secured financial conditions that defendants cannot pay in over 40 percent of all Harris County misdemeanor cases. See ODonnell, 227 F.Supp.3d at 745-46, 753-56, 2016 WL 7337549 at *27, 34-35. The court concludes that the plaintiffs are likely to succeed in proving that the County has a policy of violating equal protection by detaining indigent misdemeanor arrestees before trial.

c. The County Judges’ Policies and Customs: Due Process

Due process requires: (1) notice that the financial and other resource information Pretrial Services officers collect is for the purpose of determining a misdemeanor arrestee’s eligibility for release or detention; (2) a hearing at which the arres-tee has an opportunity to be heard and to present evidence; (3) an impartial decision-maker; (4) a written statement by the fact-finder as to the evidence relied on to find that a secured financial condition is the only reasonable way to assure the arres-tee’s appearance at hearings and law-abiding behavior before trial; and (5) timely proceedings within 24 hours of arrest.109 (Docket Entry No. 286. at 12-13,16).

The court concludes that the plaintiffs are likely to succeed on at least parts of their due process claim. Of the requirements listed above, Harris County meets only one at the probable cause and bail-setting hearing: an impartial decisionmaker. The County usually provides the hearing within 24 hours, but 20 percent of misdemeanor defendants who remain detained until the hearing wait longer than 24 hours for that hearing.110 The record evidence shows that misdemeanor defendants are sometimes confused about the financial and other resource information they are asked to provide and how it will affect their eligibility for release,111 and Hearing Officers do not make written findings or give reasons for their decisions.112

The rule requiring a Next Business Day Setting before a County Judge recently came into effect. See Rules of Court 4.3.1. Depending on the timing of arrest and booking, this first appearance may occur within 24 hours after arrest, but the record does not indicate how often that happens. Harris County’s former court administrator testified that the Next Business Day setting is not a rule change, but a codification of prior practice.113 The record shows that the practice is for County Judges to routinely deny reductions in the bail amount and to refuse release on unsecured financial conditions in more than 99 per*1154cent of cases.114 The record does not show written findings made by- County Judges-explaining why money bail must be imposed on a secured basis in any specific case.

Except for the Texas Code requirement-that misdemeanor defendants be released 24 hours after arrest if probable cause has not been found, the timing of County procedures is regulated by the County Judges’ Rules of Court promulgated by the County Judges in their legislative capacity. See Tex. Code Cbim. Pro. art. 17.033. The 'record evidence shows that thousands of .misdemeanor defendants each year are detained longer than 24 hours before they have a bail-setting hearing.115 Instead of releasing -defendants who have not had a probable -cause hearing within 24 hours, the County follows an unwritten policy of determining probable cause in absentia, using only the charging papers. The Harris'County judicial' officers agreed that bail is not meaningfully considered at .these in absentia hearings.116

' The court concludes that Harris County does not provide due process for indigent or impecunious misdemeanor defendants it detains for their inability to pay a secured financial condition of release. Those who cannot pay the secured money bail set at the probable cause hearing before a Hearing Officer must wait days, sometimes weeks, before a County Judge provides a' meaningful hearing to review the bail determination.117 Harris County is liable' for the County Judges’ policies issued in their legislative -or- rulemaking capacities that result in systemwide delays in any meaningful determination of the conditions for release. ' ■

’ If the County complied with equal protection requirements, part of the plaintiffs’ concerns about due process would be mitigated. If Hearing Officers, as they are supposed to do under the Roberson order, tailored nonfinancial ■ release conditions to address through supervision each-defendant’s risk of nonappearance or new criminal activity, and then released those defen-' dants, the need to present evidence and make written findings about financial conditions would be less urgent.-Hearing Officers do not need to issue reasoned opinions explaining them decision to detain someone using-secured money bail if the Officers cannot use secured- money -bail to detain indigent defendants-in the first place.

d. The Sheriffs Policies under Equal Protection and Due Process

“The sheriffs acquiescence in unsound and legally, insufficient procedures.- effectively create[s] a county, policy for which the county is liable” under § 1983. Doe v. Angelina County, 733 F.Supp. 245, 257 (E.D. Tex 1990). Whether a sheriffs deliberate indifference gives rise to liability for a municipal policy, including an unconstitutional custom or practice, is *1155determined by “an objective [standard]; it considers not only what the policymaker actually knew, but what he should have known, given the facts and circumstances surrounding the official policy and its impact on the plaintiffs rights.” De Luna v. Hidalgo County, 853 F.Supp.2d 623, 641 (S.D. Tex. 2012) (quoting Lawson v. Dallas County, 286 F.3d 257, 264 (5th Cir. 2002)); see also Dodds v. Logan County Sheriff’s Dept., Civil No. 8-333, 2009 WL 8747487 (W.D. Okla. Aug. 3, 2009) (the sheriff was liable for his “deliberate indifference to the due process rights of arrestees whose bail had been pré-set” by acquiescing in a policy set by the local judges); Blumel v. Mylander, 954 F.Supp. 1547, 1557 (M.D. Fla. 1997) (a sheriff and jailer were liable for violating the' right to pretrial release after 48 hours from arrést with no probable cause finding when they were “actually and constructively aware” that the 48-hour requirement had been exceeded).

In its Memorandum and Opinion on the County’s motion to dismiss, the court held that under Fifth Circuit case law, a Texas county may be liable for its sheriffs policies of detaining arrestees and enforcing orders the sheriff knows or should reasonably know are unconstitutional. ODonnell, 227 F.Supp.3d at 748-50, 2016 WL 7337549 at *30-31. At the hearing on the plaintiffs’ application for a preliminary injunction, the Harris County Sheriff testified that he knows that every day, misdemeanor arrestees who would be released if they could pay a secured financial condition of release are detained in the Harris County Jail solely because poverty prevents them from paying. Hearing Tr. 3-2:8-9, 18-19, 22-24. A major from the Sheriffs Office testified about the delays in presenting arrestees at their probable cause hearings and confirmed that in many eases of arrest by the City of Houston Police Department, the Harris County Sheriff may not even take custody of ar-restees within 24 hours and does not present those arrestees at probable cause and bail-setting hearings within 24 hours. Id. at 3-2:64, 67-71, 83. In his declaration, Sheriff Gonzalez stated that “[individuals should not be held in our Harris County jail just because they cannot pay an amount of money set according' to an arbitrary schedule. In my view, this practice violates the U.S. Constitution.” Pis. Ex. 7(r) at 2.

The Sheriffs detention of misdemeanor deféndants while knowing; (1) that the misdemeanor defendants are detained because their indigence prevents them from paying secured money bail to obtain release, and (2) that this practice violates equal protection and due process principles, is a policy choice' the Sheriff makes on Harris County’s behalf. That policy is not narrowly tailored to meet the County’s compelling. interests in ensuring misdemeanor defendants’ court appearances and law-abiding conduct before trial. The plaintiffs have demonstrated a clear likelihood of success on the merits of their claim that the Sheriff, as a County policymaker, knowingly detains misdemeanor defendants on constitutionally invalid bases.

4. Judicial Conduct that Violates Constitutional Requirements

Section .1983 does not permit injunctive relief against judicial officers acting in a judicial capacity unless either: (1) they violate a declaratory decree; or (2) declaratory relief is unavailable. 42 U.S.C. § 1983. Declaratory relief is available in this case; preliminary injunctive relief is not.118 See, *1156 e.g., MacPherson v. Town of Southampton, 664 F.Supp.2d 208, 211-12 (E.D.N.Y. 2009) (“Plaintiffs cannot allege that declaratory relief is unavailable because Plaintiffs can, and indeed have, pursued a claim seeking a declaration”); Besaro Mobile Home Park, LLC v. City of Fremont, Civil No. 10-478, 2010 WL 2991592 at *2 (N.D. Cal. July 29, 2010) (declaratory relief is unavailable when as a matter of law no cause of action for declaratory relief is provided by statute). The record evidence shows that in individual adjudications, Harris County Hearing Officers and County Judges set secured financial conditions of release in order to detain misdemeanor defendants before trial. These de facto orders of pretrial preventive detention operate only against indigent misdemeanor defendants who are unable to pay the financial condition. The minimum due process protections required to issue a" pretrial detention order are not provided in these hearings. The plaintiffs have demonstrated a clear likelihood of success on their claims for declaratory relief, but a preliminary injunction against the judicial officers in their judicial capacity is not available.

5. Conclusion on Likelihood of Success on the Merits

Harris County is liable for the unconstitutional acts of the County Judges when they act as final policymakers in their legislative and administrative capacities. The County Judges are final policymakers who administratively direct Pretrial Services to gather information on misdemeanor arres-tees and to present the information to the Hearing Officers. As a legislative and administrative body, the County Judges sitting en banc knowingly acquiesce in and ratify customs and practices so consistent and widespread as to have the force of policy. These policies systematically detain misdemeanor defendants who are otherwise eligible for release before trial but whose indigence makes them unable to pay a secured financial condition of release. These de facto detention orders are not narrowly tailored to meet a compelling government interest. The evidence shows that secured financial conditions of release are. not more effective at meeting the County’s interests than unsecured or non-fínancial conditions of release in misdemeanor cases. Instead, secured money bail operates to detain the impoverished while releasing those able to pay. This liberty deprivation based on wealth violates the Equal Protection Clause.

Harris County does not provide misdemeanor defendants notice of the significance of the financial information they are asked to give in order to even be considered for release from pretrial detention on unsecured or nonfinancial conditions. Harris County does not provide timely hearings at which misdemeanor defendants can be heard, can present evidence of their inability to pay, or can receive reasoned opinions with written findings on why a secured financial condition of release, and not a less restrictive condition, is the only reasonable means to assure their appearance at trial or law-abiding conduct before *1157trial. The lack of adequate procedures violates the Due Process Clause. ■

Harris County is also liable for the unconstitutional acts of its Sheriff when he acts as a final policymaker for, and administrator of, the Harris County Jail. The Sheriffs poliey and practice of detaining misdemeanor defendants knowing that they are eligible for release, but are detained on secured money bail, is not narrowly tailored to meet the County’s compelling interests in assuring misdemeanor defendants’ appearance at trial and law-abiding conduct before trial. ■

The plaintiffs have demonstrated a clear likelihood of success on the merits of their equal protection and due process claims against Harris County. That- showing weighs heavily in favor of granting the requested preliminary injunctive relief.119 See Rodriguez v. Providence Comm. Corr., Inc., 155 F.Supp.3d 758, 771 (M.D. Tenn. 2015), appeal dismissed, No. 16-5057 (6th Cir. Mar. 15, 2016); Walker, 2016 WL 361612 at *14, rev’d on other grounds, 682 Fed.Appx. 721, 2017 WL 929750; see also Jones, 2015 WL 5387219; Cooper, Civil No. 15-425 (M.D. Ala. June 18, 2015); Pierce, Civil No. 15-570 (E.D. Mo. June 3, 2015); Thompson, Civil No. 15-182 (S.D. Miss. Nov. 6, 2015).

C. Irreparable Injury

“When an alleged deprivation of a constitutional right is involved, ... most courts hold that no further showing of irreparable injury is necessary.” 11A Wright & Miller, Federal Practice & Procedure, § 2948.1 (3d ed. 1998). The plaintiffs have shown that Harris County detains misdemeanor defendants who are otherwise eligible for release because they cannot pay the secured financial condition necessary for release. Both the Harris County Sheriff and a County Judge credibly testified that without an injunction from this court, Harris County’s policies, practices, and customs will continue and misdemeanor defendants will be unnecessarily incarcerated. Hearing Tr. 3-1:52-53; 3-2:22-24. The incarceration deprives misdemeanor defendants of their state-created liberty interest. “Freedom from imprisonment—from government custody, detention, and other forms of physical restraint—lies at the heart of the liberty that [the Due Process] Clause protects.” Zadvydas v. Davis, 533 U.S. 678, 690, 121 S.Ct. 2491, 150 L.Ed.2d 653 (2001); Foucha v. Louisiana, 504 U.S. 71, 80, 112 S.Ct. 1780, 118 L.Ed.2d 437 (1992) (“Freedom from bodily restraint has always been at the core of the liberty protected by the Due Process Clause from arbitrary governmental action.”).

The record evidence shows that the plaintiffs’ injury is irreparable. Misdemeanor defendants detained before trial face significant pressure to plead guilty, and in fact do so at much higher rates than *1158those released before trial, in order to obtain release.120 Pretrial detention of misdemeanor defendants, for even a few days, increases the chance of conviction and of nonappearance or new criminal activity during release.121 Cumulative' disadvantages mount for already impoverished misdemeanor defendants who cannot show up to work, maintain their housing arrangements, or help them families because they are détained.122 This factor weighs strongly in favor of. granting the plaintiffs’ request for the-injunctive relief. See also Rodriguez, 155 F.Supp.3d at 771 (irreparable harm from jailing probationers on secured money bonds for probation violations supported injunction); Walker, ' 2016 WL 361612 at *14 (irreparable harm from jailing a misdemeanor defendant “simply because he could not afford to post money bail”), rev’d on other grounds, 682 Fed. Appx. 721, 2017 WL 929750.

D. Balancing the Harms

Courts “must balance the competing claims of injury and must consider the effect on each party of the granting or withholding of the requested relief.” Amoco Prod. Co. v. Village of Gambell, AK, 480 U.S. 531, 542, 107 S.Ct. 1396, 94 L.Ed.2d 542 (1987). “In exercising their sound discretion, courts of equity should pay particular regard for the public consequences in employing the extraordinary remedy of injunction.” Weinberger v. Romero-Barcelo, 456 U.S. 305, 312, 102 S.Ct. 1798, 72 L.Ed.2d 91 (1982) (citation omitted),

The defendants argue that proposed reforms expected to be implemented by July 1, 2017 will adequately address the plaintiffs’ injuries and that a court order could disrupt implementing these - reforms. (Docket Entry No. 166 at 17-23); Hearing Tr. 8-2:25-26. The defendants note that they have been working on the reforms for eighteen months. The reforms require a bottom-up “buy-in” from -Harris County’s “various criminal justice stakeholders” to be successful, not a top-down-order imposed from outside.123 (Docket Entry No. 166 at 17-18); Hearing Tr. 8-2:95.

Harris County’s adoption of the Arnold Tool and other -reforms are commendable.124 But, as noted above, the reforms will not address the, plaintiffs’ allegations that Harris County imposes secured financial conditions of release to detain indigent misdemeanor defendants who- cannot pay, despite Texas - state-law prohibitions of pretrial detention' orders for all -but one narrow category of misdemeanor defendants.125 The use of bail to detain, rather than release,, misdemeanor defendants based on their poverty is not just a possibility under the new system; it is Harris County’s stated policy purpose to use secured money bail to detain “high-risk” defendants, an as yet undefined category.126

The record evidence also calls into question the extent to which the forthcoming reforms will remedy the County’s due process violations. The Harris County Sheriff and one County Judge testified that counseled hearings' are unlikely to,,change the *1159Hearing Officers’ practice and custom of ordering indigent misdemeanor defendants to pay secured money bail, knowing that the orders operate as de facto pretrial detention orders. Hearing Tr.- 3-1:113-14, 3-2:22. The record evidence shows that despite changing the County Rules of Court to presume release on personal bonds is appropriate in twelve offense categories, Hearing Officers and County Judges continue to detain misdemeanor defendants, including the indigent, at the same rate as they did in the two years before the rule change.127 Hearing Officers and County Judges reject the recommendations the Pretrial Services officers make using the County’s current validated risk-assessment tool to release misdemeanor defendants on unsecured personal bonds about 67 percent of the time.128 The reformed system will permit Healing Officers and County Judges to continue rejecting the recommendations that result from the County’s new validated Arnold Risk-Assessment Tool at the same rate.129 Although the new inmate-processing center may help the County to provide bail-setting hearings in 24 hours after arrests for more, or even all, misdemeanor defendants, the center will not be complete until March 2018.130

The court does not intend or want to interfere with the laudable reforms that will improve the fairness of the. County’s pretrial arrest system. The proposed reforms will not take effect for months. The present system will continue during that time, detaining over 100 misdemeanor defendants every day in the Harris County Jail, defendants who are eligible for release but whose indigence makes them unable to pay a secured financial condition of release. The record shows that after July 1 (or the date the reforms are in fact implemented), the County’s system will hot remedy the constitutional infirmities of its current policies. The County Judges suggest that the court should “craft any relief to work in conjunction with these new changes, rather than .., wholly enjoining the present system.” (Docket Entry No. 166 at 28). That is the better approach. With carefully tailored relief, the balance of the harms between granting or denying a preliminary injunction strongly favors the plaintiffs.

E. The Public Interest

“It is always in the public interest to prevent the violation of a party’s constitutional rights.” Simms v. District of Columbia, 872 F.Supp.2d 90, 105 (D.D.C. 2012) (collecting cases). In an amicus brief, the Harris County District Attorney emphasizes that “[h]olding‘ un-adjudicated misdemeanor offenders in the Harris County Jail solely because they lack the money or other means of posting bail , is counterproductive to the goal of seeing that justice is done.... It makes no sense to spend public funds to house misdemean- or offenders in a high-security penal facility when the crimes themselves may not merit jail time.” (Docket Entry No. 2016 at 1-2). The court agrees. Texas state law treats misdemeanor defendants, with one narrow exception, as eligible for pretrial release. The public interest is not served by .incarcerating misdemeanor defendants who, because of poverty, are unable to pay secured money bail. This factor weighs strongly in favor of granting the plaintiffs’ request for relief.

F. Bond

A federal court may waive the bond requirement. Fed. R. Civ. Pro. 65(c); City *1160 of Atlanta v. Metro. Atlanta Rapid Transit Auth., 636 F.2d 1084, 1094 (5th Cir. Unit B 1981); Corrigan Dispatch Co. v. Casa Guzman, S.A., 569 F.2d 300, 303 (5th Cir.1978). The court finds that waiving the bond is appropriate in this case; the plaintiffs are indigent, see Wayne Chem., Inc. v. Columbus Agency Serv. Corp., 567 F.2d 692, 701 (7th Cir. 1977), and the plaintiffs have brought this suit to enforce constitutional rights, see City of Atlanta, 636 F.2d at 1094. No bond is imposed.

III. Remedy

“In view of the fact that plaintiffs established a constitutional violation, ... the task of fashioning a proper remedy is one that should be performed by the District Court after all interested parties have had an opportunity to be heard. The judicial remedy for a proven violation of law will often include commands that the law does not impose on the community at large.” Chicago Teachers Union, Local No. 1, AFT, AFL-CIO v. Hudson, 475 U.S. 292, 309 n.22, 106 S.Ct. 1066, 89 L.Ed.2d 232 (1986); see also Swann v. Charlotte-Mecklenburg Bd. of Ed., 402 U.S. 1, 15-16, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971); Gates v. Collier, 501 F.2d 1291, 1320 (5th Cir. 1974). “Every order granting an injunction ... must: (A) state the reasons why it issued; (B) state its terms specifically; and (C), describe in reasonable detail—and not by referring to the complaint or other document—the act or acts restrained or required.” Fed. R. Civ. P. 65(d)(1). “Rule 65 protects those who are enjoined by informing them of ... exactly what conduct is proscribed and ensures informed and intelligent appellate review.” Walker, 682 Fed.Appx. at 724, 2017 WL 929750 at *2 (internal quotation marks and citation omitted; alteration in original); see also Hornbeck Offshore Serv., LLC v. Salazar, 713 F.3d 787, 792 (5th Cir. 2013). The court has explained in detail its reasons for issuing preliminary injunctive relief. Several principles inform and guide the court in exercising its discretion and adhering to the record evidence and the law to fashion a suitable remedy.

First, because the plaintiffs have not alleged the facial unconstitutionality of Texas statutes or the County Rules of Court, the court will not require relief that is inconsistent with Texas law or the County Rules as written. (See Docket Entry No. 145 at 7; No. 288 at 10-11). Both Texas law and the County Rules provide for setting money bail in specific amounts, but neither requires that money bail be set in misdemeanor cases on a secured, rather than unsecured, basis.131 Using a bail schedule is not inherently unconstitutional. See Fields, 701 F.3d at 184; Terrell, 481 F.Supp.2d at 766-67. The constitutional problem in this case arises from rigid adherence to imposing secured money bail when that will obviously result in, and is often intended to effect, pretrial detention of indigent defendants charged only with misdemeanors who are eligible for release under Texas law.

Second, the court does not enjoin judicial officers acting in a judicial capacity, as prohibited by 42 U.S.C. § 1983. The County Judges argue that they are immune from an injunction in their legislative capacity as well. (Docket Entry No. 166 at 15 n.20). The court rejected that argument in its Memorandum and Opinion on the County Judges’ motion to dismiss. ODonnell, 227 F.Supp.3d at 756-57, 2016 WL 7337549 at *36. Nevertheless, the only relief against the County Judges in their legislative capacity required at this time is that they do not legislate policy that contradicts this court’s order.

*1161Third, as much as possible, the court avoids interfering -with the salutary reforms the County is proposing to implement by July 2017. The court has worked with the parties’ briefs and arguments as guidance in fashioning relief that is consistent with, and can be implemented alongside, the proposed reforms. Additionally, the court -will provide the County over fourteen days from the date of the Order of Preliminary Injunction to implement the ordered relief.

Fourth, the relief must be effective to address the serious constitutional violations proven at the motion hearing. The defendants propose that the court should substantially, repeat the Roberson order and require the Hearing Officers to “consider[ ] an arrestee’s ability to pay if they impose secured bail.” (Docket Entry No. 259 at 3). That approach appears to enjoin judicial officers acting in judicial capacities, contrary to 42 U.S.C. § 1983. The approach would also permit Harris County to continue imposing secured money bail in order to detain indigent misdemeanor , defendants who, if they could pay, would be released. These bail orders operate as de facto orders of pretrial preventive detention, without the procedures due process requires and in violation of equal protection. Adequate relief requires that those eligible for release before trial under state and federal law are released and not detained because their indigence makes them unable to pay a secured financial condition required for release.

Fifth, while relief must.be effective, it must also balance the competing interests. The plaintiffs contend that no amount of differential treatment is tolerable under the Equal Protection Clause and that indigent misdemeanor arrestees must be released at substantially the same time as those who are able to pay secured money bail. (See Docket Entry No. 257, Ex. 1 at 1, 4). Various parties also suggest changing the timeline of the arrest process. The County Judges argue .-that the Sheriff should be compelled to book misdemeanor arrestees at the County Jail within 18 hours of their arrest. (Docket Entry No. 259 at 3). The plaintiffs suggest various limits on sobriety periods and on the time it takes to process misdemeanor arrestees when they bond out of jail. (Docket Entry No. 257, Ex. 1 at 4-5). The parties did not provide detail on how to set and implement the precise timing and- speed of various procedures. Those questions are more appropriately resolved at the trial on the merits.

With these principles in mind, thé court will order the following relief, to take effect by May 15,2017, unless those enjoined move for more time and show good cause for a reasonable, brief extension. Any motions for extension will be set for prompt hearing and resolution.

• Harris County 'and its policymakers— the County Judges in their legislative and rulemaking capacity and the Harris County Sheriff in his law-enforcement capacity—are enjoined from detaining indigent misdemeanor defendants who are .otherwise eligible for release but are unable because of their poverty to pay a secured money bail.

• Pretrial Services officers, as County employees and subject to its policies, must verify an arrestee’s ability to pay a secured financial condition of release by an affidavit, and must explain to arrestees the nature and significance of the verification process.

• The purpose of the explanation is to provide the notice due process requires that a misdemeanor defendant’s right to liberty before trial is at stake in the proceedings. Pretrial Services may administer either the form of the affidavit *1162currently used to determine eligibility for appointed counsel or the adapted' form that Dr. VanNostrand testified is being prepared for Pretrial Services to be administered by July 1, 2017. See Hearing Tr. 6-1:136; see also id, at 4-1:48-49. Pretrial Services must deliver completed affidavits to the Harris County Sheriffs Office before a declar-ant’s probable cause hearing.

• The affidavit must give the misdemean-. or arrestee sufficient opportunity to de-, clare under penalty of perjury, after the significance of the information has been explained, the maximum amount of financial security the arrestee.would be able to post or pay up front within 24 hours of arrest. The question is. neither the arrestee’s immediate ability to pay with cash on hand,, nor what, assets the arrestee could eventually produce after a period of pretrial detention. The question is what amount the arrestee could reasonably pay within 24 hours of his or her arrest, from . any source, including the contributions of family and friends.

• The purpose of this requirement 'is to provide a better, easier, and faster way ’’ to get the information needed to determine a misdemeanor defendant’s ability to pay. The Hearing Officers and County Judges testified that they presently do not know who has the ability to pay. Hearing Tr. 4-1:141; 4-2:16; 5:72. The requirement is for a form of verification that Harris County already uses to determine who is indigent and therefore eligible for appointed counsel. Hearing Tr. 2-1:60-61. The affidavit can be completed within 24 hours after arrest; the current process of verifying references by phone extends for days after arrest. (See Docket Entry No. 166 at 10 n.13).

• The court does not order relief against the • Hearing • Officers or against the County Judges in their judicial capacities. The court does not order relief against the County Judges- or Sheriff in their capacities as state actors, except that they may not legislate policies that directly conflict with this court’s order,

• Misdemeanor defendants who are not subject to: (1) formal holds preventing their release from detention; (2) pending mental-health evaluations to determine competency; or (3) pretrial preventive detention orders for violating a condition of release for a crime of family violence, have á constitutionally protected state-created liberty interest in release before trial. If a misdemeanor defendant has executed an affidavit showing an inability to pay . secured, money bail and the Hearing Officer does not order release either: (1) on an unsecured personal bond ' with nónfínaneiál conditions of release; or (2) on a secured money bond for which the defendant could pay a commercial surety’s premium, as indi- ' cated on the affidavit, then the Harris County Sheriff must treat the financial condition as unsecured and release' the misdemeanor ’ defendant promptly after the probable cause ' hearing. All nohfinancial conditions of release ordered by the Hearing Officers, including protective orders, drug testing; alcohol intake ignition locks, or GPS monitoring, will remain in effect. The bail amount determined by the Hearing Officer will remain the bail required of the misdemeanor defendant, but the Sheriff must require it on an unsecured, rather than a secured, basis. An indigent defendant’s inability to pay secured money bail cannot be the basis for the Sheriff to. continue to detain that defendant.

*1163• The purpose of this requirement is to provide timely protection for the state-created liberty interest in pretrial release and to prevent the pretrial detention of a misdemeanor defendant on a financial condition when that defendant would be able to obtain release by paying but is unable to dó so. By “promptly,” the court means on the same time frame of release that a defendant who paid a secured money bail would re- •' ceive.

• The Sheriff must release on unsecured or nonfinancial conditions misdemeanor defendants identified above—those without holds preventing prompt release; pending competency evaluations; or preventive family violence detention orders—who have not had a bail-setting hearing before a Hearing Officer within 24 hours of arrest. In absentia hearings “on the papers” will not satisfy this requirement. If the City of Houston Police Department has, detained a misdemeanor defendant more than 24 hours after arrest, the Sheriff must promptly release the defendant on unsecured or nonfinancial conditions when he takes custody of the defendant, on the same time frame and procedures as if the defendant had paid a secured financial condition of release. The bail amount set by Assistant District Attorneys according to the County Judges’ bail schedule will remain the bail required of the misdemeanor defendant, but the Sheriff must require it on an unsecured, rather than a secured, basis.

• The purpose of this requirement is to give timely protection to the state-created liberty interest in release before trial and to enforce state and federal standards holding that, in Harris County, 24 hours is the outer boundary for completing the administrative incidents to arrest in misdemeanor cases. Sanders, 543 F.Supp. at 704; Roberson, Agreed ' Final Judgment, No. 84-2974 at 1. The 24-hour requirement is particularly intended to address the endemic problem of misdemeanor arrestees being detained until case disposition and pleading guilty to secure faster release from pretrial detention.

•: The Sheriff may not alter nonfinancial conditions- of release ordered by Harris County judicial officers. The Sheriff may'not alter the bail amount deter- • mined by Harris County judicial officers. The only determination the Sheriff must make under this order is the decision to require bail on a secured or unsecured basis. The decision is an objective one. If the misdemeanor defendant’s affidavit shows that the defendant is unable pay the bail up front or pay a bondman’s premium for the principal sum required by the Hearing Officers, the Sheriff must require the bail amount, but on an unsecured basis. The Sheriff may release misdemeanor defendants on an unsecured bond without a Hearing Officer’s signature on the release order. The Sheriffs acceptance of bail on an unsecured basis accords with his authority to accept bail and release misdemeanor defen- . dants under Article 17.20 of the Texas Code of Criminal Procedure.132

*1164• Texas law provides a significant role for sheriffs in setting and taking bail in misdemeanor cases. Sheriffs ordinarily defer to magistrates in setting bail, unless “no magistrate is available.” Hokr, 545 S.W.2d at 463. And a sheriff executes “legal process which it is made his duty by law to execute.” Tex. Code Chim. Pro. art. 2.16; see also Tex. Local Gov’t Code § 85.021. The purpose of this order is to inform the Harris County Sheriff that, as Sheriff Gonzalez recognized in his declaration, orders to detain misdemeanor defendants on a secured financial condition of release that they cannot pay because of their poverty are unconstitutional and invalid under federal law. See Pis. Ex. 7(r). Harris County and Sheriff Gonzalez as its policymaker are liable for, and enjoined from, executing invalid orders from the Hearing Officers or County Judges that operate to detain indigent misdemeanor defendants who are otherwise eligible for release if they cannot pay a secured financial condition of release.

• For misdemeanor defendants who are subject to formal holds and who have executed an affidavit showing an inability to pay the secured financial condition of release, the Sheriff must treat the limitations period on their holds as beginning to run the earliest of: (1) after the probable cause hearing; or (2) 24 hours after arrest. The purpose of this requirement is to ensure that misdemeanor defendants are not prevented from or delayed in addressing their holds because they are indigent and therefore cannot pay a secured financial condition of release.

• Misdemeanor defendants who do not appear competent to execute an affidavit may be evaluated under the procedures set out in the Texas Code of Criminal Procedure Article 16.22. If competence is found, the misdemeanor defendant is covered by the relief the court orders, with the exception that the 24-hour period begins to run from the finding of competence rather than from the time of arrest. As under Article 16.22, nothing in this order prevents the misdemeanor arrestee from being released on secured bail or unsecured personal bond pending the evaluation.

• The court’s relief applies to misdemeanor arrestees who are re-arrested on misdemeanor charges only or on warrants for failure to appear while on pretrial release for their misdemeanor charges. Texas does not permit preventive pretrial detention orders in misdemeanor cases, even for multiple failures to appear or for new criminal activity before trial. Misdemeanor defendants unable to pay a secured financial condition of release do not lose their state-created liberty interest in release before trial by failing to appear or by committing new misdemeanor criminal activity. Those defendants may, of course, face additional charges and exposure to longer sentences, as well as enhanced nonfinancial condi*1165tions of release, such as more demanding supervisory techniques, for their pretrial misconduct.

The court concludes that this relief strikes an equitable balance between the parties’ interests in this case. The plaintiffs seek to eliminate entirely any differential treatment between those able to pay secured money bail and those unable to do so. (See Docket Entry No; 188 at 19; No. 257, Ex. 1 at 1, 4). The defendants argue that a judicial officer should assess nonfi-nancial conditions of release and that both the judicial officer and the misdemeanor defendant require time to prepare for a full bail-setting hearing. (Docket Entry No. 260 at 5; No. 286 at 6-7,15-16). Under the court’s relief, some misdemeanor defendants may be able to pay a secured financial condition and be released between having their charges formalized (about 15 hours after arrest, see Hearing Tr. 3-2:85) and appearing before a Hearing Officer (usually within 19 to 24 hours after arrest, if the defendant remains detained, see Def. Ex. 28 at 14-15). The court considers this difference de minimis at this stage, although the plaintiffs may re-urge their position at the merits trial. The time frames are expected to substantially decrease as the County implements its' reforms. See Hearing Tr. 6-1:138-39,145-46. And the court has tailored relief to address one significant cause of the differential treatment—extended periods of detention by the City of Houston Police Department before Harris County takes custody of ar-restees.

The defendants’ many objections to relief as proposed by the plaintiffs do not apply to the relief as ordered by the court. The court is not striking down the use of secured money bail. (Cf. Docket Entry No. 161 at 8; No. 26 at 23-24). The court is not permitting arrestees to “set their own bail.” (Cf. Docket Entry No. 266 at 7; Hearing Tr. 6-1:159-60). Bail amounts— and the County’s right to collect forfeited bail—remain within the discretion Harris County officers have under state law. The County may continue to release defendants on secured financial conditions if those conditions serve to release, rather than detain, misdemeanor defendants before trial. What the County is enjoined from doing is setting the amount of bail on a secured basis in a way that detains, rather than releases, misdemeanor defendants who would be released if they could pay but who are unable to do so, in violation of the Constitution. And Harris County has long used affidavits of indigence as the basis to appoint publicly funded counsel. Hearing Tr. 2-1:60-61; 6-1:160-61. That is not a means of letting defendants confer public benefits on themselves. See, e.g., Adkins v. E.I. DuPont de Nemours & Co., 335 U.S. 331, 339, 69 S.Ct. 85, 93 L.Ed. 43 (1948). In the context ’ of pretrial bail, affidavits of indigence have been used in the United States for over 150 years.133

The defendants argue that the administration of criminal justice is a police power granted to state and local governments by the Tenth Amendment. (Docket Entry No. 259 at 4). It is. They argue that judicial officers must exercise discretion in setting conditions of release before trial. (Id. at 6, 9-10). They must. But neither police power nor judicial discretion are boundless. “Congress enacted § 1983 to enforce provisions of the Fourteenth . Amendment against those who carry a badge.of authority and represent it in some capacity, whether they act in accordance with their authority or misuse it.” Hafer v. Melo, 502 U.S. 21, 28, 112 S.Ct. 358, 116 L.Ed.2d 301 (1991) (internal quotation marks and citation omitted). That enactment would be mean*1166ingless if the Tenth Amendment exempted state officers from liability under the Fourteenth Amendment.

The defendants argue that the putative plaintiff class is bound by the Roberson order issued in 1987 and that the consent decree can only be modified under Rule 60(b). (Docket Entry No. 264 at 10-11). But a consent decree does not bind non-parties. See, e.g., Martin v. Wilks, 490 U.S. 755, 761-62, 109 S.Ct. 2180, 104 L.Ed.2d 835 (1989). On the defendants’ side, Harris County, the Hearing Officers, and the Sheriff were not parties to the Roberson litigation. See Def. Ex. 159. The Roberson defendants were twelve County Judges sued in their personal and official -capacities, See id. The plaintiffs in this case, do not seek relief against the Roberson County Judges in their personal capacities, and the number of County Judges acting in an official capacity has grown by four since 1987. On the .plaintiffs’ side, the Roberson class did not include misdemeanor defendants arrested on a warrant, as this _ one does. See Def. Ex. 160. The Roberson class did include misdemeanor defendants asserting Fourth and Sixth Amendment claims against the County Judges; this class does. not. See id.

For the reasons explained above, the Roberson order does not require the conduct the plaintiffs challenge.134 The relief ordered here is not inconsistent with Roberson and is not a modification of the consent decree. And even - if the parties were identical and the relief here was inconsistent with that ordered in Roberson, courts have held that, in the context of civil-rights litigation, a “modern successor” class action better serves the public interest and is better at resolving the parties’ disputes than a Rule 60(b) modification of a decades-old consent decree. See Coffey v. Braddy, 88 F.Supp.3d 1283, 1299 (M.D. Fla. 2015), aff'd, 834 F.3d 1184, 1193 (11th Cir. 2016).

Finally, the defendants object that they cannot implement a rule against using secured financial conditions of release as de facto orders of pretrial preventive detention, as' other jurisdictions do, because Texas does not permit transparent orders of pretrial preventive detention in misdemeanor cases that are available in other jurisdictions. {See, e.g., Docket Entry No 159 at 9-10; No. 166 at 13; Hearing Tr. 8-2:24-25). The defendants’ dissatisfaction is with Texas law, not with the plaintiffs’ claims or the relief this court ordered. It may indeed be wise to keep risky defendants, including misdemeanor defendants, in jail from arrest forward. But Texas law makes a different choice. It prohibits pretrial preventive detention of all but one category of misdemeanor cases, and in that exceptional category it provides nonfi-nancial conditions of pretrial detention with extra procedural safeguards. Jailing the indigent by setting secured money bail that they cannot pay makes an end run around a Texas-created liberty interest •without providing due process. If the defendants believe that some misdemeanor defendants present such a high risk of npnappearance or of new criminal activity as to require pretrial preventive detention, the defendants’ proper recourse is to petition the Texas Legislature to amend the Texas Constitution, not to accomplish a de facto amendment through imposing secured financial conditions of release that operate as detention orders only against those who cannot pay.

IV. Conclusion

A. Summary Judgment

In their motion for summary judgment, the defendants argue that: (1) there is no *1167constitutional right to “affordable bail”; (2) the Harris County Rules of Court pass rational basis review; and (3) the County’s arrest procedures satisfy due process as a matter of law. (Docket Entry No. 101). The court has addressed at length the standard of review required by the Supreme Court and the Fifth Circuit in this case,135 and the reasons that Harris, County’s procedures violate the Equal Protection and Due Process Clauses as a matter of law.136

Asiexplained in detail above, the issue in this case is not the right to “affordable bail.” As cases and commentaries make clear, courts may impose secured money bail beyond a defendant’s ability to pay: (1) in cases of dangerous felony; (2) after finding that no alternative to secured' money bail can reasonably assure the defendant’s appearance or public safety;-(3) with the due process of a detention- order if the secured money bail in fact operates to detain the defendant. Those factors do not apply to this case.137 Misdemeanor charges are not dangerous felonies. The credible and reliable record evidence shows that, in misdemeanor cases, secured money bail is not the only reasonable alternative to assure appearance and law-abiding conduct before trial.

That does not amount to a “right to affordable bail.” Under Texas law, Harris County magistrates—the Hearing Officers and County Judges—may weigh the state-law factors to arrive at a high amount of bail. Tex. Code Crim. Pro. art. 17.15. But they cannot, consistent with the federal Constitution, set that bail on a secured basis requiring up-front payment from indigent misdemeanor defendants otherwise eligible for release, thereby converting the inability to pay into an automatic order of detention without due process and in violation of equal protection. See Bearden, 461 U.S. at 672, 103 S.Ct. 2064; Rainwater, 572 F.2d at 1056. The motion for summary judgment is denied.

B. Preliminary Injunction

“Rules under which personal liberty is to be deprived, are limited by the constitutional guarantees of all, be they moneyed or indigent, befriended or friendless, employed or unemployed, resident or transient, of good reputation or bad.” Rainwater, 572 F.2d at 1057. Misdemean- or arrestees are often, as Judge Traman Morrison testified, people “living on the edge at the point in their lives that intersects with getting involved in an arrest,” Hearing Tr. 2-2:135. In Harris County, they may be homeless. They may lack family, friends, and “co-indemnitors.” Some are, no doubt, of bad reputation and present a risk of nonappearance or of new criminal activity. But they are not without constitutional rights to due process and the equal protection of the law.

The court has considered an extensive record consisting of hundreds of exhibits, thousands of hearing recordings,' and eight days of arguments and briefing at the motion hearing. The record evidence, the arguments of able counsel, and the extensive case law and commentary on bail and pretrial detention all show that the plaintiffs are entitled to preliminary injunctive relief. Harris County’s policy is to detain indigent misdemeanor defendants before trial, violating equal protection rights against wealth-based discrimination and violating due process protections against pretrial detention without proper procedures or an opportunity to be heard.

*1168This case is not easy. Institutions charged with safeguarding the public have an extraordinary trust and a difficult task. The difficulty and importance of the task cannot defeat an equally important public trust, which the court and the defendants share—to enforce the Constitution. The court has done its best to recognize and work toward both. Harris County is changing its bail procedures. That is commendable. The relief ordered here is intended to fit into that work, to discharge the responsibilities the court and the parties share.

The plaintiffs’ clear likelihood of success on the merits of their claims at trial, the irreparable injuries they will suffer without an order of relief from this court, the public interest, and the relative weight of the harms should the court refuse relief all weigh strongly in the plaintiffs’ favor. The Order of Preliminary Injunction is separately entered.

EDITH BROWN CLEMENT, Circuit Judge:

*152The appellees' petition for panel rehearing is granted. The prior panel opinion, ODonnell v. Harris County , 882 F.3d 528 (5th Cir. 2018) is withdrawn, and the following opinion is substituted:

Maranda ODonnell and other plaintiffs (collectively, "ODonnell") brought a class action suit against Harris County, Texas, and a number of its officials-including County Judges,1 Hearing Officers, and the Sheriff (collectively, the "County")-under 42 U.S.C. § 1983. ODonnell alleged the County's system of setting bail for indigent misdemeanor arrestees violated Texas statutory and constitutional law, as well as the equal protection and due process clauses of the Fourteenth Amendment. ODonnell moved for a preliminary injunction, and the County moved for summary judgment. After eight days of hearings, at which the parties presented numerous fact and expert witnesses and voluminous written evidence, the district court denied the County's summary judgment motion and granted ODonnell's motion for a preliminary injunction. The County then applied to this court for a stay of the injunction pending appeal, but the motion was denied, and the injunction went into effect. Before this court now is the County's appeal, seeking vacatur of the injunction and raising numerous legal challenges.

For the reasons set forth, we affirm most of the district court's rulings, including its conclusion that ODonnell established a likelihood of success on the merits of its claims that the County's policies violate procedural due process and equal protection. We disagree, however, with the district court's analysis in three respects: First, its definition of ODonnell's liberty interest under due process was too broad, and the procedures it required to protect that interest were too onerous. Second, it erred by concluding that the County Sheriff can be considered a County policymaker under § 1983. Finally, the district court's injunction was overbroad. As a result, we will vacate the injunction and order the district court to modify its terms in a manner consistent with this opinion.

*153I.

We need not conduct an exhaustive review of the facts. The district court's account is expansive: It comprised over 120 pages of factual findings, including not only the specific details of the County's bail-setting procedures, but also the history of bail and recent reform attempts nationwide.

Bail in Texas is either secured or unsecured. Secured bail requires the arrestee to post bond either out of the arrestee's pocket or from a third-party surety (often bail bondsmen, who generally require a 10% non-refundable premium in exchange for posting bond). Unsecured bail, by contrast, allows the arrestee to be released without posting bond, but if he fails to attend his court date and/or comply with any nonfinancial bail conditions, he becomes liable to the County for the bail amount. Both secured and unsecured bail may also include nonfinancial conditions to assure the detainee's attendance at future hearings.

The basic procedural framework governing the administration of bail in Harris County is set by the Texas Code of Criminal Procedure and local rules promulgated by County Judges. See Tex. Gov't Code § 75.403(f). When a misdemeanor defendant is arrested, the prosecutor submits a secured bail amount according to a bond schedule established by County Judges. See Harris County Criminal Courts at Law Rule 9 (hereinafter , "Local Rule"). Bonds are then formally set by Hearing Officers and County Judges. Tex. Code. Crim. Pro. art. 2.09, 17.15. Hearing Officers are generally responsible for setting bail amounts in the first instance. This often occurs during the arrestee's initial probable cause hearing, which must be held within 24 hours of arrest. Tex. Code Crim. Pro. art. 17.033 ; Local Rule 4.2.1.1. County Judges review the Hearing Officers' determinations and can adjust bail amounts at a "Next Business Day" hearing. Local Rule 4.3.1.

The Hearing Officers and County Judges are legally proscribed from mechanically applying the bail schedule to a given arrestee. Instead, the Texas Code requires officials to conduct an individualized review based on five enumerated factors, which include the defendant's ability to pay, the charge, and community safety. Tex. Code of Crim. Pro. art. 17.15. The Local Rules explicitly state the schedule is not mandatory. They also authorize a similar, individualized assessment using factors which partially overlap with those listed in the Code. Local Rule 4.2.4. Hearing Officers and County Judges sometimes receive assessments by Pretrial Services, which interviews the detainees prior to hearings, calculates the detainees flight and safety risk based on a point system, and then makes specific recommendations regarding bail.2

Despite these formal requirements, the district court found that, in practice, County procedures were dictated by an unwritten custom and practice that was marred by gross inefficiencies, did not achieve any individualized assessment in setting bail, and was incompetent to do so. The district court noted that the statutorily-mandated probable cause hearing (where bail is usually set) frequently does not occur within 24 hours of arrest. The hearings often last seconds, and rarely more than a few minutes. Arrestees are instructed not to *154speak, and are not offered any opportunity to submit evidence of relative ability to post bond at the scheduled amount.

The court found that the results of this flawed procedural framework demonstrate the lack of individualized assessments when officials set bail. County officials "impose the scheduled bail amounts on a secured basis about 90 percent of the time. When [they] do change the bail amount, it is often to conform the amount to what is in the bail schedule." The court further found that, when Pretrial Services recommends release on personal bond, Hearing Officers reject the suggestion 66% of the time. Because less than 10% of misdemeanor arrestees are assigned an unsecured personal bond, some amount of upfront payment is required for release in the vast majority of cases.

The court also found that the "Next Business Day" hearing before a County Judge fails to provide a meaningful review of the Hearing Officer's bail determinations. Arrestees routinely must wait days for their hearings. County Judges adjust bail amounts or grant unsecured bonds in less than 1% of cases. Furthermore, prosecutors routinely offer time-served plea bargains at the hearing, and arrestees are under immense pressure to accept the plea deals or else remain incarcerated for days or weeks until they are appointed a lawyer.

The district court further noted the various ways in which the imposition of secured bail specifically targets poor arrestees. For example, under the County's risk-assessment point system used by Pretrial Services, poverty indicators (such as not owning a car) receive the same point value as prior criminal violations or prior failures to appear in court. Thus, an arrestee's impoverishment increased the likelihood he or she would need to pay to be released.

The court also observed that Hearing Officers imposed secured bails upon arrestees after having been made aware of an arrestee's indigence by the risk-assessment reports or by the arrestee's own statements. And further, after extensive review of numerous bail hearings, the court concluded Hearing Officers were aware that, by imposing a secured bail on indigent arrestees, they were ensuring that those arrestees would remain detained.

The court rejected the argument that imposing secured bonds served the County's interest in ensuring the arrestee appeared at the future court date and committed no further crime. The court's review of reams of empirical data suggested the opposite: that "release on secured financial conditions does not assure better rates of appearance or of law-abiding conduct before trial compared to release on unsecured bonds or nonfinancial conditions of supervision." Instead, the County's true purpose was "to achieve pretrial detention of misdemeanor defendants who are too poor to pay, when those defendants would promptly be released if they could pay." In short, "secured money bail function[ed] as a pretrial detention order" against the indigent misdemeanor arrestees.

The district court also reviewed voluminous empirical data and academic literature to evaluate the impact of pretrial detention on an arrestee. The court found that the expected outcomes for an arrestee who cannot afford to post bond are significantly worse than for those arrestees who can. In general, indigent arrestees who remain incarcerated because they cannot make bail are significantly more likely to plead guilty and to be sentenced to imprisonment. They also receive sentences that are on average twice as long as their bonded *155counterparts. Furthermore, the district court found that pretrial detention can lead to loss of job, family stress, and even an increase in likeliness to commit crime.

The court concluded that ODonnell had established a likelihood of success on the merits of their claim that the County violated both the procedural due process rights and the equal protection rights of indigent misdemeanor detainees. It granted the motion for a preliminary injunction, requiring the implementation of new safeguards and the release of numerous detainees subjected to the insufficient procedures.

II.

This court reviews a "district court's grant of a preliminary injunction ... for abuse of discretion." Women's Med. Cty. of Nw. Hous. v. Bell , 248 F.3d 411, 418-19 (5th Cir. 2001). "Findings of fact are reviewed only for clear error; legal conclusions are subject to de novo review." Id. at 419. "Issuance of an injunction rests primarily in the informed discretion of the district court. Yet injunctive relief is a drastic remedy, not to be applied as a matter of course." Marshall v. Goodyear Tire & Rubber Co. , 554 F.2d 730, 733 (5th Cir. 1977) (internal citations omitted). A district court abuses its discretion if it issues an injunction that "is not narrowly tailored to remedy the specific action which gives rise to the order as determined by the substantive law at issue." Scott v. Schedler , 826 F.3d 207, 211 (5th Cir. 2016) (internal quotation marks and alterations omitted).

III.

The County raises a number of arguments that do not implicate the merits of ODonnell's constitutional claims. We address these first.

A. Status of County Judges and Sheriff as County Policymakers under § 1983

The County appeals the district court's ruling that the County Judges and Sheriff qualified as Harris County policymakers under 42 U.S.C. § 1983. Liability under § 1983 attaches to local government officers "whose [unlawful] decisions represent the official policy of the local governmental unit." Jett v. Dall. Indep. Sch. Dist. , 491 U.S. 701, 737, 109 S.Ct. 2702, 105 L.Ed.2d 598 (1989). Whether an officer has been given this authority is "a question of state law." Pembaur v. City of Cincinnati , 475 U.S. 469, 483, 106 S.Ct. 1292, 89 L.Ed.2d 452 (1986). "Official policy" includes unwritten widespread practices that are "so common and well settled as to constitute a custom that fairly represents municipal policy." Johnson v. Moore , 958 F.2d 92, 94 (5th Cir. 1992) (quoting Bennett v. City of Slidell , 735 F.2d 861, 862 (5th Cir. 1984) (en banc) ). And unlawful decisions include "acquiescence in a longstanding practice or custom which constitutes the standard operating procedure of the local governmental entity." Jett , 491 U.S. at 737, 109 S.Ct. 2702 (internal quotation marks omitted).

Though a judge is not liable when "acting in his or her judicial capacity to enforce state law," Moore , 958 F.2d at 94, we agree with the district court that the County Judges are policymakers for the municipality. Texas law explicitly establishes that the Judges are "county officers," TEX. CONST. art. V § 24, imbued with broad authority to promulgate rules that will dictate post-arrest policies consistent with the provisions of state law, Tex. Gov't Code § 75.403(f). Here, ODonnell alleged that, despite having this authority, County Judges acquiesced in an unwritten, countywide process for setting bail that violated *156both state law and the Constitution. In other words, they sue the County Judges as municipal officers in their capacity as county policymakers. Section 1983 affords them an appropriate basis to do so.

We agree with the County that its Sheriff is not an appropriate party for attaching municipal liability, however. The Sheriff does not have the same policymaking authority as the County Judges. To the contrary, the Sheriff is legally obliged to execute all lawful process and cannot release prisoners committed to jail by a magistrate's warrant-even if prisoners are committed "for want of bail." See Tex. Code Crim. Pro. arts. 2.13, 2.16, 2.18 ; Tex. Loc. Gov't Code § 351.041(a) (noting the Sheriff's authority is "subject to an order of the proper court"). State statutes, in other words, do not authorize the County Sheriff to avoid executing judicial orders imposing secured bail by unilaterally declaring them unconstitutional. Accordingly, the County Sheriff does not qualify as a municipal policymaker under § 1983.

B. Younger Abstention

The County next argues that Younger abstention precludes our review of ODonnell's claims. We are not persuaded.

The Supreme Court held in Younger v. Harris that, when a party in federal court is simultaneously defending a state criminal prosecution, federal courts "should not act to restrain [the state] criminal prosecution, when the moving party has an adequate remedy at law and will not suffer irreparable injury if denied equitable relief." 401 U.S. 37, 43-44, 91 S.Ct. 746, 27 L.Ed.2d 669 (1971). Its conclusion was motivated by the "basic doctrine of equity jurisprudence," "notion[s] of 'comity,' " and "Our Federalism." Id. Courts apply a three-part test when deciding whether to abstain under Younger . There must be (1) "an ongoing state judicial proceeding" (2) that "implicate[s] important state interests" and (3) offers "adequate opportunity" to "raise constitutional challenges." Middlesex Cty. Ethics Comm. v. Garden State Bar Ass'n , 457 U.S. 423, 432, 102 S.Ct. 2515, 73 L.Ed.2d 116 (1982).

The third prong of this test is not met. As the Supreme Court has already concluded, the relief sought by ODonnell-i.e., improvement of pretrial procedures and practice-is not properly reviewed by criminal proceedings in state court. See Gerstein v. Pugh , 420 U.S. 103, 108 n.9, 95 S.Ct. 854, 43 L.Ed.2d 54 (1975) (noting that abstention did not apply because "[t]he injunction was not directed at the state prosecutions as such, but only at the legality of pretrial detention without a judicial hearing, an issue that could not be raised in defense of the criminal prosecution"); see also Pugh v. Rainwater , 483 F.2d 778, 781-82 (5th Cir. 1973) (noting that a federal question whose "resolution ... would [only] affect state procedures for handling criminal cases .... is not 'against any pending or future court proceedings as such ' " (quoting Fuentes v. Shevin , 407 U.S. 67, 71 n.3, 92 S.Ct. 1983, 32 L.Ed.2d 556 (1972) ) ), rev'd on other grounds by Gerstein , 420 U.S. 103, 95 S.Ct. 854, 43 L.Ed.2d 54. As the district court noted, the adequacy of the state court review of bail-setting procedures is essential to ODonnell's federal cause of action. In short, "[t]o find that the plaintiffs have an adequate hearing on their constitutional claim in state court would decide [its] merits."

We also note that the policy concerns underlying this doctrine are not applicable here. The injunction sought by ODonnell seeks to impose "nondiscretionary procedural safeguard[s]," which will not require federal intrusion into pre-trial decisions on a case-by-case basis.

*157Tarter v. Hury , 646 F.2d 1010, 1013-14 (5th Cir. Unit A June 1981) ; compare O'Shea v. Littleton , 414 U.S. 488, 499-502, 94 S.Ct. 669, 38 L.Ed.2d 674 (1974) (noting that the enforcement of the improper injunction in question required "continuous supervision by the federal court over the conduct of the petitioners in the course of future criminal trial proceedings involving any of the members of the respondents' broadly defined class"). Such relief does not implicate our concerns for comity and federalism.3

C. The County's Eighth Amendment Argument

The County contends that ODonnell's complaint "is an Eighth Amendment case wearing a Fourteenth Amendment costume." The Eighth Amendment states in relevant part that "[e]xcessive bail shall not be required." U.S. CONST. amend. VIII. It is certainly true that, when a constitutional provision specifically addresses a given claim for relief under 42 U.S.C. § 1983, a party should seek to apply that provision directly. See Graham v. Connor , 490 U.S. 386, 394, 109 S.Ct. 1865, 104 L.Ed.2d 443 (1989) ; cf. Manuel v. City of Joliet , --- U.S. ----, 137 S.Ct. 911, 917, 197 L.Ed.2d 312 (2017). But we have already concluded that "[t]he incarceration of those who cannot [pay money bail], without meaningful consideration of other possible alternatives, infringes on both due process and equal protection requirements." Pugh v. Rainwater , 572 F.2d 1053, 1057 (5th Cir. 1978) (en banc). ODonnell's present claims do not run afoul of Graham .

IV.

We now address the merits of ODonnell's constitutional claims. For the reasons set forth below, we affirm the court's rulings that the County's bail system violates both due process and equal protection, though we modify the basis for its conclusion as to due process.

A. Due Process Claim

Procedural due process claims are subject to a two-step inquiry: "The first question asks whether there exists a liberty or property interest which has been interfered with by the State; the second examines whether the procedures attendant upon that deprivation were constitutionally sufficient." Meza v. Livingston , 607 F.3d 392, 399 (5th Cir. 2010) (internal quotation marks omitted). Applying this framework, we disagree with the district court's formulation of the liberty interest created by state law, but agree that the procedural protections of bail-setting procedures are nevertheless constitutionally deficient.

Liberty interests protected by the due process clause can arise from two sources, "the Due Process Clause itself and the laws of the States." Ky. Dep't of Corr. v. Thompson , 490 U.S. 454, 460, 109 S.Ct. 1904, 104 L.Ed.2d 506 (1989) (internal citation omitted). Here, our focus is the law of Texas, which has acknowledged the two-fold, conflicting purpose of bail. This tension defines the protected liberty interest at issue here.

*158On the one hand, bail is meant "to secure the presence of the defendant in court at his trial." Ex parte Vance , 608 S.W.2d 681, 683 (Tex. Crim. App. 1980). Accordingly, "ability to make bail is a factor to be considered, [but] ability alone, even indigency, does not control the amount of bail." Ex parte Charlesworth , 600 S.W.2d 316, 317 (Tex. Crim. App. 1980). On the other hand, Texas courts have repeatedly emphasized the importance of bail as a means of protecting an accused detainee's constitutional right "in remaining free before trial," which allows for the "unhampered preparation of a defense, and ... prevent[s] the infliction of punishment prior to conviction." Ex parte Anderer , 61 S.W.3d 398, 404-05 (Tex. Crim. App. 2001) (en banc). Accordingly, the courts have sought to limit the imposition of "preventive [pretrial] detention" as "abhorrent to the American system of justice." Ex parte Davis , 574 S.W.2d 166, 169 (Tex. Crim. App. 1978). Notably, state courts have recognized that "the power to ... require bail," not simply the denial of bail, can be an "instrument of [such] oppression." Taylor v. State , 667 S.W.2d 149, 151 (Tex. Crim. App. 1984) (en banc) (emphasis added).

These protections are also ensconced in the Texas Constitution. Specifically, Article 1 § 11 reads in relevant part, "[a]ll prisoners shall be bailable by sufficient sureties." TEX. CONST. art. 1, § 11. The provision is followed by a list of exceptions-i.e., circumstances in which an arrestee may be "denied release on bail." Id. §§ 11b, 11c. The only exception tied to misdemeanor charges pertains to family violence offenses. See id. § 11c. The scope of these exceptions has been carefully limited by state courts, which observe that they "include the seeds of preventive detention." Davis , 574 S.W.2d at 169.

The district court held that § 11 creates a state-made "liberty interest in misdemeanor defendants' release from custody before trial. Under Texas law, judicial officers ... have no authority or discretion to order pretrial preventive detention in misdemeanor cases." This is too broad a reading of the law. The Constitution creates a right to bail on "sufficient sureties," which includes both a concern for the arrestee's interest in pretrial freedom and the court's interest in assurance. Since bail is not purely defined by what the detainee can afford, see Charlesworth , 600 S.W.2d at 317, the constitutional provision forbidding denial of release on bail for misdemeanor arrestees does not create an automatic right to pretrial release.4

Instead, Texas state law creates a right to bail that appropriately weighs the detainees' interest in pretrial release and the court's interest in securing the detainee's attendance. Yet, as noted, state law forbids the setting of bail as an "instrument of oppression." Thus, magistrates may not impose a secured bail solely for the purpose of detaining the accused. And, when the accused is indigent, setting a secured bail will, in most cases, have the same effect as a detention order. Accordingly, such decisions must reflect a careful weighing of the individualized factors set forth by both the state Code of Criminal Procedure and Local Rules.

Having found a state-created interest, we turn now to whether the procedures in place adequately protect that interest. As always, we are guided by a *159three-part balancing test that looks to "the private interest ... affected by the official action"; "the risk of an erroneous deprivation of such interest through the procedure used, and the probable value, if any, of additional or substitute procedural safeguards"; and "the Government's interest, including the function involved and the fiscal and administrative burdens" that new procedures would impose. Meza , 607 F.3d at 402 (quoting Mathews v. Eldridge , 424 U.S. 319, 335, 96 S.Ct. 893, 47 L.Ed.2d 18 (1976) ).

As the district court found, the current procedures are inadequate-even when applied to our narrower understanding of the liberty interest at stake. The court's factual findings (which are not clearly erroneous) demonstrate that secured bail orders are imposed almost automatically on indigent arrestees. Far from demonstrating sensitivity to the indigent misdemeanor defendants' ability to pay, Hearing Officers and County Judges almost always set a bail amount that detains the indigent. In other words, the current procedure does not sufficiently protect detainees from magistrates imposing bail as an "instrument of oppression."

The district court laid out specific procedures necessary to satisfy constitutional due process when setting bail. Specifically, it found that,

Due process requires: (1) notice that the financial and other resource information Pretrial Services officers collect is for the purpose of determining a misdemeanor arrestee's eligibility for release or detention; (2) a hearing at which the arrestee has an opportunity to be heard and to present evidence; (3) an impartial decisionmaker; (4) a written statement by the factfinder as to the evidence relied on to find that a secured financial condition is the only reasonable way to assure the arrestee's appearance at hearings and law-abiding behavior before trial; and (5) timely proceedings within 24 hours of arrest.5

The County challenges these requirements on appeal. We find some of their objections persuasive.

As this court has noted, the quality of procedural protections owed a defendant is evaluated on a "spectrum" based on a case-by-case evaluation of the liberty interests and governmental burdens at issue. Meza , 607 F.3d at 408-09. We note that the liberty interest of the arrestees here are particularly important: the right to pretrial liberty of those accused (that is, presumed innocent) of misdemeanor crimes upon the court's receipt of reasonable assurance of their return. See id. So too, however, is the government's interest in efficiency. After all, the accused also stands to benefit from efficient processing because it "allow[s] [for his or her] expeditious release." United States v. Chagra , 701 F.2d 354, 363 (5th Cir. 1983) ; cf. Cty. of Riverside v. McLaughlin , 500 U.S. 44, 53, 111 S.Ct. 1661, 114 L.Ed.2d 49 (1991) (noting that defendants might be disserved by adding procedural complexity into an already complicated system). The sheer number of bail hearings in Harris County each year-according to the court, over 50,000 people were arrested on misdemeanor charges in 2015-is a significant factor militating against overcorrection.

*160With this in mind, we make two modifications to the district court's conclusions regarding the procedural floor. First, we do not require factfinders to issue a written statement of their reasons. While we acknowledge "the provision for a written record helps to insure that [such officials], faced with possible scrutiny by state officials ... [and] the courts ... will act fairly," Wolff v. McDonnell , 418 U.S. 539, 565, 94 S.Ct. 2963, 41 L.Ed.2d 935 (1974), such a drastic increase in the burden imposed upon Hearing Officers will do more harm than good. We decline to hold that the Constitution requires the County to produce 50,000 written opinions per year to satisfy due process. Cf. United States v. McConnell , 842 F.2d 105, 110 (5th Cir. 1988) (concluding that, under the Bail Reform Act of 1984, the "court must [merely] explain its reasons for concluding that the particular financial requirement is a necessary part of the conditions for release" when setting a bond that a detainee cannot pay). Moreover, since the constitutional defect in the process afforded was the automatic imposition of pretrial detention on indigent misdemeanor arrestees, requiring magistrates to specifically enunciate their individualized, case-specific reasons for so doing is a sufficient remedy.

Second, we find that the district court's 24-hour requirement is too strict under federal constitutional standards. The court's decision to impose a 24-hour limit relied not on an analysis of present Harris County procedures and their current capacity; rather, it relied on the fact that a district court imposed this requirement thirty years ago (that is, prior to modern advancements in computer and communications technology). See Sanders v. City of Hous. , 543 F.Supp. 694 (S.D. Tex. 1982). But Sander s's holding, which was not grounded in procedural due process but in the Fourth Amendment, relied on the Supreme Court opinion, Gerstein , 420 U.S. 103, 95 S.Ct. 854, 43 L.Ed.2d 54. Id. at 699. And Gerstein was later interpreted as establishing a right to a probable cause hearing within 48 hours. McLaughlin , 500 U.S. at 56-57, 111 S.Ct. 1661. Further, McLaughlin explicitly included bail hearings within this deadline. Id. at 58, 111 S.Ct. 1661.

We conclude that the federal due process right entitles detainees to a hearing within 48 hours. Our review of the due process right at issue here counsels against an expansion of the right already afforded detainees under the Fourth Amendment by McLaughlin . We note in particular that the heavy administrative burden of a 24-hour requirement on the County is evidenced by the district court's own finding: the fact that 20% of detainees do not receive a probable cause hearing within 24 hours despite the statutory requirement. Imposing the same requirement for bail would only exacerbate such issues.

The court's conclusion was also based on its interpretation of state law. But while state law may define liberty interests protected under the procedural due process clause, it does not define the procedure constitutionally required to protect that interest. See Wansley v. Miss. Dep't of Corr. , 769 F.3d 309, 313 (5th Cir. 2014) (noting that state law cannot serve as "the source of ... process due"); Giovanni v. Lynn , 48 F.3d 908, 912 (5th Cir. 1995) ("[W]here a liberty ... interest is infringed, the process which is due under the United States Constitution is that measured by the due process clause, not that called for by state regulations. Mere failure to accord the procedural protections called for by state law or regulation does not of itself amount to a denial of due process." (internal citation omitted) ). Accordingly, although the parties contest *161whether state law imposes a 24- or 48-hour requirement, we need not resolve this issue because state law procedural requirements do not impact our federal due process analysis.

The district court's definition of ODonnell's liberty interests is too broad, and the procedural protections it required are too strict. Nevertheless, even under our more forgiving framework, we agree that the County procedures violate ODonnell's due process rights.

B. Equal Protection

The district court held that the County's bail-setting procedures violated the equal protection clause of the Fourteenth Amendment because they treat otherwise similarly-situated misdemeanor arrestees differently based solely on their relative wealth. The County makes three separate arguments against this holding. It argues: (1) ODonnell's disparate impact theory is not cognizable under the equal protection clause, see Johnson v. Rodriguez , 110 F.3d 299, 306 (5th Cir. 1997) ; (2) rational basis review applies and is satisfied; (3) even if heightened scrutiny applies, it is satisfied. We disagree.

First, the district court did not conclude that the County policies and procedures violated the equal protection clause solely on the basis of their disparate impact. Instead, it found the County's custom and practice purposefully "detain[ed] misdemeanor defendants before trial who are otherwise eligible for release, but whose indigence makes them unable to pay secured financial conditions of release." The conclusion of a discriminatory purpose was evidenced by numerous, sufficiently supported factual findings, including direct evidence from bail hearings. This custom and practice resulted in detainment solely due to a person's indigency because the financial conditions for release are based on predetermined amounts beyond a person's ability to pay and without any "meaningful consideration of other possible alternatives." Rainwater , 572 F.2d at 1057. Under this circuit's binding precedent, the district court was therefore correct to conclude that this discriminatory action was unconstitutional. Id. at 1056-57 (noting that pretrial "imprisonment solely because of indigent status is invidious discrimination and not constitutionally permissible" under both "due process and equal protection requirements"); see also Griffin v. Illinois , 351 U.S. 12, 18, 76 S.Ct. 585, 100 L.Ed. 891 (1956) (noting that the indigent are protected by equal protection "at all stages of [criminal] proceedings"). Because this conclusion is sufficient to decide this case, we need not determine whether the equal protection clause requires a categorical bar on secured money bail for indigent misdemeanor arrestees who cannot pay it.

Second, the district court's application of intermediate scrutiny was not in error. It is true that, ordinarily, "[n]either prisoners nor indigents constitute a suspect class." Carson v. Johnson , 112 F.3d 818, 821-22 (5th Cir. 1997). But the Supreme Court has found that heightened scrutiny is required when criminal laws detain poor defendants because of their indigence. See, e.g. , Tate v. Short , 401 U.S. 395, 397-99, 91 S.Ct. 668, 28 L.Ed.2d 130 (1971) (invalidating a facially neutral statute that authorized imprisonment for failure to pay fines because it violated the equal protection rights of indigents); Williams v. Illinois , 399 U.S. 235, 241-42, 90 S.Ct. 2018, 26 L.Ed.2d 586 (1970) (invalidating a facially neutral statute that required convicted defendants to remain in jail beyond the maximum sentence if they could not pay other fines associated with their sentences because it violated the equal protection rights of indigents). Reviewing this case law, the Supreme Court *162later noted that indigents receive a heightened scrutiny where two conditions are met: (1) "because of their impecunity they were completely unable to pay for some desired benefit," and (2) "as a consequence, they sustained an absolute deprivation of a meaningful opportunity to enjoy that benefit." San Antonio Indep. Sch. Dist. v. Rodriguez , 411 U.S. 1, 20, 93 S.Ct. 1278, 36 L.Ed.2d 16 (1973).

We conclude that this case falls into the exception created by the Court. Both aspects of the Rodriguez analysis apply here: indigent misdemeanor arrestees are unable to pay secured bail, and, as a result, sustain an absolute deprivation of their most basic liberty interests-freedom from incarceration. Moreover, this case presents the same basic injustice: poor arrestees in Harris County are incarcerated where similarly situated wealthy arrestees are not, solely because the indigent cannot afford to pay a secured bond. Heightened scrutiny of the County's policy is appropriate.6

Third, we discern no error in the court's conclusion that the County's policy failed to meet the tailoring requirements of intermediate scrutiny. In other words, we will not disturb the court's finding that, although the County had a compelling interest in the assurance of a misdemeanor detainee's future appearance and lawful behavior, its policy was not narrowly tailored to meet that interest.

The court's thorough review of empirical data and studies found that the County had failed to establish any "link between financial conditions of release and appearance at trial or law-abiding behavior before trial." For example, both parties' experts agreed that the County lacked adequate data to demonstrate whether secured bail was more effective than personal bonds in securing a detainee's future appearance. Notably, even after analyzing the incomplete data that were available, neither expert discerned more than a negligible comparative impact on detainees' attendance. Additionally, the court considered a comprehensive study of the impact of Harris County's bail system on the behavior of misdemeanor detainees between 2008 and 2013. The study found that the imposition of secured bail might increase the likelihood of unlawful behavior. See Paul Heaton et al., The Downstream Consequences of Misdemeanor Pretrial Detention , 69 STAN. L. REV. 711, 786-87 (2017) (estimating that the release on personal bond of the lowest-risk detainees would have resulted in 1,600 fewer felonies and 2,400 fewer misdemeanors within the following eighteen months). These findings mirrored those of various empirical studies from other jurisdictions.

The County, of course, challenges these assertions with empirical studies of its own. But its studies at best cast some doubt on the court's conclusions. They do not establish clear error. We are satisfied that the court had sufficient evidence to *163conclude that Harris County's use of secured bail violated equal protection.

In sum, the essence of the district court's equal protection analysis can be boiled down to the following: take two misdemeanor arrestees who are identical in every way-same charge, same criminal backgrounds, same circumstances, etc.-except that one is wealthy and one is indigent. Applying the County's current custom and practice, with their lack of individualized assessment and mechanical application of the secured bail schedule, both arrestees would almost certainly receive identical secured bail amounts. One arrestee is able to post bond, and the other is not. As a result, the wealthy arrestee is less likely to plead guilty, more likely to receive a shorter sentence or be acquitted, and less likely to bear the social costs of incarceration. The poor arrestee, by contrast, must bear the brunt of all of these, simply because he has less money than his wealthy counterpart. The district court held that this state of affairs violates the equal protection clause, and we agree.

V.

Having largely affirmed the district court's determinations that constitutional violations occurred, we turn to the court's remedy. When crafting an injunction, district courts are guided by the Supreme Court's instruction that "the scope of injunctive relief is dictated by the extent of the violation established." Califano v. Yamasaki , 442 U.S. 682, 702, 99 S.Ct. 2545, 61 L.Ed.2d 176 (1979). A district court abuses its discretion if it does not "narrowly tailor an injunction to remedy the specific action which gives rise to the order." John Doe # 1 v. Veneman , 380 F.3d 807, 818 (5th Cir. 2004). Thus, an injunction must be vacated if it "fails to meet these standards" and "is overbroad." Id. "The broadness of an injunction refers to the range of proscribed activity .... [and] is a matter of substantive law." U.S. Steel Corp. v. United Mine Workers of Am. , 519 F.2d 1236, 1246 n.19 (5th Cir. 1975).

The County argues that, even if the panel credits every one of the district court's factual findings and conclusions of law, the injunction it ultimately crafted is still overbroad. We agree. There is a significant mismatch between the district court's procedure-focused legal analysis and the sweeping injunction it implemented.

The fundamental source of constitutional deficiency in the due process and equal protection analyses is the same: the County's mechanical application of the secured bail schedule without regard for the individual arrestee's personal circumstances. Thus, the equitable remedy necessary to cure the constitutional infirmities arising under both clauses is the same: the County must implement the constitutionally-necessary procedures to engage in a case-by-case evaluation of a given arrestee's circumstances, taking into account the various factors required by Texas state law (only one of which is ability to pay). These procedures are: notice, an opportunity to be heard and submit evidence within 48 hours of arrest, and a reasoned decision by an impartial decisionmaker.

That is not what the preliminary injunction does, however. Rather, it amounts to the outright elimination of secured bail for indigent misdemeanor arrestees. That remedy makes some sense if one assumes a fundamental substantive due process right to be free from any form of wealth-based detention. But, as the foregoing analysis establishes, no such right is in view. The sweeping injunction is overbroad.

*164We therefore conclude that the district court abused its discretion in crafting an injunction that was not "narrowly tailor[ed] ... to remedy the specific action which gives rise to the order." Veneman , 380 F.3d at 818. We will vacate the injunction and remand to allow the court to craft a remedy more finely tuned to address the harm.

The following represents the sort of modification that would be appropriate here, although we leave the details to the district court's discretion:

With these principles in mind, the court will order the following relief, to take effect within 30 days, unless those enjoined move for and show good cause for a reasonable, brief extension. Any motions for extension will be set for prompt hearing and resolution.

• Harris County is enjoined from imposing prescheduled bail amounts as a condition of release on arrestees who attest that they cannot afford such amounts without providing an adequate process for ensuring that there is individual consideration for each arrestee of whether another amount or condition provides sufficient sureties.

• Pretrial Services officers, as County employees and subject to its policies, must verify an arrestee's ability to pay a prescheduled financial condition of release by an affidavit, and must explain to arrestees the nature and significance of the verification process.

• The purpose of the explanation is to provide the notice due process requires that a misdemeanor defendant's state constitutional right to be bailable by sufficient sureties is at stake in the proceedings. Pretrial Services may administer either the form of the affidavit currently used to determine eligibility for appointed counsel or the adapted form that Dr. VanNostrand testified was prepared for Pretrial Services to be administered by July 1, 2017, if they comply with the below guidelines. Pretrial Services must deliver completed affidavits to the Harris County Sheriff's Office before a declarant's probable cause hearing.

• The affidavit must give the misdemeanor arrestee sufficient opportunity to declare under penalty of perjury, after the significance of the information has been explained, the maximum amount of financial security the arrestee would be able to post or pay up front within 24 hours of arrest. The affidavit should ask the arrestee to provide details about their financial situation sufficient to help the County make reliable determinations regarding the amount of bail that would provide sufficient sureties, including: 1) arrestee and spouse's income from employment, real property, interest and dividends, gifts, alimony, child support, retirement, disability, unemployment payments, public-assistance, and other sources; 2) arrestee and spouse's employment history for the prior two years and gross monthly pay; 3) arrestee and spouse's present cash available and any financial institutions where cash is held; 4) assets owned, e.g., real estate and motor vehicles; 5) money owed to arrestee and spouse; 6) dependents of arrestee and spouse, and their ages; 7) estimation of itemized monthly expenses; 8) taxes and legal costs; 9) expected major *165changes in income or expenses; 10) additional information the arrestee wishes to provide to help explain the inability to pay. The question is neither the arrestee's immediate ability to pay with cash on hand, nor what assets the arrestee could eventually produce after a period of pretrial detention. The question is what amount the arrestee could reasonably pay within 24 hours of his or her arrest, from any source, including the contributions of family and friends.

• The purpose of this requirement is to provide a better, easier, and faster way to get the information needed to determine a misdemeanor defendant's ability to pay. The Hearing Officers and County Judges testified that they presently do not know who has the ability to pay. The affidavit can be completed within 24 hours after arrest; the current process of verifying references by phone extends for days after arrest.

• The court does not order relief against the Hearing Officers or against the County Judges in their judicial or legislative capacities.

• Misdemeanor defendants who are not subject to: (1) formal holds preventing their release from detention; (2) pending mental-health evaluations to determine competency; or (3) pretrial preventive detention orders for violating a condition of release for a crime of family violence, have a constitutionally protected state-created liberty interest in being bailable by sufficient sureties before trial. If a misdemeanor defendant has executed an affidavit showing an inability to pay prescheduled money bail and has not been released either: (1) on an unsecured personal bond with nonfinancial conditions of release; or (2) on a secured money bond for which the defendant could pay a commercial surety's premium, as indicated on the affidavit, then the defendant is entitled to a hearing within 48 hours of arrest in which an impartial decision-maker conducts an individual assessment of whether another amount of bail or other condition provides sufficient sureties. At the hearing, the arrestee must have an opportunity to describe evidence in his or her favor, and to respond to evidence described or presented by law enforcement. If the decision-maker declines to lower bail from the prescheduled amount to an amount the arrestee is able to pay, then the decisionmaker must provide written factual findings or factual findings on the record explaining the reason for the decision, and the County must provide the arrestee with a formal adversarial bail review hearing before a County Judge. The Harris County Sheriff is therefore authorized to decline to enforce orders requiring payment of prescheduled bail amounts as a condition of release for said defendants if the orders are not accompanied by a record showing that the required individual assessment was made and an opportunity for formal review was provided. All nonfinancial conditions of release ordered by the Hearing Officers, including protective orders, drug testing, alcohol intake ignition locks, or GPS monitoring, will remain in effect.

*166• The purpose of this requirement is to provide timely protection for the state-created liberty interest in being bailable by sufficient sureties and to prevent the automatic imposition of prescheduled bail amounts without an adequate process for ensuring that there is individualized consideration of whether another amount or condition provides sufficient sureties.

• To enforce the 48-hour timeline, the County must make a weekly report to the district court of misdemeanor defendants identified above for whom a timely individual assessment has not been held. The County must also notify the defendant's counsel and/or next of kin of the delay. A pattern of delays might warrant further relief from the district court. Because the court recognizes that the County might need additional time to comply with this requirement, the County may propose a reasonable timeline for doing so.

• The purpose of this requirement is to give timely protection to the state-created liberty interest in being bailable by sufficient sureties by enforcing federal standards indicating that 48 hours is a reasonable timeframe for completing the administrative incidents to arrest. The 48-hour requirement is intended to address the endemic problem of misdemeanor arrestees being detained until case disposition and pleading guilty to secure faster release from pretrial detention.

• For misdemeanor defendants who are subject to formal holds and who have executed an affidavit showing an inability to pay the prescheduled financial condition of release, the Sheriff must treat the limitations period on their holds as beginning to run the earliest of: (1) after the probable cause hearing; or (2) 24 hours after arrest. The purpose of this requirement is to ensure that misdemeanor defendants are not prevented from or delayed in addressing their holds because they are indigent and therefore cannot pay a prescheduled financial condition of release.

• Misdemeanor defendants who do not appear competent to execute an affidavit may be evaluated under the procedures set out in the Texas Code of Criminal Procedure Article 16.22. If competence is found, the misdemeanor defendant is covered by the relief the court orders, with the exception that the 48-hour period begins to run from the finding of competence rather than from the time of arrest. As under Article 16.22, nothing in this order prevents the misdemeanor arrestee from being released on secured bail or unsecured personal bond pending the evaluation.

VI.

For the forgoing reasons, we AFFIRM the district court's findings of fact. We AFFIRM its conclusions of law except its conclusion that the County Sheriff qualifies as a municipal policymaker under § 1983 and its determination of the specific procedural protections owed under procedural due process. On those issues, we REVERSE the district court's conclusions. Accordingly, we VACATE the preliminary injunction as overbroad and REMAND to the district court to craft a revised injunction-one that is narrowly tailored to cure the constitutional deficiencies the district *167court properly identified. But we also STAY the vacatur pending implementation of the revised injunction, so as to maintain a stable status quo.