The question of jurisdiction was submitted to the court without argument, by P.B. Key, for the appellants, and Harper, for the appellees.

On a subsequent day,

MARSHALL, Ch. J. delivered the opinion of the court.

The court has considered this case, and is of opinion that the jurisdiction cannot be supported.

The words of the act of congress are, "where an alien is a party; or the suit is between a citizen of a state where the suit is brought, and a citizen of another state."

The court understands these expressions to meant that each distinct interest should be represented by persons, all of whom are entitled to sue, or may be sued, in the federal courts. That is, that where the interest is joint, each of the persons concerned in that interest must be competent to sue, or liable to be sued, in those courts.

But the court does not mean to give an opinion in the case where several parties represent several distinct interests, [268] and some of those parties are, and others are not, competent to sue, or liable to be sued, in the courts of the United States.

Decree affirmed.

[63] Binney, for the plaintiffs in error.

P.B. Key, contra.

[84] MARSHALL, Ch. J. delivered the opinion of the court as follows:

Two points have been made in this cause.

1. That a corporation, composed of citizens of [85] one state, may sue a citizen of another state, in the federal courts.

2. That a right to sue in those courts is conferred on this bank by the law which incorporates it.

The last point will be first considered.

The judicial power of the United States, as defined in the constitution, is dependent, 1st. On the nature of the case; and, 2d. On the character of the parties.

By the judicial act, the jurisdiction of the circuit courts is extended to cases where the constitutional right to plead and be impleaded, in the courts of the union, depends on the character of the parties; but where that right depends on the nature of the case, the circuit courts derive no jurisdiction from that act, except in the single case of a controversy between citizens of the same state, claiming lands under grants from different states.

Unless, then, jurisdiction over this cause has been given to the circuit court by some other than the judicial act, the bank of the United States had not a right to sue in that court, upon the principle that the case arises under a law of the United States.

The plaintiffs contend that the incorporating act confers this jurisdiction.

That act creates the corporation, gives it a capacity to make contracts and to acquire property, and enables it "to sue and be sued, plead and be impleaded, answer and be answered, defend and be defended, in courts of record, or any other place whatsoever."

This power, if not incident to a corporation, is conferred by every incorporating act, and is not understood to enlarge the jurisdiction of any particular court, but to give a capacity to the corporation to [86] appear, as a corporation, in any court which would, by law, have cognisance of the cause, if brought by individuals. If jurisdiction is given by this clause to the federal courts, it is equally given to all courts having original jurisdiction, and for all sums however small they may be.

But the 9th article of the 7th section of the act furnishes a conclusive argument against the construction for which the plaintiffs contend. That section subjects the president and directors, in their individual capacity, to the suit of any person aggrieved by their putting into circulation more notes than is permitted by law, and expressly authorizes the bringing of that action in the federal or state courts.

This evinces the opinion of congress, that the right to sue does not imply a right to sue in the courts of the union, unless it be expressed. This idea is strengthened also by the law respecting patent rights. That law expressly recognizes the right of the patentee to sue in the circuit courts of the United States.

The court, then, is of opinion, that no right is conferred on the bank, by the act of incorporation, to sue in the federal courts.

2. The other point is one of much more difficulty.

The jurisdiction of this court being limited, so far as respects the character of the parties in this particular case, "to controversies between citizens of different states," both parties must be citizens, to come within the description.

That invisible, intangible, and artificial being, that mere legal entity, a corporation aggregate, is certainly not a citizen; and, consequently, cannot sue or be sued in the courts of the United States, unless the rights of the members, in this respect, can be exercised in their corporate name. If the corporation [87] be considered as a mere faculty, and not as a company of individuals, who, in transacting their joint concerns, may use a legal name, they must be excluded from the courts of the union.

The duties of this court, to exercise jurisdiction where it is conferred, and not to usurp it where it is not conferred, are of equal obligation. The constitution, therefore, and the law, are to be expounded, without a leaning the one way or the other, according to those general principles which usually govern in the construction of fundamental or other laws.

A constitution, from its nature, deals in generals, not in detail. Its framers cannot perceive minute distinctions which arise in the progress of the nation, and therefore confine it to the establishment of broad and general principles.

The judicial department was introduced into the American constitution under impressions, and with views, which are too apparent not to be perceived by all. However true the fact may be, that the tribunals of the states will administer justice as impartially as those of the nation, to parties of every description, it is not less true that the constitution itself either entertains apprehensions on this subject, or views with such indulgence the possible fears and apprehensions of suitors, that it has established national tribunals for the decision of controversies between aliens and a citizen, or between citizens of different states. Aliens, or citizens of different states, are not less susceptible of these apprehensions, nor can they be supposed to be less the objects of constitutional provision, because they are allowed to sue by a corporate name. That name, indeed, cannot be an alien or a citizen; but the persons whom it represents may be the one or the other; and the controversy is, in fact and in law, between those persons suing in their corporate character, by their corporate name, for a corporate right, and the individual against whom the suit may be instituted. Substantially [88] and essentially, the parties in such a case, where the members of the corporation are aliens, or citizens of a different state from the opposite party, come within the spirit and terms of the jurisdiction conferred by the constitution on the national tribunals.

Such has been the universal understanding on the subject. Repeatedly has this court decided causes between a corporation and an individual without feeling a doubt respecting its jurisdiction. Those decisions are not cited as authority; for they were made without considering this particular point; but they have much weight, as they show that this point neither occurred to the bar or the bench; and that the common understanding of intelligent men is in favor of the right of incorporated aliens, or citizens of a different state from the defendant, to sue in the national courts. It is by a course of acute, metaphysical and abstruse reasoning, which has been most ably employed on this occasion, that this opinion is shaken.

As our ideas of a corporation, its privileges and its disabilities, are derived entirely from the English books, we resort to them for aid, in ascertaining its character. It is defined as a mere creature of the law, invisible, intangible, and incorporeal. Yet, when we examine the subject further, we find that corporations have been included within terms of description appropriated to real persons.

The statute of Henry VIII. concerning bridges and highways, enacts, that bridges and highways shall be made and repaired by the "inhabitants of the city, shire, or riding," and that the justices shall have power to tax every "inhabitant of such city," &c.; and that the collectors may "distrain every such inhabitant as shall be taxed and refuse payment thereof, in his lands, goods and chattels."

Under this statute those have been construed inhabitants who hold lands within the city where the [89] bridge to be repaired lies, although they reside elsewhere.

Lord Coke says, "every corporation and body politic residing in any county, riding, city, or town corporate, or having lands or tenements in any shire, quæ propriis manibus et sumptibus possident et habent, are said to be inhabitants there, within the purview of this statute."

The tax is not imposed on the person, whether he be a member of the corporation or not, who may happen to reside on the lands; but is imposed on the corporation itself, and, consequently, this ideal existence is considered as an inhabitant, when the general spirit and purpose of the law requires it.

In the case of The King v. Gardner, reported by Cowper, a corporation was decided, by the court of king's bench, to come within the description of "occupiers or inhabitants." In that case the poor rates, to which the lands of the corporation were declared to be liable, were not assessed to the actual occupant, for there was none, but to the corporation. And the principle established by the case appears to be, that the poor rates, on vacant ground belonging to a corporation, may be assessed to the corporation, as being inhabitants or occupiers of that ground. In this case Lord Mansfield notices and overrules an inconsiderate dictum of Justice Yates, that a corporation could not be an inhabitant or occupier.

These opinions are not precisely in point; but they serve to show that, for the general purposes and objects of a law, this invisible, incorporeal creature of the law may be considered as having corporeal qualities.

It is true that as far as these cases go they serve to show that the corporation itself, in its incorporeal character, may be considered as an inhabitant or an occupier; and the argument from them would be more strong in favour of considering the corporation [90] itself as endowed for this special purpose with the character of a citizen, than to consider the character of the individuals who compose it as a subject which the court can inspect, when they use the name of the corporation, for the purpose of asserting their corporate rights. Still the cases show that this technical definition of a corporation does not uniformly circumscribe its capacities, but that courts for legitimate purposes will contemplate it more substantially.

There is a case, however, reported in 12 Mod. which is thought precisely in point. The corporation of London brought a suit against Wood, by their corporate name, in the mayor's court. The suit was brought by the mayor and commonalty, and was tried before the mayor and aldermen. The judgment rendered in this cause was brought before the court of king's bench and reversed, because the court was deprived of its jurisdiction by the character of the individuals who were members of the corporation.

In that case the objection, that a corporation was an invisible, intangible thing, a mere incorporeal legal entity, in which the characters of the individuals who composed it were completely merged, was urged and was considered. The judges unanimously declared that they could look beyond the corporate name, and notice the character of the individual. In the opinions, which were delivered seriatim, several cases are put which serve to illustrate the principle, and fortify the decision.

The case of The Mayor and Commonalty v. Wood, is the stronger, because it is on the point of jurisdiction. It appears to the court to be a full authority for the case now under consideration. It seems not possible to distinguish them from each other.

If, then, the congress of the United States had, in terms, enacted that incorporated aliens might suc [91] a citizen, or that the incorporated citizens of one state might sue a citizen of another state, in the federal courts, by its corporate name, this court would not have felt itself justified in declaring that such a law transcended the constitution.

The controversy is substantially between aliens, suing by a corporate name, and a citizen, or between citizens of one state, suing by a corporate name, and those of another state. When these are said to be substantially the parties to the controversy, the court does not mean to liken it to the case of a trustee. A trustee is a real person capable of being a citizen or an alien, who has the whole legal estate in himself. At law, he is the real proprietor, and he represents himself, and sues in his own right. But in this case the corporate name represents persons who are members of the corporation.

If the constitution would authorize congress to give the courts of the union jurisdiction in this case, in consequence of the character of the members of the corporation, then the judicial act ought to be construed to give it. For the term citizen ought to be understood as it is used in the constitution, and as it is used in other laws. That is, to describe the real persons who come into court, in this case, under their corporate name.

That corporations composed of citizens are considered by the legislature as citizens, under certain circumstances, is to be strongly inferred from the registering act. It never could be intended that an American registered vessel, abandoned to an insurance company composed of citizens, should lose her character as an American vessel; and yet this would be the consequence of declaring that the members of the corporation were, to every intent and purpose, out of view, and merged in the corporation.

The court feels itself authorized by the case in 12 Mod. on a question of jurisdiction, to look to [92] the character of the individuals who compose the corporation, and they think that the precedents of this court, though they were not decisions on argument, ought not to be absolutely disregarded.

If a corporation may sue in the courts of the union, the court is of opinion that the averment in this case is sufficient.

Being authorized to sue in their corporate name, they could make the averment, and it must apply to the plaintiffs as individuals, because it could not be true as applied to the corporation.

Judgment reversed; plea in abatement overruled, and cause remanded.

Judge LIVINGSTON, having an interest in the question, gave no opinion.

[83] [84] Alan M. Dubow, White Plains, N.Y. (Bruce Minkoff, Robinowitz Cohlan & Dubow, White Plains, N.Y., on the brief), for plaintiff-appellant.

Larry H. Krantz, New York City (Ronald W. Meister, Meister Leventhal & Slade, New York City, George Forman, Alexander & Karshmer, Berkeley, Cal., on brief), for defendant-appellee.

Before NEWMAN and KEARSE, Circuit Judges, and STANTON, District Judge.^[1]

KEARSE, Circuit Judge:

Plaintiff A.F.A. Tours, Inc., doing business as Alumni Flights Abroad ("AFA"), appeals from a final judgment of the United States District Court for the Southern District of New York, Vincent L. Broderick, Judge, dismissing for lack of subject matter jurisdiction this diversity action against defendant Desmond Whitchurch for misappropriation of trade secrets. The district court summarily dismissed the complaint on the ground that it would not be possible for AFA to prove damages amounting to more than $50,000. For the reasons below, we vacate and remand for further proceedings.

I. BACKGROUND

According to the complaint, AFA operates a travel and tour business, specializing in deluxe tours for United States residents to overseas destinations including Australia, New Zealand, and New Guinea. It expended large sums of money and invested significant time and labor to develop, inter alia, a client and customer list, marketing information, and tour information. It regarded this information as confidential trade secrets.

From 1972 through 1989, Whitchurch was employed by AFA as its exclusive tour escort in the above areas. In that position, Whitchurch was privy to certain of the above confidential information. The complaint alleged that in or about October 1989, Whitchurch resigned from AFA, misappropriated the confidential information known to him, and organized his own tour business. Since that time, he has offered or intends to offer tours that compete with those offered by AFA; in connection with his own tours, he has solicited or intends to solicit participants from AFA's customer list.

AFA commenced the present diversity action in the district court for misappropriation of its trade secrets, seeking an injunction against any use by Whitchurch of confidential AFA information, and damages "in an amount which is not presently ascertainable, but which is believed to exceed the sum of $50,000.00." It also sought punitive damages of "no less than $250,000.00."

Whitchurch denied all of the material allegations of the complaint and quickly moved for summary judgment on the ground that the AFA information possessed by Whitchurch was not confidential. Whitchurch's supporting affidavits stated, inter alia, that, though Whitchurch had led an average of seven AFA tours per year for 17 years, he had had no written contract with AFA. He stated that AFA had never informed him that the customer names were confidential, or should not be [85] shared with others, or should be returned to AFA upon Whitchurch's resignation. Rather, he stated, AFA had freely disseminated lists of the names and addresses of its customers to "countless individuals and entities," including hotels, booking and travel agents, and the tour participants themselves.

Whitchurch stated that from February to May 1990, he had attempted to organize a tour on his own. In this effort he had, inter alia, written to former AFA tour participants he had escorted, informed them of his separation from AFA, and solicited their participation. Only two people reserved space on the planned tour, however, and it was canceled. He opposed the present lawsuit because he remained interested in conducting tours in the future.

AFA opposed Whitchurch's summary judgment motion, contending that there were genuine issues of fact as to, inter alia, the confidential nature of the AFA information and Whitchurch's knowledge of that confidentiality. It submitted affidavits describing its own operations and detailing steps it had taken to maintain the confidentiality of its customer list, including the following.

The names and addresses of AFA customers were not readily obtainable from any public source. AFA catered to alumni and alumnae of certain United States universities that did not generally sell their alumni lists to outside businesses for commercial purposes. AFA's customer list, maintained in a computer file to which access is denied without a password, was compiled largely from responses to AFA advertisements in alumni magazines and from referrals by AFA tour participants. AFA was currently spending more than $100,000 per year in advertising and promotion and in all had spent more than $1,000,000 in developing the customer list. The most valuable names on the list were those of persons who had traveled on past AFA tours.

AFA disputed Whitchurch's contention that its customer list had been freely disseminated, stating that the names on that list were not disclosed except on a need-to-know basis. Thus, Whitchurch himself had not been given the whole list but only the names and addresses of the persons he was to escort. Lists of persons participating in a particular tour obviously had to be disclosed to providers of services such as hotels, but the disclosures were limited and were made with a notation of confidentiality. AFA also asserted that Whitchurch was fully aware of the confidential nature of that information and that as part of his duties he had delivered preregistration information to hotels, including an AFA cover letter stating that the information provided "is STRICTLY CONFIDENTIAL and is supplied to you only for the purpose of pre-registering the members of the tour."

AFA never sold or traded its list to others. Occasionally it received calls from persons considering a tour and requesting the names and addresses of former AFA tour participants; it refused to give out that information, informing the caller that the information was confidential. Even when a former AFA tour participant would call AFA to request the address or telephone number of a coparticipant, AFA's response was to relay the request to the coparticipant rather than giving the information to the caller. These practices were designed to prevent rival tours from gaining AFA confidential information by ruse.

At the oral argument of Whitchurch's motion, Whitchurch's attorney began by characterizing the motion as "turn[ing] on a very narrow issue" of whether the AFA information constituted trade secrets (Hearing Transcript, November 16, 1990 ("Tr."), at 1), and stating that the case was important to Whitchurch because "he may one day in the future wish to organize a tour, and write letters to individuals again" (id. at 2). The court, however, asked whether it even had to reach the question of trade secrets, raising sua sponte the question of whether the value of AFA's claims exceeded $50,000, a jurisdictional prerequisite for a diversity action. In response to the court's jurisdictional question, Whitchurch's attorney stated that Whitchurch, in soliciting for his planned tour, had written to 100-200 former AFA [86] tour participants but had received favorable responses from only two.

THE COURT: .... Why did you think your client was so unsuccessful in sparking interest in a tour under his direction?

MR. KRANTZ [counsel for Whitchurch]: I am sure one factor was that he was writing to people who had already been there, and so he was looking for only people who were interested in going twice. What other factors there were, I am really not certain....

Certainly any damages here are d[e] min[imi]s. I would absolutely agree with the Court.

THE COURT: Well, the Supreme Court has told us, as I read what it has said, that if there is a speculative question of fact but no reasonable possibility of damage, then summary judgement is appropriate....

(Id. at 5.)

AFA's attorney argued that Whitchurch's lack of success on his first effort was hardly dispositive of the issue of the amount of damages AFA might suffer, in light of Whitchurch's desire to conduct other tours in the future. In his letter to former AFA tour participants, for example, Whitchurch had offered the addressees their "first" opportunity to join his planned 28-day tour and said, "If you would like to be on my mailing list for future tours, please fill out and return the enclosed form. I will also be available to arrange special interest tours to the South Pacific." AFA's attorney, noting that over the years Whitchurch had escorted some 1,500 AFA clients on tours and had indicated that he would conduct a number of tours, argued that AFA's damages would be substantial. He stated that a single 10-customer tour to the area in question would easily generate more than $50,000:

THE COURT: How do you get up to fifty thousand dollars?

MR. DUBOW: ....

One tour of ten of your customers is a lot more than fifty thousand dollars. It's about seven to ten thousand dollars per customer on that one tour, and he's offering five or six different tours to areas of the world. And he even said in his letter that he is going to solicit customers for trips to that area of the world and expand on that. So there is no doubt, he doesn't deny that.

(Id. at 15.) Dubow also stated that "one tour alone will produce more than fifty thousand dollars to Mr. Whitchurch...." (Id. at 19.)

At the close of this hearing, the court granted summary judgment in favor of Whitchurch, stating as follows:

THE COURT: I grant defendant's motion for summary judgment. There may be a theoretical question of fact here, but it is not the same type of material question of fact that precludes summary judgment.

We have lists here that were available to the defendant for the seventeen years, with respect to some of the older witnesses [sic], I suspect that the people traveled on those lists[, m]any of whom have reached retirement age at that time are probably no longer available for touring purposes.

The defendant's track record in soliciting is a pretty bad one. Two people signed up with him. I see there is no way any fact finder in this case will possibly reach a point where it will award damages to the plaintiff. I see no possible reason that the fifty thousand — no possible basis upon which the fifty thousand dollar number, which I think does apply here, could ever be reached. We're not dealing here with the plaintiff's massive lists. We're dealing with a circumscribed portion of those lists where the customers have already been tapped and have taken their tours.

I have [a] very serious question as to whether the lists involved here are trade secrets. It maybe [sic], as the plaintiff argues, that this is a question of fact that I should not be deciding. But, I'm taking to heart what the Supreme Court has said and that is a summary judgment is an effective way to eliminate from the Court the matters cluttering up our calendars.

I see this case is one where there is no reasonable possibility that any jury will [87] render a verdict for the plaintiff, and on that basis, I grant summary judgement.

(Id. at 33-35.)

Judgment was entered dismissing the complaint, and this appeal followed.

II. DISCUSSION

On appeal, AFA contends that the dismissal for lack of jurisdiction was improper because the court (1) failed to give AFA an appropriate opportunity to show that it satisfied the jurisdictional amount, and (2) failed to apply the proper standard to AFA's requests for (a) damages and (b) injunctive relief. Whitchurch seeks to defend the judgment on the ground that the district court did not dismiss on the jurisdictional ground and that summary judgment on the merits was proper.

We reject Whitchurch's contention that the district court did not dismiss for lack of jurisdiction. Though the court styled its decision as one granting summary judgment and stated that it had a "very serious question" as to whether the AFA lists were trade secrets, it did not purport to resolve that question. Rather, it stated that it could see "no possible basis upon which the fifty thousand dollar number could ever be reached." The decision itself, therefore, especially in light of the colloquy that preceded it, indicates that the court dismissed because it had concluded that AFA could not satisfy the minimum jurisdictional amount.

For the reasons below, we conclude that the dismissal on the jurisdictional ground was improper. We also note that summary judgment dismissing AFA's trade secrets claims on the merits would have been improper.

A. The Jurisdictional Amount

The district courts have jurisdiction over civil diversity suits "where the matter in controversy exceeds the sum or value of $50,000, exclusive of interest and costs." 28 U.S.C. § 1332 (1988). The test for determining whether a plaintiff meets the jurisdictional amount, established by the Supreme Court in St. Paul Mercury Indemnity Co. v. Red Cab Co., 303 U.S. 283, 58 S.Ct. 586, 82 L.Ed. 845 (1938), is as follows:

The rule governing dismissal for want of jurisdiction in cases brought in the federal court is that, unless the law gives a different rule, the sum claimed by the plaintiff controls if the claim is apparently made in good faith. It must appear to a legal certainty that the claim is really for less than the jurisdictional amount to justify a dismissal.

303 U.S. at 288-89, 58 S.Ct. at 590 (emphasis added).

The amount of damages recoverable in an action for misappropriation of trade secrets may be measured either by the plaintiff's losses, see, e.g., Timely Products Corp. v. Arron, 523 F.2d 288, 304 (2d Cir.1975); see generally 2 R. Milgrim, Milgrim on Trade Secrets ("Milgrim") § 7.08[3][a], at 7-314 to 7-318 (1990), or by the profits unjustly received by the defendant, see id.; Electro-Miniatures Corp. v. Wendon Co., 771 F.2d 23, 27 (2d Cir.1985). In addition, if punitive damages are permitted under the controlling law, the demand for such damages may be included in determining whether the jurisdictional amount is satisfied. See generally 14A C. Wright & A. Miller, Federal Practice and Procedure § 3702, at 44 (1985) ("Wright & Miller"). New York law apparently allows the recovery of punitive damages in a trade secrets case if the defendant's conduct has been sufficiently "gross and wanton." See, e.g., Huschle v. Battelle, 33 A.D.2d 1017, 308 N.Y.S.2d 235 (1st Dep't 1970), aff'd, 31 N.Y.2d 767, 338 N.Y.S.2d 622, 290 N.E.2d 823 (1972).

Further, in appropriate circumstances, the owner of trade secrets may obtain an injunction against their use or disclosure by another in breach of his confidential relationship with the owner. See Bridge C.A.T. Scan Associates v. Technicare Corp., 710 F.2d 940, 946 (2d Cir.1983). Where the plaintiff seeks injunctive relief, the value of his claim is generally assessed with reference to the right he seeks to protect and measured by the extent of the impairment to be prevented by the injunction. See generally 1 Moore's Federal [88] Practice ¶ 0.96[2] (2d ed. 1991); 14A Wright & Miller § 3708, at 143-44. In calculating that impairment, the court may look not only at past losses but also at potential harm. See generally id. at 146-49.

Before making a determination that the plaintiff's claim does not meet the jurisdictional minimum, the court must afford the plaintiff an "appropriate and reasonable opportunity to show good faith in believing that a recovery in excess of [the jurisdictional amount] is reasonably possible." Arnold v. Troccoli, 344 F.2d 842, 846 (2d Cir.1965). Under these substantive and procedural principles, although the record indicates that AFA has not yet suffered actual damages even approaching $50,000, we have difficulty with the district court's decision.

First, though AFA did not make an evidentiary showing in support of its contention that the value of its claims exceeded $50,000, it was not afforded a proper opportunity to do so. The issue of the jurisdictional amount was first raised by the district court sua sponte at the argument on the summary judgment motion, and the court rendered its decision at the end of that argument. To the extent that the court thought AFA could not meet the jurisdictional minimum, it should not have dismissed without giving AFA an opportunity to present substantiation directed toward that issue.

Second, despite AFA's lack of an opportunity to present evidence addressed directly to the jurisdictional question, there was evidence in the record to suggest that the matter could not be conclusively resolved against it, for the oral arguments made by AFA's attorney to show that its claims were worth more than $50,000 had some support from documents already before the court. For example, he said Whitchurch had the names of some 1,500 AFA customers; this was consistent with (a) Whitchurch's own statement that in the 17 years he was employed by AFA he had led approximately seven tours each year (thus totaling some 119 tours) and (b) his attorney's statement that there were usually 10-15 people per tour. If there were no repeat customers among the persons Whitchurch escorted, he could have escorted some 1,785 AFA tour participants (119 × 15). As to AFA's attorney's estimate that a single tour of this type "[i]s about seven to ten thousand dollars per customer," the evidence in the record as to the destination and deluxe nature of the tours, including evidence that some participants traveled first class and reserved preferred hotel accommodations, supports an inference that a 28-day tour could well cost $10,000 per person.

What this means in terms of loss of earnings to a tour operator, however, is not revealed by the present record. AFA's suggestion that the tour operator himself would earn $10,000 per tourist (i.e., that such a tour would "produce more than fifty thousand dollars to Mr. Whitchurch" (Tr. 19 (emphasis added))) does not have the same record support and seems questionable. The district court was also undoubtedly correct in its assumption that many of the persons who traveled to the South Pacific with Whitchurch during the 17 years he was with AFA are not likely to travel to that area again. But it could not be said to a legal certainty that no one would return to that area. There was ample support in the record for the proposition that AFA has the prospects for repeat customers. For example, AFA had submitted from one of its brochures two pages excerpting comments from participants in a recent AFA tour (the commenters being identified only by their initials and city of residence); nearly one-third of those quoted indicated that they either had been on other AFA tours or would hope to go on future AFA tours. Indeed, an exhibit submitted by Whitchurch stated that one of the couples in a group he was to escort had been on 11 AFA tours. And two of the 100-200 former AFA South Pacific tour participants contacted by Whitchurch signed up for Whitchurch's proposed tour to the same area. If a tour operator could earn 17% of the price of a tour, and if Whitchurch were eventually successful in soliciting even 30 of the approximately 1,500 AFA participants [89] he has escorted (i.e., 2%, which may reflect the ratio of his success on his first attempt), the profit he could siphon from AFA would total $51,000. Thus, on the present record, the court could not conclude to a legal certainty that the value of AFA's claims did not exceed the jurisdictional minimum.

Further, AFA requested injunctive relief not just against Whitchurch's solicitation of its customers but also against any use of the information. Presumably such an injunction would include a prohibition against Whitchurch's sale or disclosure of the names and addresses of AFA's customers to other tour operators who might be better equipped than Whitchurch to exploit the information and attract more than 2% of the persons whose names Whitchurch could provide them. In addition, AFA's request for punitive damages in the amount of $250,000 might provide a basis for satisfaction of the jurisdictional amount. Whether or not AFA will be able to prove that Whitchurch's conduct was "gross and wanton" and warrants the recovery of such damages under New York law is an open question. But the present record does not foreclose that possibility.

In all the circumstances, we conclude that the record as it existed in the district court did not permit the court to find with legal certainty that the value of AFA's claims did not exceed $50,000.

B. The Merits of AFA's Trade Secrets Claims

Finally, even if Whitchurch were correct in his contention that the court dismissed AFA's trade secrets claims on the merits rather than for lack of jurisdiction, the dismissal could not stand. The question of whether or not a customer list is a trade secret is generally a question of fact. See, e.g., Chevron U.S.A., Inc. v. Roxen Service, Inc., 813 F.2d 26, 29-30 (2d Cir.1987); Defiance Button Machine Co. v. C & C Metal Products Corp., 759 F.2d 1053, 1063 (2d Cir.), cert. denied, 474 U.S. 844, 106 S.Ct. 131, 88 L.Ed.2d 108 (1985); see also 1 Milgrim § 2.03, at 2-47 to 2-49 (1990) ("existence of a trade secret is a question of fact for the determination of the trier of fact, secrecy being a basic element"). The answer depends in part on the subsidiary fact questions of whether or not the owner took "reasonable measures to protect [the] secrecy" of the list, see, e.g., id. § 2.04, at 2-55; Defiance Button Machine Co. v. C & C Metal Products Corp., 759 F.2d at 1063; and the ease or difficulty with which the information could properly be obtained from other sources, see, e.g., id.; Restatement of Torts § 757, comment b (1939). The mere fact that an employee has access to information the employer regards as confidential is not inconsistent with treatment of the information as a trade secret. The employer must, however, take appropriate precautions to alert the employee to the need to maintain the confidentiality. 1 Milgrim § 2.04, at 2-75 to 2-76.

In assessing the record to determine whether there is a genuine issue as to any material fact, the court is required to resolve all ambiguities and draw all factual inferences in favor of the party against whom summary judgment is sought. See, e.g., Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 255, 106 S.Ct. 2505, 2513-14, 91 L.Ed.2d 202 (1986). As set forth in Part I above, in opposition to Whitchurch's motion for summary judgment, AFA submitted several affidavits describing both the efforts it had made to develop a customer list that was not available from any other source, and efforts it had made to prevent dissemination of the names on that list except to persons actually participating in the tours and persons who needed to know those names in order to service the AFA tours. Drawing all permissible fact inferences in favor of AFA, the court could not have concluded that as a matter of law the list was not confidential or was unworthy of protection. The record as it stands on this question is rife with genuine issues of fact.

CONCLUSION

We have considered all of Whitchurch's arguments in support of the district court's judgment and have found them to be without [90] merit. The judgment is vacated, and the matter is remanded for further proceedings not inconsistent with the foregoing.

[1] Honorable Louis L. Stanton, of the United States District Court for the Southern District of New York, sitting by designation.

ORDER AND REASONS

MARTIN L. C. FELDMAN, District Judge

Before the Court is the plaintiff's motion to remand. For the reasons that follow, the plaintiff's motion is GRANTED.

Background

Angela Ochoa was injured while riding in a Ford Explorer that was rear-ended by Paul Gulley, who was driving a Budget rental car. The accident happened on February 2, 2006 in New Orleans. Ms. Ochoa, a Louisiana citizen, sued Paul Gulley, Budget Rent A Car System, Inc., and PV Holding Corporation (the title holder of Budget rental cars) in state court on October 5, 2006. Budget and PV Holding were served on October 31, 2006. They removed the suit to this Court on November 30, 2006, invoking the Court's diversity jurisdiction. Paul Gulley was served on December 21, 2006.

The plaintiff now moves to remand, contending that the Court lacks subject matter jurisdiction. Budget is a citizen of Delaware. Before his apartment was destroyed by Hurricane Katrina on August 29, 2005, Paul Gulley had lived in Orleans Parish, Louisiana. He evacuated to Arlington, Texas, where he has been living since the storm. The issue before the Court is the evacuee Paul Gulley's domicile at the time the state court petition was filed.

I. Standard for Remand

Although the plaintiffs challenge removal in this case, the removing defendants carry the burden of showing the propriety of this Court's removal jurisdiction. See Jernigan v. Ashland Oil, Inc., 989 F.2d 812, 815 (5th Cir.), cert. denied, 510 U.S. 868, 114 S. Ct. 192, 126 L.Ed.2d 150 (1993); Willy v. Coastal Corp., 855 F.2d 1160, 1164 (5th Cir. 1988). In addition, any ambiguities are construed against removal, Butler v. Polk, 592 F.2d 1293, 1296 (5th Cir. 1979), as the removal statute should be strictly construed in favor of remand. York v. Horizon Fed. Sav. and Loan Ass'n, 712 F. Supp. 85, 87 (E.D. La. 1989); see also Shamrock Oil & Gas Corp. v. Sheets, 313 U.S. 100 (1941).

Budget and PV Holding removed this case on the basis of diversity jurisdiction. The plaintiff initially contends that Paul Gulley's failure to consent to the removal renders removal procedurally defective. Alternatively, the plaintiff contends that the Court lacks subject matter jurisdiction because Paul Gulley was a Louisiana domiciliary at the time the complaint and removal petition were filed.

The Court agrees that it lacks subject matter jurisdiction, but first addresses the plaintiff's procedural defect ground for remand.

A defect in the procedure for removal, if timely asserted, may be grounds for remand. 28 U.S.C. § 1447(c) (providing 30-day window for challenges to procedural defects in removal); Caterpillar, Inc. v. Lewis, 519 U.S. 61, (1996). The "rule of unanimity" requires that, absent exceptional circumstances, all served defendants must join or otherwise file a written notice of consent to removal before the expiration of the 30-day removal period in 28 U.S.C. § 1446. Getty Oil Corp. v. Insurance Co. of North America, 841 F.2d 1254, 1261-62 (5th Cir. 1988) (holding that all served defendants are required to join in petition for removal no later than 30 days from the date on which the first defendant was served); Gillis v. Louisiana, 294 F.3d 755, 759 (5th Cir. 2002).^[1] In other words, "[d]efendants (at least those not citizens of the forum state) who are unserved when the removal petition is filed need not join in it." Getty Oil Corp. v. Ins. Co. of North America, 841 F.2d at 1262 n.9 (5th Cir. 1988) (citing Pullman Co. v. Jenkins, 305 U.S. 534 (1939)). It is undisputed that Paul Gulley had not been served at the time the served defendants filed their notice of removal. His consent was not required.^[2]

II. Diversity Jurisdiction

To exercise diversity jurisdiction, complete diversity must exist between the plaintiffs and all of the properly joined defendants, and the amount in controversy must exceed $75,000. See 28 U.S.C. § 1332.

By providing that the judicial power of the United States shall extend to controversies "between Citizens of Different States," Article III, §2 of the United States Constitution vests diversity jurisdiction in federal courts. For a court to have subject matter jurisdiction over a case based upon 28 U.S.C. § 1332, complete diversity must exist between the plaintiff and the defendant (no plaintiff may be a citizen of the same state as any defendant). Strawbridge v. Curtiss, 7 U.S. (3 Cranch) 267, 2 L.Ed. 435 (1806); Freeman v. Northwest Acceptance Corp., 754 F.2d 553, 555 (5th Cir. 1985); Mas v. Perry, 489 F.2d 1396, 1398-00 (5th Cir. 1974), cert. denied, 419 U.S. 842 (1974). The determination of state citizenship for diversity purposes is a matter of federal common law and is not determined by state law. Mas v. Perry, 489 F.2d 1396, 1399 (5th Cir. 1974); Coury v. Prot, 85 F.3d 244, 248 (5th Cir. 1996). Diversity of citizenship must be present at the time the complaint is filed; therefore, the Court examines a litigant's domicile at the time when the complaint was filed. Mas, 489 F.2d at 1399.

The citizenship of an individual is synonymous with his domicile. Freeman, 754 F.2d at 555. The United States Supreme Court defined domicile as "residence at a particular place accompanied with positive or presumptive proof of an intention to remain there for an unlimited time." Mitchell v. United States, 88 U.S. (21 Wall.) 350, 352, 22 L.Ed. 584 (1874). The Fifth Circuit has refined the definition: "A person's domicile is the place of his `true, fixed, and permanent home and principal establishment, and to which he has the intention of returning whenever he is absent therefrom.'" Mas, 489 F.2d at 1399 (quoting Stine v. Moore, 213 F.2d 446, 448 (5th Cir. 1954)). There is no minimum period of residence required, but the intent to remain at the particular place for an unlimited or indefinite period of time is required. Freeman, 754 F.2d at 555.

The law of domicile is a mixed question of law and fact. Welch v. American Surety Co.. 186 F.2d 16 (5th Cir. 1951); Knapp v. State Farm Ins., 584 F. Supp. 905, 906 (E.D. La. 1984); Whitney Nat'l Bank v. Chatelain, 1991 WL 213917 (E.D. La. 1991). The party invoking federal jurisdiction has the burden of pleading the diverse citizenship, and if diversity jurisdiction is properly challenged, that party also bears the burden of proof. Mas, 489 F.2d at 1399 (citations omitted).

In ascertaining domicile, the Court is not limited to the pleadings; rather, it may review record evidence, affidavits, and testimony concerning facts underlying the citizenship of the parties. Coury v. Prot, 85 F.3d 244, 249 (5th Cir. 1996). While relevant to the determination of domicile, a statement of intent is "entitled to little weight if it conflicts with objective facts." Freeman v. Northwest Acceptance Corp., 754 F.2d 553, 556 (5th Cir. 1985).

In determining one's domicile, the Court should address a variety of factors, none of which is itself determinative. Coury v. Prot, 85 F.3d 244, 251 (5th Cir. 1996). In fact, "[t]he court should look to all evidence shedding light on the litigant's intention to establish domicile." Id. The actual fact of residence and the real intention of remaining there, as disclosed by a person's entire course of conduct, are the controlling factors. See Freeman, 754 F.2d at 555-56.^[3]

III.

Federal subject matter jurisdiction here turns on where was Gulley's domicile on October 5, 2006 when this lawsuit was filed. The defendants have the burden of showing that removal was proper; thus, they must show that Gulley was a Texas domiciliary at the time the plaintiff filed her complaint. The Court finds that the defendants have not carried their burden.

The parties agreed to limited discovery relating to the issue of Gulley's domicile and he was deposed in December 2006. Gulley testified that he was born in New Orleans and lived there until Hurricane Katrina. Though he was living in Texas, he was "visiting family" in New Orleans when the car accident giving rise to this lawsuit happened in February 2006.

The plaintiff characterizes Gulley as being domiciled in Louisiana in February 2006 when the accident happened and months later when she filed this lawsuit. She points to Gulley's Louisiana driver's license, which he presented at the time of the accident, and the fact that he did not have permanent employment in Texas until shortly before his deposition.

The defendants characterize Gulley as being domiciled in Texas, pointing to his deposition testimony that he has resided in Texas since he evacuated there shortly after the storm, does not presently intend to live in Louisiana, and recently got a job in Texas.

The Court notes that the record provides few objective facts concerning Gulley's domicile in October 2006; indeed, the record consists only of Gulley's deposition testimony, which is attached (albeit unauthenticated) to the defendants' opposition to the plaintiff's motion to remand. Accordingly, the defendants have not addressed many of the objective facts that typically aid the Court in ascertaining domicile, e.g., the places where the litigant exercises civil and political rights, pays taxes, owns real and personal property, has his drivers and other licenses, maintains bank accounts, belongs to clubs and churches, has places of business or employment, and maintains a home for his family. See Coury v. Prot, 85 F.3d 244, 251 (5th Cir. 1996) (citations omitted).

The record shows that Gulley works and resides in Texas. The defendants do not dispute that Gulley has a Louisiana driver's license. He has lived in New Orleans all of his life until he was forced to leave because Hurricane Katrina destroyed his home; he has family in New Orleans. He was in New Orleans at the time of the accident, though the record does not show the duration of his stay or how often he travels between Texas and Louisiana, nor when he developed his present intention to stay in Texas "indefinitely."

Gulley's recent employment in Texas and his subjective statement that he does not presently intend to live in New Orleans fall short of establishing that his domicile several months ago — at the time the complaint was filed — was Texas and not Louisiana. There is little objective evidence to corroborate his subjective present statement of intent. The defendants bear the burden of showing diverse citizenship because they are attempting to invoke this Court's subject matter jurisdiction. They have failed to carry that burden to show that Gulley was a citizen of Texas at the time the complaint was filed; thus, the Court lacks removal jurisdiction based on diversity of citizenship.

Accordingly, the plaintiff's motion to remand is GRANTED. The case is hereby remanded to the Civil District Court for the Parish of Orleans.

[1] "It is well settled in the Fifth Circuit that all defendants who have been served before removal must consent to removal within 30 days after service on the first defendant." Forman v. Equifax Credit Info. Services, Inc., 1997 WL 162008, at *1 (E.D. La. April 4, 1997) (Clement, J.) (citing Doe v. Kerwood, 969 F.2d 165, 167 (5 Cir. 1992) and Getty Oil Corp. v. Ins. Co.th of North America, 841 F.2d 1254, 1263 (5th Cir. 1988)).

[2] To the extent the plaintiff is contending that removal is defective because Gulley is in fact a Louisiana citizen and not a citizen of Texas, these contentions are addressed in the Court's analysis of whether the parties are completely diverse.

[3] Some factors the Court may examine include the places where the litigant exercises civil and political rights, pays taxes, owns real and personal property, has his drivers and other licenses, maintains bank accounts, belongs to clubs and churches, has places of business or employment, and maintains a home for his family. Coury v. Prot, 85 F.3d 244, 251 (5th Cir. 1996) (citations omitted).

[744] Mr. Hammond, for the appellants.

Mr. Clay, for the respondents.

[816] Mr. Chief Justice MARSHALL delivered the opinion of the Court, and, after stating the case, proceeded as follows:

At the close of the argument, a point was suggested, of such vital importance, as to induce the Court to request that it might be particularly spoken to. That point is, the right of the Bank to sue in the Courts of the United States. It has been argued, and ought to be disposed of, before we proceed to the actual exercise of jurisdiction, by deciding on the rights of the parties.

[817] The appellants contest the jurisdiction of the Court on two grounds:

1st. That the act of Congress has not given it.

2d. That, under the constitution, Congress cannot give it.

1. The first part of the objection depends entirely on the language of the act. The words are, that the Bank shall be "made able and capable in law," "to sue and be sued, plead and be impleaded, answer and be answered, defend and be defended, in all State Courts having competent jurisdiction, and in any Circuit Court of the United States."

These words seem to the Court to admit of but one interpretation. They cannot be made plainer by explanation. They give, expressly, the right "to sue and be sued," "in every Circuit Court of the United States," and it would be difficult to substitute other terms which would be more direct and appropriate for the purpose. The argument of the appellants is founded on the opinion of this Court, in The Bank of the United States v. Deveaux, (5 Cranch, 85.) In that case it was decided, that the former Bank of the United States was not enabled, by the act which incorporated it, to sue in the federal Courts. The words of the 3d section of that act are, that the Bank may "sue and be sued," &c.; "in Courts of record, or any other place whatsoever." The Court was of opinion, that these general words, which are usual in all acts of incorporation, gave only a general capacity to sue, not a particular privilege to sue in the [818] Courts of the United States; and this opinion was strengthened by the circumstance that the 9th rule of the 7th section of the same act, subjects the directors, in case of excess in contracting debt, to be sued in their private capacity, "in any Court of record of the United States, or either of them." The express grant of jurisdiction to the federal Courts, in this case, was considered as having some influence on the construction of the general words of the 3d section, which does not mention those Courts. Whether this decision be right or wrong, it amounts only to a declaration, that a general capacity in the Bank to sue, without mentioning the Courts of the Union, may not give a right to sue in those Courts. To infer from this, that words expressly conferring a right to sue in those Courts, do not give the right, is surely a conclusion which the premises do not warrant.

The act of incorporation, then, confers jurisdiction on the Circuit Courts of the United States, if Congress can confer it.

2. We will now consider the constitutionality of the clause in the act of incorporation, which authorizes the Bank to sue in the federal Courts.

In support of this clause, it is said, that the legislative, executive, and judicial powers, of every well constructed government, are co-extensive with each other; that is, they are potentially co-extensive. The executive department may constitutionally execute every law which the Legislature may constitutionally make, and the judicial department may receive from the Legislature the power of construing every such law. All governments [819] which are not extremely defective in their organization, must possess, within themselves, the means of expounding, as well as enforcing, their own laws. If we examine the constitution of the United States, we find that its framers kept this great political principle in view. The 2d article vests the whole executive power in the President; and the 3d article declares, "that the judicial power shall extend to all cases in law and equity arising under this constitution, the laws of the United States, and treaties made, or which shall be made, under their authority."

This clause enables the judicial department to receive jurisdiction to the full extent of the constitution, laws, and treaties of the United States, when any question respecting them shall assume such a form that the judicial power is capable of acting on it. That power is capable of acting only when the subject is submitted to it by a party who asserts his rights in the form prescribed by law. It then becomes a case, and the constitution declares, that the judicial power shall extend to all cases arising under the constitution, laws, and treaties of the United States.

The suit of The Bank of the United States v. Osborn and others, is a case, and the question is, whether it arises under a law of the United States?

The appellants contend, that it does not, because several questions may arise in it, which depend on the general principles of the law, not on any act of Congress.

If this were sufficient to withdraw a case from [820] the jurisdiction of the federal Courts, almost every case, although involving the construction of a law, would be withdrawn; and a clause in the constitution, relating to a subject of vital importance to the government, and expressed in the most comprehensive terms, would be construed to mean almost nothing. There is scarcely any case, every part of which depends on the constitution, laws, or treaties of the United States. The questions, whether the fact alleged as the foundation of the action, be real or fictitious; whether the conduct of the plaintiff has been such as to entitle him to maintain his action; whether his right is barred; whether he has received satisfaction, or has in any manner released his claims, are questions, some or all of which may occur in almost every case; and if their existence be sufficient to arrest the jurisdiction of the Court, words which seem intended to be as extensive as the constitution, laws, and treaties of the Union, which seem designed to give the Courts of the government the construction of all its acts, so far as they affect the rights of individuals, would be reduced to almost nothing.

In those cases in which original jurisdiction is given to the Supreme Court, the judicial power of the United States cannot be exercised in its appellate form. In every other case, the power is to be exercised in its original or appellate form, or both, as the wisdom of Congress may direct. With the exception of these cases, in which original jurisdiction is given to this Court, there is none to which the judicial power extends, from which the original jurisdiction of the inferior Courts is excluded [821] by the constitution. Original jurisdiction, so far as the constitution gives a rule, is co-extensive with the judicial power. We find, in the constitution, no prohibition to its exercise, in every case in which the judicial power can be exercised. It would be a very bold construction to say, that this power could be applied in its appellate form only, to the most important class of cases to which it is applicable.

The constitution establishes the Supreme Court, and defines its jurisdiction. It enumerates cases in which its jurisdiction is original and exclusive; and then defines that which is appellate, but does not insinuate, that in any such case, the power cannot be exercised in its original form by Courts of original jurisdiction. It is not insinuated, that the judicial power, in cases depending on the character of the cause, cannot be exercised in the first instance, in the Courts of the Union, but must first be exercised in the tribunals of the State; tribunals over which the government of the Union has no adequate control, and which may be closed to any claim asserted under a law of the United States.

We perceive, then, no ground on which the proposition can be maintained, that Congress is incapable of giving the Circuit Courts original jurisdiction, in any case to which the appellate jurisdiction extends.

We ask, then, if it can be sufficient to exclude this jurisdiction, that the case involves questions depending on general principles? A cause may depend on several questions of fact and law. Some [822] of these may depend on the construction of a law of the United States; others on principles unconnected with that law. If it be a sufficient foundation for jurisdiction, that the title or right set up by the party, may be defeated by one construction of the constitution or law of the United States, and sustained by the opposite construction, provided the facts necessary to support the action be made out, then all the other questions must be decided as incidental to this, which gives that jurisdiction. Those other questions cannot arrest the proceedings. Under this construction, the judicial power of the Union extends effectively and beneficially to that most important class of cases, which depend on the character of the cause. On the opposite construction, the judicial power never can be extended to a whole case, as expressed by the constitution, but to those parts of cases only which present the particular question involving the construction of the constitution or the law. We say it never can be extended to the whole case, because, if the circumstance that other points are involved in it, shall disable Congress from authorizing the Courts of the Union to take jurisdiction of the original cause, it equally disables Congress from authorizing those Courts to take jurisdiction of the whole cause, on an appeal, and thus will be restricted to a single question in that cause; and words obviously intended to secure to those who claim rights under the constitution, laws, or treaties of the United States, a trial in the federal Courts, will be restricted to the insecure remedy of an appeal upon an insulated point, after it has [823] received that shape which may be given to it by another tribunal, into which he is forced against his will.

We think, then, that when a question to which the judicial power of the Union is extended by the constitution, forms an ingredient of the original cause, it is in the power of Congress to give the Circuit Courts jurisdiction of that cause, although other questions of fact or of law may be involved in it.

The case of the Bank is, we think, a very strong case of this description. The charter of incorporation not only creates it, but gives it every faculty which it possesses. The power to acquire rights of any description, to transact business of any description, to make contracts of any description, to sue on those contracts, is given and measured by its charter, and that charter is a law of the United States. This being can acquire no right, make no contract, bring no suit, which is not authorized by a law of the United States. It is not only itself the mere creature of a law, but all its actions and all its rights are dependant on the same law. Can a being, thus constituted, have a case which does not arise literally, as well as substantially, under the law?

Take the case of a contract, which is put as the strongest against the Bank.

When a Bank sues, the first question which presents itself, and which lies at the foundation of the cause, is, has this legal entity a right to sue? Has it a right to come, not into this Court particularly, but into any Court? This depends on a [824] law of the United States. The next question is, has this being a right to make this particular contract? If this question be decided in the negative, the cause is determined against the plaintiff; and this question, too, depends entirely on a law of the United States. These are important questions, and they exist in every possible case. The right to sue, if decided once, is decided for ever; but the power of Congress was exercised antecedently to the first decision on that right, and if it was constitutional then, it cannot cease to be so, because the particular question is decided. It may be revived at the will of the party, and most probably would be renewed, were the tribunal to be changed. But the question respecting the right to make a particular contract, or to acquire a particular property, or to sue on account of a particular injury, belongs to every particular case, and may be renewed in every case. The question forms an original ingredient in every cause. Whether it be in fact relied on or not, in the defence, it is still a part of the cause, and may be relied on. The right of the plaintiff to sue, cannot depend on the defence which the defendant may choose to set up. His right to sue is anterior to that defence, and must depend on the state of things when the action is brought. The questions which the case involves, then, must determine its character, whether those questions be made in the cause or not.

The appellants say, that the case arises on the contract; but the validity of the contract depends on a law of the United States, and the plaintiff is [825] compelled, in every case, to show its validity. The case arises emphatically under the law. The act of Congress is its foundation. The contract could never have been made, but under the authority of that act. The act itself is the first ingredient in the case, is its origin, is that from which every other part arises. That other questions may also arise, as the execution of the contract, or its performance, cannot change the case, or give it any other origin than the charter of incorporation. The action still originates in, and is sustained by, that charter.

The clause giving the Bank a right to sue in the Circuit Courts of the United States, stands on the same principle with the acts authorizing officers of the United States who sue in their own names, to sue in the Courts of the United States. The Postmaster General, for example, cannot sue under that part of the constitution which gives jurisdiction to the federal Courts, in consequence of the character of the party, nor is he authorized to sue by the Judiciary Act. He comes into the Courts of the Union under the authority of an act of Congress, the constitutionality of which can only be sustained by the admission that his suit is a case arising under a law of the United States. If it be said, that it is such a case, because a law of the United States authorizes the contract, and authorizes the suit, the same reasons exist with respect to a suit brought by the Bank. That, too, is such a case; because that suit, too, is itself authorized, and is brought on a contract authorized by a law of the United States. It depends absolutely [826] on that law, and cannot exist a moment without its authority.

If it be said, that a suit brought by the Bank may depend in fact altogether on questions unconnected with any law of the United States, it is equally true, with respect to suits brought by the Postmaster General. The plea in bar may be payment, if the suit be brought on a bond, or non-assumpsit, if it be brought on an open account, and no other question may arise than what respects the complete discharge of the demand. Yet the constitutionality of the act authorizing the Postmaster General to sue in the Courts of the United States, has never been drawn into question. It is sustained singly by an act of Congress, standing on that construction of the constitution which asserts the right of the Legislature to give original jurisdiction to the Circuit Courts, in cases arising under a law of the United States.

The clause in the patent law, authorizing suits in the Circuit Courts, stands, we think, on the same principle. Such a suit is a case arising under a law of the United States. Yet the defendant may not, at the trial, question the validity of the patent, or make any point which requires the construction of an act of Congress. He may rest his defence exclusively on the fact, that he has not violated the right of the plaintiff. That this fact becomes the sole question made in the cause, cannot oust the jurisdiction of the Court, or establish the position, that the case does not arise under a law of the United States.

It is said, that a clear distinction exists between [827] the party and the cause; that the party may originate under a law with which the cause has no connexion; and that Congress may, with the same propriety, give a naturalized citizen, who is the mere creature of a law, a right to sue in the Courts of the United States, as give that right to the Bank.

This distinction is not denied; and, if the act of Congress was a simple act of incorporation, and contained nothing more, it might be entitled to great consideration. But the act does not stop with incorporating the Bank. It proceeds to bestow upon the being it has made, all the faculties and capacities which that being possesses. Every act of the Bank grows out of this law, and is tested by it. To use the language of the constitution, every act of the Bank arises out of this law.

A naturalized citizen is indeed made a citizen under an act of Congress, but the act does not proceed to give, to regulate, or to prescribe his capacities. He becomes a member of the society, possessing all the rights of a native citizen, and standing, in the view of the constitution, on the footing of a native. The constitution does not authorize Congress to enlarge or abridge those rights. The simple power of the national Legislature, is to prescribe a uniform rule of naturalization, and the exercise of this power exhausts it, so far as respects the individual. The constitution then takes him up, and, among other rights, extends to him the capacity of suing in the Courts of the United States, precisely under the same circumstances under which a native might sue. He is [828] distinguishable in nothing from a native citizen, except so far as the constitution makes the distinction. The law makes none.

There is, then, no resemblance between the act incorporating the Bank, and the general naturalization law.

Upon the best consideration we have been able to bestow on this subject, we are of opinion, that the clause in the act of incorporation, enabling the Bank to sue in the Courts of the United States, is consistent with the constitution, and to be obeyed in all Courts.

We will now proceed to consider the merits of the cause.

The appellants contend, that the decree of the Circuit Court is erroneous —

1. Because no authority is shown in the record, from the Bank, authorizing the institution or prosecution of the suit.

2. Because, as against the defendant, Sullivan, there are neither proofs nor admissions, sufficient to sustain the decree.

3. Because, upon equitable principles, the case made in the bill, does not warrant a decree against either Osborn or Harper, for the amount of coin and notes in the bill specified to have passed through their hands.

4. Because, the defendants are decreed to pay interest upon the coin, when it was not in the power of Osborn or Harper, and was stayed in the hands of Sullivan by injunction.

5. Because, the case made in the bill does not [829] warrant the interference of a Court of Chancery, by injunction.

6. Because, if any case is made in the bill proper for the interference of a Court of Chancery, it is against the State of Ohio, in which case the Circuit Court could not exercise jurisdiction.

7. Because, the decree assumes that the Bank of the United States is not subject to the taxing power of the State of Ohio, and decides that the law of Ohio, the execution of which is enjoined, is unconstitutional.

These points will be considered in the order in which they are made.

1. It is admitted that a corporation can only appear by attorney, and it is also admitted, that the attorney must receive the authority of the corporation to enable him to represent it. It is not admitted that this authority must be under seal. On the contrary, the principle decided in the cases of the Bank of Columbia v. Patterson, &c.; is supposed to apply to this case, and to show that the seal may be dispensed with. It is, however, unnecessary to pursue this inquiry, since the real question is, whether the non-appearance of the power in the record be error, not whether the power was insufficient in itself.

Natural persons may appear in Court, either by themselves, or by their attorney. But no man has a right to appear as the attorney of another, without the authority of that other. In ordinary cases, the authority must be produced, because there is, in the nature of things, no prima facie evidence that one man is in fact the attorney of another. [830] The case of an attorney at law, an attorney for the purpose of representing another in Court, and prosecuting or defending a suit in his name, is somewhat different. The power must indeed exist, but its production has not been considered as indispensable. Certain gentlemen, first licensed by government, are admitted by order of Court, to stand at the bar, with a general capacity to represent all the suitors in the Court. The appearance of any one of these gentlemen in a cause, has always been received as evidence of his authority; and no additional evidence, so far as we are informed, has ever been required. This practice, we believe, has existed from the first establishment of our Courts, and no departure from it has been made in those of any State, or of the Union.

The argument supposes some distinction, in this particular, between a natural person and a corporation; but the Court can perceive no reason for this distinction. A corporation, it is true, can appear only by attorney, while a natural person may appear for himself. But when he waives this privilege, and elects to appear by attorney, no reason is perceived why the same evidence should not be required, that the individual professing to represent him has authority to do so, which would be required if he were incapable of appearing in person. The universal and familiar practice, then, of permitting gentlemen of the profession to appear without producing a warrant of attorney, forms a rule, which is as applicable in reason to their appearance for a corporation, as for a natural person. Were it even otherwise, the practice is [831] as uniform and as ancient, with regard to corporations, as to natural persons. No case has ever occurred, so far as we are informed, in which the production of a warrant of attorney has been supposed a necessary preliminary to the appearance of a corporation, either as plaintiff or defendant, by a gentleman admitted to the bar of the Court. The usage, then, is as full authority for the case of a corporation, as of an individual. If this usage ought to be altered, it should be a rule to operate prospectively, not by the reversal of a decree pronounced in conformity with the general course of the Court, in a case in which no doubt of the legality of the appearance had ever been suggested.

In the statutes of jeofails and amendment, which respect this subject, the non-appearance of a warrant of attorney in the record, has generally been treated as matter of form; and the 32d section of the Judiciary Act may very well be construed to comprehend this formal defect in its general terms, in a case of law. No reason is perceived why the Courts of Chancery should be more rigid in exacting the exhibition of a warrant of attorney than a Court of laws and, since the practice has, in fact, been the same in both Courts, an appellate Court ought, we think, to be governed in both by the same rule.

2. The second point is one on which the productiveness of any decree in favour of the plaintiffs most probably depends; for, if the claim be not satisfied with the money found in the possession of Sullivan, it is, at best, uncertain whether [832] a fund, out of which it can be satisfied, is to be found elsewhere.

In inquiring whether the proofs or admissions in the cause be sufficient to charge Sullivan, the Court will look into the answer of Currie, as well as into that of Sullivan. In objection to this course, it is said, that the answer of one defendant cannot be read against another. This is generally, but not universally, true. Where one defendant succeeds to another, so that the right of the one devolves on the other, and they become privies in estate, the rule is not admitted to apply. Thus, if an ancestor die, pending a suit, and the proceedings be revived against his heir, or if a suit be revived against an executor or administrator, the answer of the deceased person, or any other evidence, establishing any fact against him, might be read also against the person who succeeds to him. So, a pendente lite purchaser is bound by the decree, without being even made a party to the suit; a fortiori, he would, if made a party, be bound by the testimony taken against the vendor.

In this case, if Currie received the money taken out of the Bank, and passed it over to Sullivan, the establishment of this fact, in a suit against Currie, would seem to bind his successor, Sullivan, both as a privy in estate, and as a person getting possession pendente lite, if the original suit had been instituted against Currie. We can perceive no difference, so far as respects the answer of Currie, between the case supposed, and the case as it stands. If Currie, who was the predecessor of Sullivan, admits that he received the money of [833] the Bank, the fact seems to bind all those coming in under him, as completely as it binds himself. This, therefore, appears to the Court to be a case in which, upon principle, the answer of Currie may be read.

His answer states, that on or about the 19th or 20th of September, 1819, the defendant, Harper, delivered to him, in coin and notes, the sum of 98,000 dollars, which he was informed, and believed to be the money levied on the Bank as a tax, in pursuance of the law of the State of Ohio. After consulting counsel on the question, whether he ought to retain this sum within his individual control, or pass it to the credit of the State on the books of the treasury, he adopted the latter course, but retained it carefully in a trunk, separate from the other funds of the treasury. The money afterwards came to the hands of Sullivan, the gentleman who succeeded him as treasurer, and gave him a receipt for all the money in the treasury, including this, which was still kept separate from the rest.

We think no reasonable doubt can be entertained, but that the 98,000 dollars, delivered by Harper to Currie, were taken out of the Bank. Currie understood and believed it to be the fact. When did he so understand and believe it? At the time when he received the money. And from whom did he derive his understanding and belief? The inference is irresistible, that he derived it from his own knowledge of circumstances, for they were of public notoriety, and from the information of Harper. In the necessary course of things, Harper, who was sent, as Currie must have known, on this [834] business, brings with him to the treasurer of the State, a sum of money, which, by the law, was to be taken out of the Bank, pays him 98,000 dollars thereof, which the treasurer receives and keeps, as being money taken from the Bank, and so enters it on the books of the treasury. In a suit brought against Mr. Currie for this money, by the State of Ohio, if he had failed to account for it, could any person doubt the competency of the testimony to charge him? We think no mind could hesitate in such a case.

Currie, then, being clearly in possession of this money, and clearly liable for it, we are next to look into Sullivan's answer, for the purpose of inquiring whether he admits any facts which show him to be liable also.

Sullivan denies all personal knowledge of the transaction; that is, he was not in office when it took place, and was not present when the money was taken out of the Bank, or when it was delivered to Currie. But when he entered the treasury office, he received this sum of 98,000 dollars, separate from the other money of the treasury, which, he understood from report, and was informed by his predecessor, from whom he received it, was the money taken out of the Bank. This sum has remained untouched ever since, from respect to the injunction awarded by the Court.

We ask, if a rational doubt can remain on this subject.

Mr. Currie, as treasurer of the State of Ohio, receives 98,000 dollars, as being the amount of a tax imposed by the Legislature of that State on [835] the Bank of the United States; enters the same on the books of the treasury; and, the legality of the act by which the money was levied being questioned, puts it in a trunk, and keeps it apart from the other money belonging to the public. He resigns his office, and is succeeded by Mr. Sullivan, to whom he delivers the money, informing him, at the same time, that it is the money raised from the Bank; and Mr. Sullivan continues to keep it apart, and abstains from the use of it, out of respect to an injunction, forbidding him to pay it away, or in any manner to dispose of it. Is it possible to doubt the identity of this money?

Even admitting that the answer of Currie, though establishing his liability as to himself, could not prove even that fact as to Sullivan; the answer of Sullivan is itself sufficient, we think, to charge him. He admits that these 98,000 dollars were delivered to him, as being the money which was taken out of the Bank, and that he so received it; for, he says, he understood this sum was the same as charged in the bill; that his information was from report, and from his predecessor; and that the money has remained untouched, from respect to the injunction. This declaration, then, is a part of the fact. The fact, as admitted in his answer, is not simply that he received 98,000 dollars, but that he received 98,000 dollars, as being the money taken out of the Bank — the money to which the writ of injunction applied.

In a common action between two private individuals, such an admission would, at least, be sufficient to throw on the defendant the burthen of [836] proving that the money, which he acknowledges himself to have received and kept as the money of the plaintiff, was not that which it was declared to be on its delivery. A declaration, accompanying the delivery, and constituting a part of it, gives a character to the transaction, and is not to be placed on the same footing with a declaration made by the same person at a different time. The answer of Sullivan, then, is, in the opinion of the Court, sufficient to show that these 98,000 dollars were the specific dollars for which this suit was brought. This sum having come to his possession with full knowledge of the fact, in a separate trunk, unmixed with money, and with notice that an injunction had been awarded respecting it, he would seem to be responsible to the plaintiff for it, unless he can show sufficient matter to discharge himself.

3. The next objection is, to the decree against Osborn and Harper, as to whom the bill was taken for confessed.

The bill charges, that Osborn employed John L. Harper to collect the tax, who proceeded by violence to enter the office of discount and deposit at Chilicothe, and forcibly took therefrom 100,000 dollars in specie and bank notes; and that, at the time of the seizure, Harper well knew, and was duly notified, that an injunction had been allowed, which money was delivered either to Currie or Osborn.

So far as respects Harper and Osborn, these allegations are to be considered as true. If the act of the Legislature of Ohio, and the official [837] character of Osborn, constitute a defence, neither of these defendants are liable, and the whole decree is erroneous; but if the act be unconstitutional and void, it can be no justification, and both these defendants are to be considered as individuals who are amenable to the laws. Considering them, for the present, in this character, the fact, as made out in the bill, is, that Osborn employed Harper to do an illegal act, and that Harper has done that act; and that they are jointly responsible for it, is supposed to be as well settled as any principle of law whatever.

We think it unnecessary, in this part of the case, to enter into the inquiry respecting the effect of the injunction. No injunction is necessary to attach responsibility on those who conspire to do an illegal act, which this is, if not justified by the authority under which it was done.

4. The next objection is, to the allowance of interest on the coin, which constituted a part of the sum decreed to the complainants. Had the complainants, without the intervention of a Court of equity, resorted to their legal remedy for the injury sustained, their right to principal and interest would have stood on equal ground. The same rule would be adopted in a Court of equity, had the subject been left under the control of the party in possession, while the right was in litigation. But the subject was not left under the control of the party. The Court itself interposed, and forbade the person, in whose possession the property was, to make any use of it. This order having been obeyed, places the defendant in the same [838] situation, so far as respects interest, as if the Court had taken the money into its own custody. The defendant, in obeying the mandate of the Court, becomes its instrument, as entirely as the Clerk of the Court would have been, had the money been placed in his hands. It does not appear reasonable, that a decree which proceeds upon the idea, that the injunction of the Court was valid, ought to direct interest to be paid on the money which that injunction restrained the defendant from using.

5. The fifth objection to the decree is, that the case made in the bill does not warrant the interference of a Court of Chancery.

In examining this question, it is proper that the Court should consider the real case, and its actual circumstances. The original bill prays for an injunction against Ralph Osborn, Auditor of the State of Ohio, to restrain him from executing a law of that State, to the great oppression and injury of the complainants, and to the destruction of rights and privileges conferred on them by their charter, and by the constitution of the United States. The true inquiry is, whether an injunction can be issued to restrain a person, who is a State officer, from performing any official act enjoined by statute; and whether a Court of equity can decree restitution, if the act be performed. In pursuing this inquiry, it must be assumed, for the present, that the act is unconstitutional, and furnishes no authority or protection to the officer who is about to proceed under it. This must be assumed, because, in the arrangement of his argument, [839] the counsel who opened the cause, has chosen to reserve that point for the last, and to contend that, though the law be void, no case is made out against the defendants. We suspend, also, the consideration of the question, whether the interest of the State of Ohio, as disclosed in the bill, shows a want of jurisdiction in the Circuit Court, which ought to have arrested its proceedings. That question, too, is reserved by the appellants, and will be subsequently considered. The sole inquiry, for the present, is, whether, stripping the case of these objections, the plaintiffs below were entitled to relief in a Court of equity, against the defendants, and to the protection of an injunction. The appellants expressly waive the extravagant proposition, that a void act can afford protection to the person who executes it, and admits the liability of the defendants to the plaintiffs, to the extent of the injury sustained, in an action at law. The question, then, is reduced to the single inquiry, whether the case is cognizable in a Court of equity. If it is, the decree must be affirmed, so far as it is supported by the evidence in the cause.

The appellants allege, that the original bill contains no allegation which can justify the application for an injunction, and treat the declarations of Ralph Osborn, the Auditor, that he should execute the law, as the light and frivolous threats of an individual, that he would commit an ordinary trespass. But surely this is not the point of view in which the application for an injunction is to be considered. The Legislature of Ohio had passed [840] a law for the avowed purpose of expelling the Bank from the State; and had made it the duty of the Auditor to execute it as a ministerial officer. He had declared that he would perform this duty. The law, if executed, would unquestionably effect its object, and would deprive the Bank of its chartered privileges, so far as they were to be exercised in that State. It must expel the Bank from the State; and this is, we think, a conclusion which the Court might rightfully draw from the law itself. That the declarations of the Auditor would be fulfilled, did not admit of reasonable doubt. It was to be expected, that a person continuing to hold an office, would perform a duty enjoined by his government, which was completely within his power. This duty was to be repeated until the Bank should abandon the exercise of its chartered rights.

To treat this as a common casual trespass, would be to disregard entirely its true character and substantial merits. The application to the Court was, to interpose its writ of injunction, to protect the Bank, not from the casual trespass of an individual, who might not perform the act he threatened, but from the total destruction of its franchise, of its chartered privileges, so far as respected the State of Ohio. It was morally certain, that the Auditor would proceed to execute the law, and it was morally certain, that the effect must be the expulsion of the Bank from the State. An annual charge of 100,000 dollars, would more than absorb all the advantages of the privilege, and would consequently annul it.

[841] The appellants admit, that injunctions are often awarded for the protection of parties in the enjoyment of a franchise; but deny that one has ever been granted in such a case as this. But, although the precise case may never have occurred, if the same principle applies, the same remedy ought to be afforded. The interference of the Court in this class of cases, has most frequently been to restrain a person from violating an exclusive privilege, by participating in it. But if, instead of a continued participation in the privilege, the attempt be to disable the party from using it, is not the reason for the interference of the Court rather strengthened than weakened? Had the privilege of the Bank been exclusive, the argument admits that any other person, or company, might have been enjoined, according to the regular course of the Court of Chancery, from using or exercising the same business. Why would such person or company have been enjoined? To prevent a permanent injury from being done to the party entitled to the franchise or privilege; which injury, the appellants say, cannot be estimated in damages. It requires no argument to prove, that the injury is greater, if the whole privilege be destroyed, than if it be divided; and, so far as respects the estimate of damages, although precise accuracy may not be attained, yet a reasonable calculation may be made of the amount of the injury, so as to satisfy the Court and Jury. It will not be pretended, that, in such a case, an action at law could not be maintained, or that the materials do not exist on which a verdict might be [842] found, and a judgment rendered. But in this, and many other cases of continuing injuries, as in the case of repeated ejectments, a Court of Chancery will interpose. The injury done, by denying to the Bank the exercise of its franchise in the State of Ohio, is as difficult to calculate, as the injury done by participating in an exclusive privilege. The single act of levying the tax in the first instance, is the cause of an action at law; but that affords a remedy only for the single act, and is not equal to the remedy in Chancery, which prevents its repetition, and protects the privilege. The same conservative principle, which induces the Court to interpose its authority for the protection of exclusive privileges, to prevent the commission of waste, even in some cases of trespass, and in many cases of destruction, will, we think, apply to this. Indeed, trespass is destruction, where there is no privity of estate.

If the State of Ohio could have been made a party defendant, it can scarcely be denied, that this would be a strong case for an injunction. The objection is, that, as the real party cannot be brought before the Court, a suit cannot be sustained against the agents of that party; and cases have been cited, to show that a Court of Chancery will not make a decree, unless all those who are substantially interested, be made parties to the suit.

This is certainly true, where it is in the power of the plaintiff to make them parties; but if the person who is the real principal, the person who is the true source of the mischief, by whose power and for whose advantage it is done, be himself [843] above the law, be exempt from all judicial process, it would be subversive of the best established principles, to say that the laws could not afford the same remedies against the agent employed in doing the wrong, which they would afford against him, could his principal be joined in the suit. It is admitted, that the privilege of the principal is not communicated to the agent; for the appellants acknowledge that an action at law would lie against the agent, in which full compensation ought to be made for the injury. It being admitted, then, that the agent is not privileged by his connexion with his principal, that he is responsible for his own act, to the full extent of the injury, why should not the preventive power of the Court also be applied to him? Why may it not restrain him from the commission of a wrong, which it would punish him for committing? We put out of view the character of the principal as a sovereign State, because that is made a distinct point, and consider the question singly as respects the want of parties. Now, if the party before the Court would be responsible for the whole injury, why may he not be restrained from its commission, if no other party can be brought before the Court? The appellants found their distinction on the legal principle, that all trespasses are several as well as joint, without inquiry into the validity of this reason, if true. We ask, if it be true? Will it be said, that the action of trespass is the only remedy given for this injury? Can it be denied, that an action on the case, for money had and received to the plaintiff's use, might be maintained? [844] We think it cannot; and if such an action might be maintained, no plausible reason suggests itself to us, for the opinion, that an injunction may not be awarded to restrain the agent, with as much propriety as it might be awarded to restrain the principal, could the principal be made a party.

We think the reason for an injunction is much stronger in the actual, than it would be in the supposed case. In the regular course of things, the agent would pay over the money immediately to his principal, and would thus place it beyond the reach of the injured party, since his principal is not amenable to the law. The remedy for the injury, would be against the agent only; and what agent could make compensation for such an injury? The remedy would have nothing real in it. It would be a remedy in name only, not in substance. This alone would, in our opinion, be a sufficient reason for a Court of equity. The injury would, in fact, be irreparable; and the cases are innumerable, in which injunctions are awarded on this ground.

But, were it even to be admitted, that the injunction, in the first instance, was improperly awarded, and that the original bill could not be maintained, that would not, we think, materially affect the case. An amended and supplemental bill, making new parties, has been filed in the cause, and on that bill, with the proceedings under it, the decree was pronounced. The question is, whether that bill and those proceedings support the decree.

The case they make, is, that the money and [845] notes of the plaintiffs, in the Circuit Court, have been taken from them without authority, and are in possession of one of the defendants, who keeps them separate and apart from all other money and notes. It is admitted, that this defendant would be liable for the whole amount in an action at law; but it is denied that he is liable in a Court of equity.

We think it a case in which a Court of equity ought to interpose, and that there are several grounds on which its jurisdiction may be placed.

One, which appears to be ample for the purpose, is, that a Court will always interpose, to prevent the transfer of a specific article, which, if transferred, will be lost to the owner. Thus, the holder of negotiable securities, indorsed in the usual manner, if he has acquired them fraudulently, will be enjoined from negotiating them; because if negotiated, the maker or indorser must pay them.^[1] Thus, too, a transfer of stock will be restrained in favour of a person having the real property in the article. In these cases, the injured party would have his remedy at law; and the probability that this remedy would be adequate, is stronger in the cases put in the books, than in this, where the sum is so greatly beyond the capacity of an ordinary agent to pay. But it is the province of a Court of equity, in such cases, to arrest the injury, and prevent the wrong. The remedy is more beneficial and complete, than the law can give. The money of the Bank, if mingled with the other money [846] in the treasury, and put into circulation, would be totally lost to the owners; and the reason for an injunction is, at least, as strong in such a case, as in the case of a negotiable note.

6. We proceed now to the 6th point made by the appellants, which is, that if any case is made in the bill, proper for the interference of a Court of Chancery, it is against the State of Ohio, in which case the Circuit Court could not exercise jurisdiction.

The bill is brought, it is said, for the purpose of protecting the Bank in the exercise of a franchise granted by a law of the United States, which franchise the State of Ohio asserts a right to invade, and is about to invade. It prays the aid of the Court to restrain the officers of the State from executing the law. It is, then, a controversy between the Bank and the State of Ohio. The interest of the State is direct and immediate, not consequential. The process of the Court, though not directed against the State by name, acts directly upon it, by restraining its officers. The process, therefore, is substantially, though not in form, against the State, and the Court ought not to proceed without making the State a party. If this cannot be done, the Court cannot take jurisdiction of the cause.

The full pressure of this argument is felt, and the difficulties it presents are acknowledged. The direct interest of the State in the suit, as brought, is admitted; and, had it been in the power of the Bank to make it a party, perhaps no decree ought to have been pronounced in the cause, until the [847] State was before the Court. But this was not in the power of the Bank. The eleventh amendment of the constitution has exempted a State from the suits of citizens of other States, or aliens; and the very difficult question is to be decided, whether, in such a case, the Court may act upon the agents employed by the State, and on the property in their hands.

Before we try this question by the constitution, it may not be time misapplied, if we pause for a moment, and reflect on the relative situation of the Union with its members, should the objection prevail.

A denial of jurisdiction forbids all inquiry into the nature of the case. It applies to cases perfectly clear in themselves; to cases where the government is in the exercise of its best established and most essential powers, as well as to those which may be deemed questionable. It asserts, that the agents of a State, alleging the authority of a law void in itself, because repugnant to the constitution, may arrest the execution of any law in the United States. It maintains, that if a State shall impose a fine or penalty on any person employed in the execution of any law of the United States, it may levy that fine or penalty by a ministerial officer, without the sanction even of its own Courts; and that the individual, though he perceives the approaching danger, can obtain no protection from the judicial department of the government. The carrier of the mail, the collector of the revenue, the marshal of a district, the recruiting officer, may all be inhibited, under ruinous [848] penalties, from the performance of their respective duties; the warrant of a ministerial officer may authorize the collection of these penalties, and the person thus obstructed in the performance of his duty, may indeed resort to his action for damages, after the infliction of the injury, but cannot avail himself of the preventive justice of the nation to protect him in the performance of his duties. Each member of the Union is capable, at its will, of attacking the nation, of arresting its progress at every step, of acting vigorously and effectually in the execution of its designs, while the nation stands naked, stripped of its defensive armour, and incapable of shielding its agent or executing its laws, otherwise than by proceedings which are to take place after the mischief is perpetrated, and which must often be ineffectual, from the inability of the agents to make compensation.

These are said to be extreme cases; but the case at bar, had it been put by way of illustration in argument, might have been termed an extreme case; and, if a penalty on a revenue officer, for performing his duty, be more obviously wrong than a penalty on the Bank, it is a difference in degree, not in principle. Public sentiment would be more shocked by the infliction of a penalty on a public officer for the performance of his duty, than by the infliction of this penalty on a Bank which, while carrying on the fiscal operations of the government, is also transacting its own business; but, in both cases, the officer levying the penalty acts under a void authority, and the power [849] to restrain him is denied as positively in the one as in the other.

The distinction between any extreme case, and that which has actually occurred, if, indeed, any difference of principle can be supposed to exist between them, disappears, when considering the question of jurisdiction; for, if the Courts of the United States cannot rightfully protect the agents who execute every law authorized by the constitution, from the direct action of State agents in the collection of penalties, they cannot rightfully protect those who execute any law.

The question, then, is, whether the constitution of the United States has provided a tribunal which can peacefully and rightfully protect those who are employed in carrying into execution the laws of the Union, from the attempts of a particular State to resist the execution of those laws.

The State of Ohio denies the existence of this power, and contends, that no preventive proceedings whatever, or proceedings against the very property which may have been seized by the agent; of a State, can be sustained against such agent, because they would be substantially against the State itself, in violation of the 11th amendment of the constitution.

That the Courts of the Union cannot entertain a suit brought against a State by an alien, or the citizen of another State, is not to be controverted. Is a suit, brought against an individual, for any cause whatever, a suit against a State, in the sense of the constitution?

[850] The 11th amendment is the limitation of a power supposed to be granted in the original instrument; and to understand accurately the extent of the limitation, it seems proper to define the power that is limited.

The words of the constitution, so far as they respect this question, are, "The judicial power shall extend to controversies between two or more States, between a State and citizens of another State, and between a State and foreign states, citizens, or subjects."

A subsequent clause distributes the power previously granted, and assigns to the Supreme Court original jurisdiction in those cases in which "a State shall be a party."

The words of the 11th amendment are, "The judicial power of the United States shall not be construed to extend to any suit in law or equity, commenced or prosecuted against one of the United States, by citizens of another State, or by citizens or subjects of a foreign state."

The Bank of the United States contends, that in all cases in which jurisdiction depends on the character of the party, reference is made to the party on the record, not to one who may be interested, but is not shown by the record to be a party.

The appellants admit, that the jurisdiction of the Court is not ousted by any incidental or consequential interest, which a State may have in the decision to be made, but is to be considered as a party where the decision acts directly and immediately upon the State, through its officers.

[851] If this question were to be determined on the authority of English decisions, it is believed that no case can be adduced, where any person has been considered as a party, who is not made so in the record. But the Court will not review those decisions, because it is thought a question growing out of the constitution of the United States, requires rather an attentive consideration of the words of that instrument, than of the decisions of analogous questions by the Courts of any other country.

Do the provisions, then, of the American constitution, respecting controversies to which a State may be a party, extend, on a fair construction of that instrument, to cases in which the State is not a party on the record?

The first in the enumeration, is a controversy between two or more States.

There are not many questions in which a State would be supposed to take a deeper or more immediate interest, than in those which decide on the extent of her territory. Yet the constitution, not considering the State as a party to such controversies, if not plaintiff or defendant on the record, has expressly given jurisdiction in those between citizens claiming lands under grants of different States. If each State, in consequence of the influence of a decision on her boundary, had been considered, by the framers of the constitution, as a party to that controversy, the express grant of jurisdiction would have been useless. The grant of it certainly proves, that the constitution [852] does not consider the State as a party in such a case.

Jurisdiction is expressly granted, in those cases only where citizens of the same State claim lands under grants of different States. If the claimants be citizens of different States, the Court takes jurisdiction for that reason. Still, the right of the State to grant, is the essential point in dispute; and in that point the State is deeply interested. If that interest converts the State into a party, there is an end of the cause; and the constitution will be construed to forbid the Circuit Courts to take cognizance of questions to which it was thought necessary expressly to extend their jurisdiction, even when the controversy arose between citizens of the same State.

We are aware, that the application of these cases may be denied, because the title of the State comes on incidentally, and the appellants admit the jurisdiction of the Court, where its judgment does not act directly upon the property or interests of the State; but we deemed it of some importance to show, that the framers of the constitution contemplated the distinction between cases in which a State was interested, and those in which it was a party, and made no provision for a case of interest, without being a party on the record.

In cases where a State is a party on the record, the question of jurisdiction is decided by inspection. If jurisdiction depend, not on this plain fact, but on the interest of the State, what rule has the constitution given, by which this interest [853] is to be measured? If no rule be given, is it to be settled by the Court? If so, the curious anomaly is presented, of a Court examining the whole testimony of a cause, inquiring into, and deciding on, the extent of a State's interest, without having a right to exercise any jurisdiction in the case. Can this inquiry be made without the exercise of jurisdiction?

The next in the enumeration, is a controversy between a State and the citizens of another State.

Can this case arise, if the State be not a party on the record? If it can, the question recurs, what degree of interest shall be sufficient to change the parties, and arrest the proceedings against the individual? Controversies respecting boundary have lately existed between Virginia and Tennessee, between Kentucky and Tennessee, and now exist between New-York and New-Jersey. Suppose, while such a controversy is pending, the collecting officer of one State should seize property for taxes belonging to a man who supposes himself to reside in the other State, and who seeks redress in the federal Court of that State in which the officer resides. The interest of the State is obvious. Yet it is admitted, that in such a case the action would lie, because the officer might be treated as a trespasser, and the verdict and judgment against him would not act directly on the property of the State. That it would not so act, may, perhaps, depend on circumstances. The officer may retain the amount of the taxes in his hands, and, on the proceedings of the State against him, may plead in bar the judgment of a Court of [854] competent jurisdiction. If this plea ought to be sustained, and it is far from being certain that it ought not, the judgment so pleaded would have acted directly on the revenue of the State, in the hands of its officer. And yet the argument admits, that the action, in such a case, would be sustained. But, suppose, in such a case, the party conceiving himself to be injured, instead of bringing an action sounding in damages, should sue for the specific thing, while yet in possession of the seizing officer. It being admitted, in argument, that the action sounding in damages would lie, we are unable to perceive the line of distinction between that and the action of detinue. Yet the latter action would claim the specific article seized for the tax, and would obtain it, should the seizure be deemed unlawful.

It would be tedious to pursue this part of the inquiry farther, and it would be useless, because every person will perceive that the same reasoning is applicable to all the other enumerated controversies to which a State may be a party. The principle may be illustrated by a reference to those other controversies where jurisdiction depends on the party. But, before we review them, we will notice one where the nature of the controversy is, in some degree, blended with the character of the party.

If a suit be brought against a foreign minister, the Supreme Court alone has original jurisdiction, and this is shown on the record. But, suppose a suit to be brought which affects the interest of a foreign minister, or by which the person of his secretary, [855] or of his servant, is arrested. The minister does not, by the mere arrest of his secretary, or his servant, become a party to this suit, but the actual defendant pleads to the jurisdiction of the Court, and asserts his privilege. If the suit affects a foreign minister, it must be dismissed, not because he is a party to it, but because it affects him. The language of the constitution in the two cases is different. This Court can take cognizance of all cases "affecting" foreign ministers; and, therefore, jurisdiction does not depend on the party named in the record. But this language changes, when the enumeration proceeds to States. Why this change? The answer is obvious. In the case of foreign ministers, it was intended, for reasons which all comprehend, to give the national Courts jurisdiction over all cases by which they were in any manner affected. In the case of States, whose immediate or remote interests were mixed up with a multitude of cases, and who might be affected in an almost infinite variety of ways, it was intended to give jurisdiction in those cases only to which they were actual parties.

In proceeding with the cases in which jurisdiction depends on the character of the party, the first in the enumeration is, "controversies to which the United States shall be a party."

Does this provision extend to the cases where the United States are not named in the record, but claim, and are actually entitled to, the whole subject in controversy?

Let us examine this question.

Suits brought by the Postmaster-General are [856] for money due to the United States. The nominal plaintiff has no interest in the controversy, and the United States are the only real party. Yet, these suits could not be instituted in the Courts of the Union, under that clause which gives jurisdiction in all cases to which the United States are a party; and it was found necessary to give the Court jurisdiction over them, as being cases arising under a law of the United States.

The judicial power of the Union is also extended to controversies between citizens of different States; and it has been decided, that the character of the parties must be shown on the record. Does this provision depend on the character of those whose interest is litigated, or of those who are parties on the record? In a suit, for example, brought by or against an executor, the creditors or legatees of his testator are the persons really concerned in interest; but it has never been suspected that, if the executor be a resident of another State, the jurisdiction of the federal Courts could be ousted by the fact, that the creditors or legatees were citizens of the same State with the opposite party. The universally received construction in this case is, that jurisdiction is neither given nor ousted by the relative situation of the parties concerned in interest, but by the relative situation of the parties named on the record. Why is this construction universal? No case can be imagined, in which the existence of an interest out of the party on the record is more unequivocal than in that which has been just stated. Why, then, is it universally admitted, that this interest in [857] no manner affects the jurisdiction of the Court? The plain and obvious answer is, because the jurisdiction of the Court depends, not upon this interest, but upon the actual party on the record.

Were a State to be the sole legatee, it will not, we presume, be alleged, that the jurisdiction of the Court, in a suit against the executor, would be more affected by this fact, than by the fact that any other person, not suable in the Courts of the Union, was the sole legatee. Yet, in such a case, the Court would decide directly and immediately on the interest of the State.

This principle might be further illustrated by showing that jurisdiction, where it depends on the character of the party, is never conferred in consequence of the existence of an interest in a party not named; and by showing that, under the distributive clause of the 2d section of the 3d article, the Supreme Court could never take original jurisdiction, in consequence of an interest in a party not named in the record.

But the principle seems too well established to require that more time should be devoted to it. It may, we think, be laid down as a rule which admits of no exception, that, in all cases where jurisdiction depends on the party, it is the party named in the record. Consequently, the 11th amendment, which restrains the jurisdiction granted by the constitution over suits against States, is, of necessity, limited to those suits in which a State is a party on the record. The amendment has its full effect, if the constitution be construed as it [858] would have been construed, had the jurisdiction of the Court never been extended to suits brought against a State, by the citizens of another State, or by aliens.

The State not being a party on the record, and the Court having jurisdiction over those who are parties on the record, the true question is, not one of jurisdiction, but whether, in the exercise of its jurisdiction, the Court ought to make a decree against the defendants; whether they are to be considered as having a real interest, or as being only nominal parties.

In pursuing the arrangement which the appellants have made for the argument of the cause, this question has already been considered. The responsibility of the officers of the State for the money taken out of the Bank, was admitted, and it was acknowledged that this responsibility might be enforced by the proper action. The objection is, to its being enforced against the specific article taken, and by the decree of this Court. But, it has been shown, we think, that an action of detinue might be maintained for that article, if the Bank had possessed the means of describing it, and that the interest of the State would not have been an obstacle to the suit of the Bank against the individual in possession of it. The judgment in such a suit might have been enforced, had the article been found in possession of the individual defendant. It has been shown, that the danger of its being parted with, of its being lost to the plaintiff, and the necessity of a discovery, justified the application to a Court of equity. It was in a [859] Court of equity alone that the relief would be real, substantial, and effective. The parties must certainly have a real interest in the case, since their personal responsibility is acknowledged, and, if denied, could be demonstrated.

It was proper, then, to make a decree against the defendants in the Circuit Court, if the law of the State of Ohio be repugnant to the constitution, or to a law of the United States made in pursuance thereof, so as to furnish no authority to those who took, or to those who received, the money for which this suit was instituted.

7. Is that law unconstitutional?

This point was argued with great ability, and decided by this Court, after mature and deliberate consideration, in the case of M'Culloch v. The State of Maryland. A revision of that opinion has been requested; and many considerations combine to induce a review of it.

The foundation of the argument in favour of the right of a State to tax the Bank, is laid in the supposed character of that institution. The argument supposes the corporation to have been originated for the management of an individual concern, to be founded upon contract between individuals, having private trade and private profit for its great end and principal object.

If these premises were true, the conclusion drawn from them would be inevitable. This mere private corporation, engaged in its own business, with its own views, would certainly be subject to the taxing power of the State, as any individual would be; and the casual circumstance of its being [860] employed by the government in the transaction of its fiscal affairs, would no more exempt its private business from the operation of that power, than it would exempt the private business of any individual employed in the same manner. But the premises are not true. The Bank is not considered as a private corporation, whose principal object is individual trade and individual profit; but as a public corporation, created for public and national purposes. That the mere business of banking is, in its own nature, a private business, and may be carried on by individuals or companies having no political connexion with the government, is admitted; but the Bank is not such an individual or company. It was not created for its own sake, or for private purposes. It has never been supposed that Congress could create such a corporation. The whole opinion of the Court, in the case of M'Culloch v. The State of Maryland, is founded on, and sustained by, the idea that the Bank is an instrument which is "necessary and proper for carrying into effect the powers vested in the government of the United States." It is not an instrument which the government found ready made, and has supposed to be adapted to its purposes; but one which was created in the form in which it now appears, for national purposes only. It is, undoubtedly, capable of transacting private as well as public business. While it is the great instrument by which the fiscal operations of the government are effected, it is also trading with individuals for its own advantage. The appellants endeavor to distinguish between this trade and its [861] agency for the public, between its Banking operations and those qualities which it possesses in common with every corporation, such as individuality, immortality, &c.; While they seem to admit the right to preserve this corporate existence, they deny the right to protect it in its trade and business.

If there be any thing in this distinction, it would tend to show that so much of the act as incorporates the Bank is constitutional, but so much of it as authorizes its Banking operations is unconstitutional. Congress can make the inanimate body, and employ the machine as a depository of, and vehicle for, the conveyance of the treasure of the nation, if it be capable of being so employed, but cannot breathe into it the vital spirit which alone can bring it into useful existence.

Let this distinction be considered.

Why is it that Congress can incorporate or create a Bank? This question was answered in the case of M`Culloch v. The State of Maryland. It is an instrument which is "necessary and proper" for carrying on the fiscal operations of government. Can this instrument, on any rational calculation, effect its object, unless it be endowed with that faculty of lending and dealing in money, which is conferred by its charter? If it can, if it be as competent to the purposes of government without, as with this faculty, there will be much difficulty in sustaining that essential part of the charter. If it cannot, then this faculty is necessary to the legitimate operations of government, and was constitutionally and rightfully engrafted on the institution. It is, in that view of the subject, [862] the vital part of the corporation; it is its soul; and the right to preserve it originates in the same principle, with the right to preserve the skeleton or body which it animates. The distinction between destroying what is denominated the corporate franchise, and destroying its vivifying principle, is precisely as incapable of being maintained, as a distinction between the right to sentence a human being to death, and a right to sentence him to a total privation of sustenance during life. Deprive a Bank of its trade and business, which is its sustenance, and its immortality, if it have that property, will be a very useless attribute.

This distinction, then, has no real existence. To tax its faculties, its trade, and occupation, is to tax the Bank itself? To destroy or preserve the one, is to destroy or preserve the other.

It is urged, that Congress has not, by this act of incorporation, created the faculty of trading in money; that it had anterior existence, and may be carried on by a private individual, or company, as well as by a corporation. As this profession or business may be taxed, regulated, or restrained, when conducted by an individual, it may, likewise, be taxed, regulated, or restrained, when conducted by a corporation.

The general correctness of these propositions need not be controverted. Their particular application to the question before the Court, is alone to be considered. We do not maintain that the corporate character of the Bank exempts its operations from the action of State authority. If an individual were to be endowed with the same faculties, [863] for the same purposes, he would be equally protected in the exercise of those faculties. The operations of the Bank are believed not only to yield the compensation for its services to the government, but to be essential to the performance of those services. Those operations give its value to the currency in which all the transactions of the government are conducted. They are, therefore, inseparably connected with those transactions. They enable the Bank to render those services to the nation for which it was created, and are, therefore, of the very essence of its character, as national instruments. The business of the Bank constitutes its capacity to perform its functions, as a machine for the money transactions of the government. Its corporate character is merely an incident, which enables it to transact that business more beneficially.

Were the Secretary of the Treasury to be authorized, by law, to appoint agencies throughout the Union, to perform the public functions of the Bank, and to be endowed with its faculties, as a necessary auxiliary to those functions, the operations of those agents would be as exempt from the control of the States as the Bank, and not more so. If, instead of the Secretary of the Treasury, a distinct office were to be created for the purpose, filled by a person who should receive, as a compensation for his time, labour, and expense, the profits of the banking business, instead of other emoluments, to be drawn from the treasury, which banking business was essential to the operations of the government, would each State in the Union possess a right to [864] control these operations? The question on which this right would depend must always be, are these faculties so essential to the fiscal operations of the government, as to authorize Congress to confer them? Let this be admitted, and the question, does the right to preserve them exist? must always be answered in the affirmative.

Congress was of opinion that these faculties were necessary, to enable the Bank to perform the services which are exacted from it, and for which it was created. This was certainly a question proper for the consideration of the national Legislature. But, were it now to undergo revision, who would have the hardihood to say, that, without the employment of a banking capital, those services could be performed? That the exercise of these faculties greatly facilitates the fiscal operations of the government, is too obvious for controversy; and who will venture to affirm, that the suppression of them would not materially affect those operations, and essentially impair, if not totally destroy, the utility of the machine to the government? The currency which it circulates, by means of its trade with individuals, is believed to make it a more fit instrument for the purposes of government, than it could otherwise be; and, if this be true, the capacity to carry on this trade, is a faculty indispensable to the character and objects of the institution.

The appellants admit, that, if this faculty be necessary, to make the Bank a fit instrument for the purposes of the government, Congress possesses the same power to protect the machine in [865] this, as in its direct fiscal operations; but they deny that it is necessary to those purposes, and insist that it is granted solely for the benefit of the members of the corporation. Were this proposition to be admitted, all the consequences which are drawn from it might follow. But it is not admitted. The Court has already stated its conviction, that without this capacity to trade with individuals, the Bank would be a very defective instrument, when considered with a single view to its fitness for the purposes of government. On this point the whole argument rests.

It is contended, that, admitting Congress to possess the power, this exemption ought to have been expressly asserted in the act of incorporation; and, not being expressed, ought not to be implied by the Court.

It is not unusual, for a legislative act to involve consequences which are not expressed. An officer, for example, is ordered to arrest an individual. It is not necessary, nor is it usual, to say that he shall not be punished for obeying this order. His security is implied in the order itself. It is no unusual thing for an act of Congress to imply, without expressing, this very exemption from State control, which is said to be so objectionable in this instance. The collectors of the revenue, the carriers of the mail, the mint establishment, and all those institutions which are public in their nature, are examples in point. It has never been doubted, that all who are employed in them, are protected, while in the line of duty; and yet this protection is not expressed in any act of Congress. It is incidental [866] to, and is implied in, the several acts by which these institutions are created, and is secured to the individuals employed in them, by the judicial power alone; that is, the judicial power is the instrument employed by the government in administering this security.

That department has no will, in any case. If the sound construction of the act be, that it exempts the trade of the Bank, as being essential to the character of a machine necessary to the fiscal operations of the government, from the control of the States, Courts are as much bound to give it that construction, as if the exemption had been established in express terms. Judicial power, as contradistinguished from the power of the laws, has no existence. Courts are the mere instruments of the law, and can will nothing. When they are said to exercise a discretion, it is a mere legal discretion, a discretion to be exercised in discerning the course prescribed by law; and, when that is discerned, it is the duty of the Court to follow it. Judicial power is never exercised for the purpose of giving effect to the will of the Judge; always for the purpose of giving effect to the will of the Legislature; or, in other words, to the will of the law.

The appellants rely greatly on the distinction between the Bank and the public institutions, such as the mint or the post office. The agents in those offices are, it is said, officers of government, and are excluded from a seat in Congress. Not so the directors of the Bank. The connexion of the government with the Bank, is likened to that with contractors.

It will not be contended, that the directors, or [867] other officers of the Bank, are officers of government. But it is contended, that, were their resemblance to contractors more perfect than it is, the right of the State to control its operations, if those operations be necessary to its character, as a machine employed by the government, cannot be maintained. Can a contractor for supplying a military post with provisions, be restrained from making purchases within any State, or from transporting the provisions to the place at which the troops were stationed? or could he be fined or taxed for doing so? We have not yet heard these questions answered in the affirmative. It is true, that the property of the contractor may be taxed, as the property of other citizens; and so may the local property of the Bank. But we do not admit that the act of purchasing, or of conveying the articles purchased, can be under State control.

If the trade of the Bank be essential to its character, as a machine for the fiscal operations of the government, that trade must be as exempt from State control as the actual conveyance of the public money. Indeed, a tax bears upon the whole machine; as well upon the faculty of collecting and transmitting the money of the nation, as on that of discounting the notes of individuals. No distinction is taken between them.

Considering the capacity of carrying on the trade of banking, as an important feature in the character of this corporation, which was necessary, to make it a fit instrument for the objects for which it was created, the Court adheres to its decision in the case of M`Culloch against The State [868] of Maryland, and is of opinion, that the act of the State of Ohio, which is certainly much more objectionable than that of the State of Maryland, is repugnant to a law of the United States, made in pursuance of the constitution, and, therefore, void. The counsel for the appellants are too intelligent, and have too much self respect, to pretend, that a void act can afford any protection to the officers who execute it. They expressly admit that it cannot.

It being then shown, we think conclusively, that the defendants could derive neither authority nor protection from the act which they executed, and that this suit is not against the State of Ohio within the view of the constitution, the State being no party on the record, the only real question in the cause is, whether the record contains sufficient matter to justify the Court in pronouncing a decree against the defendants? That this question is attended with great difficulty, has not been concealed or denied. But when we reflect that the defendants, Osborne and Harper, are incontestably liable for the full amount of the money taken out of the Bank; that the defendant, Currie, is also responsible for the sum received by him, it having come to his hands with full knowledge of the unlawful means by which it was acquired; that the defendant, Sullivan, is also responsible for the sum specifically delivered to him, with notice that it was the property of the Bank, unless the form of having made an entry on the books of the treasury can countervail the fact, that it was, in truth, kept untouched, in a trunk, by itself, as a deposit, to await [869] the event of the pending suit respecting it; we may lay it down as a proposition, safely to be affirmed, that all the defendants in the cause were liable in an action at law for the amount of this decree. If the original injunction was properly awarded, for the reasons stated in the preceding part of this opinion, the money, having reached the hands of all those to whom it afterwards came with notice of that injunction, might be pursued, so long as it remained a distinct deposit, neither mixed with the money of the treasury, nor put into circulation. Were it to be admitted, that the original injunction was not properly awarded, still the amended and supplemental bill, which brings before the Court all the parties who had been concerned in the transaction, was filed after the cause of action had completely accrued. The money of the Bank had been taken, without authority, by some of the defendants, and was detained by the only person who was not an original wrong doer, in a specific form; so that detinue might have been maintained for it, had it been in the power of the Bank to prove the facts which are necessary to establish the identity of the property sued for. Under such circumstances, we think, a Court of equity may afford its aid, on the ground that a discovery is necessary, and also on the same principle that an injunction issues to restrain a person who has fraudulently obtained possession of negotiable notes, from putting them into circulation; or a person having the apparent ownership of stock really belonging to another, from transferring it. The suit, then, might be as well sustained in a [870] Court of equity as in a Court of law, and the objection that the interests of the State are committed to subordinate agents, if true, is the unavoidable consequence of exemption from being sued — of sovereignty. The interests of the United States are sometimes committed to subordinate agents. It was the case in Hoyt v. Gelston, in the case of The Apollon, and in the case of Doddridge's Lessee v. Thompson and Wright, and in many others. An independent foreign sovereign cannot be sued, and does not appear in Court. But a friend of the Court comes in, and, by suggestion, gives it to understand, that his interests are involved in the controversy. The interests of the sovereign, in such a case, and in every other where he chooses to assert them under the name of the real party to the cause are as well defended as if he were a party to the record. But his pretensions, where they are not well founded, cannot arrest the right of a party having a right to the thing for which he sues. Where the right is in the plaintiff, and the possession in the defendant, the inquiry cannot be stopped by the mere assertion of title in a sovereign. The Court must proceed to investigate the assertion, and examine the title. In the case at bar, the tribunal established by the constitution, for the purpose of deciding, ultimately, in all cases of this description, had solemnly determined, that a State law imposing a tax on the Bank of the United States, was unconstitutional and void, before the wrong was committed for which this suit was brought.

We think, then, that there is no error in the decree [871] of the Circuit Court for the district of Ohio, so far as it directs restitution of the specific sum of 98,000 dollars, which was taken out of the Bank unlawfully, and was in the possession of the defendant, Samuel Sullivan, when the injunction was awarded, in September, 1820, to restrain him from paying it away, or in any manner using it; and so far as it directs the payment of the remaining sum of 2000 dollars, by the defendants, Ralph Osborne and John L. Harper; but that the same is erroneous, so far as respects the interest on the coin, part of the said 98,000 dollars, it being the opinion of this Court, that, while the parties were restrained by the authority of the Circuit Court from using it, they ought not to be charged with interest. The decree of the Circuit Court for the district of Ohio is affirmed, as to the said sums of 98,000 dollars, and 2000 dollars; and reversed, as to the residue.

Mr. Justice JOHNSON.

The argument in this cause presents three questions: 1. Has Congress granted to the Bank of the United States, an unlimited right of suing in the Courts of the United States? 2. Could Congress constitutionally grant such a right? and 3. Has the power of the Court been legally and constitutionally exercised in this suit?

I have very little doubt that the public mind will be easily reconciled to the decision of the Court here rendered; for, whether necessary or unnecessary originally, a state of things has now grown up, in some of the States, which renders all [872] the protection necessary, that the general government can give to this Bank. The policy of the decision is obvious, that is, if the Bank is to be sustained; and few will bestow upon its legal correctness, the reflection, that it is necessary to test it by the constitution and laws, under which it is rendered.

The Bank of the United States, is now identified with the administration of the national government. It is an immense machine, economically and beneficially applied to the fiscal transactions of the nation. Attempts have been made to dispense with it, and they have failed; serious and very weighty doubts have been entertained of its constitutionality, but they have been abandoned; and it is now become the functionary that collects, the depository that holds, the vehicle that transports, the guard that protects, and the agent that distributes and pays away, the millions that pass annually through the national treasury; and all this, not only without expense to the government, but after paying a large bonus, and sustaining actual annual losses to a large amount; furnishing the only possible means of embodying the most ample security for so immense a charge.

Had its effects, however, and the views of its framers, been confined exclusively to its fiscal uses, it is more than probable that this suit, and the laws in which it originated, would never have had existence. But it is well known, that with that object was combined another, of a very general, and not less important character.

The expiration of the charter of the former Bank, led to State creations of Banks; each new Bank increased [873] the facilities of creating others; and the necessities of the general government, both to make use of the State Banks for their deposits, and to borrow largely of all who would lend to them, produced that rage for multiplying Banks, which, aided by the emoluments derived to the States in their creation, and the many individual incentives which they developed, soon inundated the country with a new description of bills of credit, against which it was obvious that the provisions of the constitution opposed no adequate inhibition.

A specie-paying Bank, with an overwhelming capital, and the whole aid of the government deposits, presented the only resource to which the government could resort, to restore that power over the currency of the country, which the framers of the constitution evidently intended to give to Congress alone. But this necessarily involved a restraint upon individual cupidity, and the exercise of State power; and, in the nature of things, it was hardly possible for the mighty effort necessary to put down an evil spread so wide, and arrived to such maturity, to be made without embodying against it an immense moneyed combination, which could not fail of making its influence to be felt, wherever its claimances could reach, or its industry and wealth be brought to operate.

I believe, that the good sense of a people, who know that they govern themselves, and feel that they have no interests distinct from those of their government, would readily concede to the Bank, thus circumstanced, some, if not all the rights here [874] contended for. But I cannot persuade myself, that they have been conceded in the extent which this decision affirms. Whatever might be proper to be done by an amendment of the constitution, this Court is only, at present, expounding its existing provisions.

In the present instance, I cannot persuade myself, that the constitution sanctions the vesting of the right of action in this Bank, in cases in which the privilege is exclusively personal, or in any case, merely on the ground that a question might possibly be raised in it, involving the constitution, or constitutionality of a law, of the United States.

When laws were heretofore passed for raising a revenue by a duty on stamped paper, the tax was quietly acquiesced in, notwithstanding it entrenched so closely on the unquestionable power of the States over the law of contracts; but had the same law which declared void contracts not written upon stamped paper, declared, that every person holding such paper should be entitled to bring his action "in any Circuit Court" of the United States, it is confidently believed that there could have been but one opinion on the constitutionality of such a provision. The whole jurisdiction over contracts, might thus have been taken from the State Courts, and conferred upon those of the United States. Nor would the evil have rested there; by a similar exercise of power, imposing a stamp on deeds generally, jurisdiction over the territory of the State, whoever might be parties, even between citizens of the same State — jurisdiction of suits instituted for the recovery of legacies [875] or distributive portions of intestates' estates — jurisdiction, in fact, over almost every possible case, might be transferred to the Courts of the United States. Wills may be required to be executed on stamped paper; taxes may be, and have been, imposed upon legacies and distributions; and, in all such cases, there is not only a possibility, but a probability, that a question may arise, involving the constitutionality, construction, &c.; of a law of the United States. If the circumstance, that the questions which the case involves, are to determine its character, whether those questions be made in the case or not, then every case here alluded to, may as well be transferred to the jurisdiction of the United States, as those to which this Bank is a party. But still farther, as was justly insisted in argument, there is not a tract of land in the United States, acquired under laws of the United States, whatever be the number of mesne transfers that it may have undergone, over which the jurisdiction of the Courts of the United States might not be extended by Congress, upon the very principle on which the right of suit in this Bank is here maintained. Nor is the case of the alien, put in argument, at all inapplicable. The one acquires its character of individual property, as the other does his political existence, under a law of the United States; and there is not a suit which may be instituted to recover the one, nor an action of ejectment to be brought by the other, in which a right acquired under a law of the United States, does not lie as essentially at the basis of the right of action, as in the suits brought by this Bank. [876] It is no answer to the argument, to say, that the law of the United States is but ancillary to the constitution, as to the alien; for the constitution could do nothing for him without the law: and, whether the question be upon law or constitution, still if the possibility of its arising be a sufficient circumstance to bring it within the jurisdiction of the United States Courts, that possibility exists with regard to every suit affected by alien disabilities; to real actions in time of peace — to all actions in time of war.

I cannot persuade myself, then, that, with these palpable consequences in view, Congress ever could have intended to vest in the Bank of the United States, the right of suit to the extent here claimed. And, notwithstanding the confidence with which this point has been argued, an examination of the terms of the act, and a consideration of them with a view to the context, will be found to leave it by no means a clear case, that such is the legal meaning of the act of incorporation. To be sure, if the act had simply and substantively given the right "to sue and be sued in the Circuit Courts of the United States," there could have been no question made upon the construction of those words. But such is not the fact. The words are, not that the Bank shall be made able and capable in law, to sue, &c.;, but that it shall, "by a certain name," be made able and capable in law to do the various acts therein enumerated. And these words, under the force of which this suit is instituted, are found in the ordinary incorporating clause of this act, a clause [877] which is well understood to be, and which this Court, in the case of Deveaux, has recognised to be, little more than the mere common place or formula of such an act. The name of a corporation is the symbol of its personal existence; a misnomer there is fatal to a suit, (and still more fatal as to other transactions.) By the incorporating clause, a name is given it, and, with that name, a place among created beings; then usually follows an enumeration of the ordinary acts in which it may personate a natural man; and among those acts, the right to sue and be sued, of which the Court, in Deveaux's case, very correctly remarks, that it is "a power which if not incident to a corporation, is conferred by every incorporating act, and is not understood to enlarge the jurisdiction of any particular Court, but to give a capacity to the corporation to appear as a corporation in any Court which would by law have cognizance of the cause if brought by individuals." With this qualification, the clause in question will be construed, as an enumeration of incidents, instead of a string of enactments; and such a construction is strongly countenanced by the concluding sentence of the section; for, after running through the whole routine of powers, most of which are unquestionably incidental, and needed no enactment to vest them, it concludes thus: "and generally to do and execute all and singular the acts, matters, and things, which to them it shall and may appertain to do." And, in going over the act, it will be found, that whenever it is contemplated to vest a power not incidental, it is done by a specific provision, made [878] the subject of a distinct clause; such is that power to transact the business of the loan-office of the United States. And, indeed, there is one section of the act, which strikingly exhibits the light in which the law-makers considered the incorporating clause. I mean the tenth; which, notwithstanding that the same clause in the seventh section, which is supposed to confer this sweeping power to sue, confers also, in terms equally comprehensive, the power to make laws for the institution, and "to do and execute all and singular the matters and things, which to them it shall and may appertain to do," contains an enactment in the following words: "that they shall have power to appoint such officers, clerks, and servants, under them, for executing the business of the corporation, and to allow them such compensation for their services respectively, as shall be reasonable; and shall be capable of exercising such other powers and authorities for the well governing and ordering the officers of the said corporation, as shall be prescribed by the laws, regulations, and ordinances, of the same;" a section which would have been altogether unnecessary, had the seventh section been considered as enacting, instead of enumerating and limiting. I consider the incorporating clause, then, not as purporting the absolute investment of any power, but as the usual and formal declaration of the extent to which this artificial should personate the natural person, in the transactions incident to ordinary life, or to the peculiar objects of its creation; and, therefore, not vesting the right to sue in the Courts of the United [879] States, but only the right of personating the natural man in the Courts of the United States, as it might, upon general principles, in any other Courts of competent jurisdiction. And this, I say, is consonant to the decision in Deveaux's case, and sustained by abundant evidence on the face of the act itself. Indeed, any other view of the effect of the section, converts some of its provisions into absolute nonsense.

It has been argued, and I have no objection to admit, that the phraseology of this act has been varied from that incorporating the former Bank, with a view to meet the decision in Deveaux's case. But it is perfectly obvious, that in the prosecution of that design, the purport of Deveaux's case has been misapprehended. The Court there decide, that the jurisdiction of the United States depended, (1.) on the character of the cause, (2.) on the character of the parties; that the Judiciary Act confined the jurisdiction of the Circuit Courts to the second class of cases, and the incorporating act contained no words that purported to carry it further. Whether the legislative power of the United States could extend it as far as is here insisted on, or what words would be adequate to that purpose, the case neither called on the Court to decide, nor has it proposed to decide. If any thing is to be inferred from that decision on those points, it is unfavourable to the sufficiency of the words inserted in the present act. For, the argument of the Court intimates, that where the Legislature propose to give jurisdiction to the Courts of the United States, they do [880] it by a separate provision, as in the case of the action of debt for exceeding the sum authorized to be loaned. And on the words of the incorporating section, it makes this remark, "that it is not understood to enlarge the jurisdiction of any particular Court, but to give a capacity to the corporation to appear as a corporation in any Court, which would by law have cognizance of the cause if brought by individuals. If jurisdiction is given by this clause to the federal Courts, it is equally given to all Courts having original jurisdiction, and for all sums, however small they be." Now, the difference of phraseology between the former act and the present, in the clause in question, is this: the former has these words, "may sue and be sued, &c.; in Courts of record or any other place whatsoever;" the present act has substituted these words, "in all State Courts having competent jurisdiction, and in any Circuit Court of the United States." Now, the defect here could not have been the want of adequate words, had the intent appeared to have been, to enlarge the jurisdiction of any particular Court. For, if the Circuit Courts were Courts of record, the right of suit given was as full as any other words could have made it. But, as the Court in its own words assigns the ground of its decision, the clause could not have been intended to enlarge the jurisdiction of the State Courts, and therefore could not have been intended to enlarge that of the federal Courts, much less to have extended it to the smallest sum possible. Therefore it concludes, that the clause is one of mere enumeration, containing, [881] as it expresses it, "the powers which, if not incident to a corporation, are conferred by every incorporating act, and are not understood to enlarge," &c.; If, then, this variation had in view the object which is attributed to it, the words intended to answer that object have been inserted so unhappily as to neutralize its influence; but, I think it much more consistent with the respect due to the draftsman, who was known to have been an able lawyer, to believe that, with such an object in view, he would have pursued a much more plain and obvious course, and given it a distinct and unequivocal section to itself, or at least have worded it with more marked attention. This opinion is further supported, by considering the absurdities that a contrary opinion would lead to.

A literal translation of the words in question is impossible. Nothing but inconsistencies present themselves, if we attempt to apply it without a reference to the laws and constitution of the United States, forming together the judicial system of the Union. The words are, "may sue and be sued, &c.;, in any State Court having competent jurisdiction, and in any Circuit Court of the United States." But why should one member of the passage be entitled to an enacting effect, and not the residue? Yet, who will impute to the Legislature or the draftsman, and intention to vest a jurisdiction by these words in a State Court? I do not speak of the positive effect; since the failure of one enactment, because of a want either of power to give or capacity to receive, will not control [882] the effect as to any other enactment. I speak of the intent or understanding of the law-maker; who must have used these words, as applicable to the State Courts, in an enacting sense, if we suppose him to have used them in that sense, as to the Courts of the United States. Yet I should be very unwilling to impute to him, or to the Legislature of the country, ignorance of the fact, that such an enactment, if it was one, could not give a right to sue in the State Courts, if the right did not exist without it. Or, in fact, that such enactment was altogether unnecessary, if the legislative power, which must give effect to such an enactment, was adequate to constitute effectually this body corporate.

But why should this supposed enactment go still farther, and confer the capacity to be sued, as well as to sue, either in the Courts of the one jurisdiction or the other? Did the lawgivers suppose that this corporation would not be subject to suit, without an express enactment for that purpose also? Or was it guilty of the more unaccountable mistake, of supposing that it could confer upon individuals, indiscriminately, this privilege of bringing suits in the Courts of the United States against the Bank? that too, for a cause of action originating, say, in work and labour, or in a special action on the case, or perhaps, ejectment to try title to land mortgaged by a person not having the estate in him, or purchased of a tortious holder for a banking house? I cannot acquiesce in the supposition; and yet, if one is an enactment, and [883] takes effect as such, they are all enactments, for they are uttered eodem flatu.

My own conclusion is, that none of them are enactments, but all merely declaratory; or, at most, only enacting, in the words of the Court, in the case of Deveaux, that the Bank may, by its corporate name and metaphysical existence, bring suit, or personate the natural man, in the Courts specified, as though it were in fact a natural person; that is, in those cases in which, according to existing laws, suits may be brought in the Courts specified respectively.

Indeed, a more unrestricted sense given to the words of the act, could not be carried into execution; a literal exercise of the right of suit, supposed to be granted, would be impossible. Can the Bank of the United States be sued (in the literal language of the act) "in any Circuit Court of the United States?" in that of Ohio, or Louisiana, for instance? Locality, in this respect, cannot be denied to such an institution; or, at least, it is only incidentally, by distress infinite, or attachment, for instance, that such a suit could be maintained. Nor, on the other hand, could the Bank sue literally in any Circuit Court of the United States. It must, of necessity, be confined to the Circuit Court of that district in which the defendant resides, or is to be found. And thus, at last, we circumscribe these general words, by reference to the judicial system of the United States, as it existed at the time. And why the same restriction should not have been imposed as to amount, which is imposed as to all other suitors, [884] to wit, 500 dollars and upwards, is to me inscrutable, except on the supposition that this clause was not intended for any other purpose than that which I have supposed. The United States have suffered no other suitors to institute a suit in its Courts for less than that sum, and it is hard to conceive why the Bank should be permitted to institute a suit to recover, if it will, a single cent. This consideration is expressly drawn into notice by this Court, in the case of Deveaux, and if it was entitled to weight then, in fixing the construction of the incorporating section, I see no reason why it should be unnoticed now.

I will dwell no longer on a point, which is in fact secondary and subordinate; for if Congress can vest this jurisdiction, and the people will it, the act may be amended, and the jurisdiction vested. I next proceed to consider, more distinctly, the constitutional question, on the right to vest the jurisdiction to the extent here contended for.

And here I must observe, that I altogether misunderstood the counsel, who argued the cause for the plaintiff in error, if any of them contended against the jurisdiction, on the ground that the cause involved questions depending on general principles. No one can question, that the Court which has jurisdiction of the principal question, must exercise jurisdiction over every question. Neither did I understand them as denying, that if Congress could confer on the Circuit Courts appellate, they could confer original jurisdiction. The argument went to deny the right to assume jurisdiction on a mere hypothesis. It was one of [885] description, identity, definition; they contended, that until a question involving the construction or administration of the laws of the United States did actually arise, the casus federis was not presented, on which the constitution authorized the government to take to itself the jurisdiction of the cause. That until such a question actually arose, until such a case was actually presented, non constat, but the cause depended upon general principles, exclusively cognizable in the State Courts; that neither the letter nor the spirit of the constitution sanctioned the assumption of jurisdiction on the part of the United States at any previous stage.

And this doctrine has my hearty concurrence in its general application. A very simple case may be stated, to illustrate its bearing on the question of jurisdiction between the two governments. By virtue of treaties with Great Britain, aliens holding lands were exempted from alien disabilities, and made capable of holding, aliening, and transmitting their estates, in common with natives. But why should the claimants of such lands, to all eternity, be vested with the privilege of bringing an original suit in the Courts of the United States? It is true, a question might be made, upon the effect of the treaty, on the rights claimed by or through the alien; but until that question does arise, nay, until a decision against the right takes place, what end has the United States to subserve in claiming jurisdiction of the cause? Such is the present law of the United States, as to all but this one distinguished party; and that law was [886] passed when the doctrines, the views, and ends of the constitution, were, at least, as well understood as they are at present. I attach much importance to the 25th section of the judiciary act, not only as a measure of policy, but as a contemporaneous exposition of the constitution on this subject; as an exposition of the words of the constitution, deduced from a knowledge of its views and policy. The object was, to secure a uniform construction and a steady execution of the laws of the Union. Except as far as this purpose might require, the general government had no interest in stripping the State Courts of their jurisdiction; their policy would rather lead to avoid incumbering themselves with it. Why then should it be vested with jurisdiction in a thousand causes, on a mere possibility of a question arising, which question, at last, does not occur in one of them? Indeed, I cannot perceive how such a reach of jurisdiction can be asserted, without changing the reading of the constitution on this subject altogether. The judicial power extends only to "cases arising," that is, actual, not potential cases. The framers of the constitution knew better, than to trust such a quo minus fiction in the hands of any government.

I have never understood any one to question the right of Congress to vest original jurisdiction in its inferior Courts, in cases coming properly within the description of "cases arising under the laws of the United States;" but surely it must first be ascertained, in some proper mode, that the cases are such as the constitution describes. By possibility, a constitutional question may be raised in [887] any conceivable suit that may be instituted; but that would be a very insufficient ground for assuming universal jurisdiction; and yet, that a question has been made, as that, for instance, on the Bank charter, and may again be made, seems still worse, as a ground for extending jurisdiction. For, the folly of raising it again in every suit instituted by the Bank, is too great, to suppose it possible. Yet this supposition, and this alone, would seem to justify vesting the Bank with an unlimited right to sue in the federal Courts. Indeed, I cannot perceive how, with ordinary correctness, a question can be said to be involved in a cause, which only may possibly be made, but which, in fact, is the very last question that there is any probability will be made; or rather, how that can any longer be denominated a question, which has been put out of existence by a solemn decision. The constitution presumes, that the decisions of the supreme tribunal will be acquiesced in; and after disposing of the few questions which the constitution refers to it, all the minor questions belong properly to the State jurisdictions, and never were intended to be taken away in mass.

Efforts have been made to fix the precise sense of the constitution, when it vests jurisdiction in the general government, in "cases arising under the laws of the United States." To me, the question appears susceptible of a very simple solution; that all depends upon the identity of the case supposed; according to which idea, a case may be such in its very existence, or it may become such in its progress. An action may "live, move, and have [888] its being," in a law of the United States; such is that given for the violation of a patent-right, and four or five different actions given by this act of incorporation; particularly that against the President and Directors for over-issuing; in all of which cases the plaintiff must count upon the law itself as the ground of his action. And of the other description, would have been an action of trespass, in this case, had remedy been sought for an actual levy of the tax imposed. Such was the case of the former Bank against Deveaux, and many others that have occurred in this Court, in which the suit, in its form, was such as occur in ordinary cases, but in which the pleadings or evidence raised the question on the law or constitution of the United States. In this class of cases, the occurrence of a question makes the case, and transfers it, as provided for under the twenty-fifth section of the Judiciary Act, to the jurisdiction of the United States. And this appears to me to present the only sound and practical construction of the constitution on this subject; for no other cases does it regard as necessary to place under the control of the general government. It is only when the case exhibits one or the other of these characteristics, that it is acted upon by the constitution. Where no question is raised, there can be no contrariety of construction; and what else had the constitution to guard against? As to cases of the first description, ex necessitate rei, the Courts of the United States must be susceptible of original jurisdiction; and as to all other cases, I should hold them, also, susceptible of original jurisdiction, if it were practicable, [889] in the nature of things, to make out the definition of the case, so as to bring it under the constitution judicially, upon an original suit. But until the plaintiff can control the defendant in his pleadings, I see no practical mode of determining when the case does occur, otherwise than by permitting the cause to advance until the case for which the constitution provides shall actually arise. If it never occurs, there can be nothing to complain of; and such are the provisions of the twenty-fifth section. The cause might be transferred to the Circuit Court before an adjudication takes place; but I can perceive no earlier stage at which it can possibly be predicated of such a case, that it is one within the constitution; nor any possible necessity for transferring it then, or until the Court has acted upon it to the prejudice of the claims of the United States. It is not, therefore, because Congress may not vest an original jurisdiction, where they can constitutionally vest in the Circuit Courts appellate jurisdiction, that I object to this general grant of the right to sue; but, because that the peculiar nature of this jurisdiction is such, as to render it impossible to exercise it in a strictly original form, and because the principle of a possible occurrence of a question as a ground of jurisdiction, is transcending the bounds of the constitution, and placing it on a ground which will admit of an enormous accession, if not an unlimited assumption, of jurisdiction.

But, dismissing the question of possibility, which, I must think, would embrace every other case as well as those to which this Bank is a party, in what [890] sense can it be predicated of this case, that it is one arising under a law of the United States? It cannot be denied, that jurisdiction of this suit in equity could not be entertained, unless the Court could have had jurisdiction of the action of trespass, which this injunction was intended to anticipate. And, in fact, there is no question, that the Bank here maintains, that the right to sue extends to common trespass, as well as to contracts, or any other cause of action. But suppose trespass in the common form instituted; the declaration is general, and the defendant pleads not guilty, and goes to trial. Where is the feature in such a cause that can give the Court jurisdiction? What question arises under a law of the United States? or what question that must not be decided exclusively upon the lex loci, upon State laws? Take also the case of a contract, and in what sense can it be correctly predicated of that, that in common with every other act of the Bank, it arises out of the law that incorporates it? May it not with equal propriety be asserted, that all the crimes and all the controversies of mankind, arise out of the fiat that called their progenitor into existence? It is not because man was created, that he commits a trespass, or incurs a debt; but because, being indued with certain faculties and propensities, he is led by an appropriate motive to the one action or the other. Sound philosophy attributes effects to their proximate causes. It is but pursuing the grade of creation from one step to another, to deduce the acts of this Bank from State law, or even divine law, with as much correctness as from the law of [891] its immediate creation. Its contracts arise under its own acts, and not under a law of the United States; so far from it, indeed, that their effect, their construction, their limitation, their concoction, are all the creatures of the respective State laws in which they originate. There is a satisfactory illustration of the distinction between contracts which draw their existence from statutes, and those which originate in the acts of man, afforded by this act of incorporation itself. It will be unnecessary to look beyond it. The action of debt before alluded to, given by the ninth clause of the seventh section, against the directors, to any one who will sue, is one of those factitious or statute contracts which exist in, and expire with, the statute that creates it. Not so with the ordinary contracts of the Bank; upon the expiration of the charter, they would be placed in the state of the credits of an intestate before administration; there is no one to sue for them; but the moral obligation would remain, and a Court of equity would enforce it against their debtors, at the suit of the individual stockholders. Nor would this be on the principle of contracts executed under power of attorney; for, the law applicable to principals would govern every question in such causes. All the acts of the corporation are executed in their own right, and not in the right of another. A personal existence, with all its incidents, is given to them, and it is in right of that existence that they are capable of acting, and do act.

Nor, indeed, in another point of view, is it strictly predicable of this Bank, that its acts arise [892] out of, because its existence is drawn from, a law of the United States. It is because it is incorporated, not because incorporated by a law of the United States, that it is made capable of exercising certain powers incidentally, and of being vested with others expressly. The same effects would follow, if incorporated by any other competent legislative power. The law of the United States creates the Bank, and the common law, or State law more properly, takes it up and makes it what it is. Who can deny, that in many points the incidents to such an institution may vary in different States, although its existence be derived from the general government? It is the case with the natural alien, when adopted into the national family. His rights, duties, powers, &c.;, receive always a shade from the lex loci of the State in which he fixes his domicil.

If this right to sue could be vested at all in the Bank, it is obvious that it must have been for one or more of three causes: 1. That a law of the United States incorporated it; 2. That a law of the United States vested in it the power to sue; or, 3. That the power to defend itself from trespasses as applicable to this case strictly, or to contract debts as applicable to the Georgia case, was conferred on it by a law of the United States expressly.

The first I have considered. On the second, no one would have the hardihood to contend, that such a grant has any efficacy, unless the suits come within the description of cases arising under a law of the United States, independently of the [893] grant of the right to sue; and it only remains to add a few more remarks on the third ground.

Of the power to repel trespasses, and to enter into contracts, as mere incidents to its creation, I trust I have shown, that neither comes within the description of a case arising under a law of the United States. But where will we find, in the law in question, any express grant of power relative to either? The contracts on which the Georgia case is founded, are declared on as common promissory notes, payable to bearer. Now, as mere incidents, I have no doubt of an action being sustainable in a State Court in both cases. But if an express grant is relied on, as bringing this, or the case of a contract, within the description of "a case arising under a law of the United States," then I look through the law in vain for any express grant, either to make the contract, or repel the trespass. It is true, the sweeping terms with which the incorporating section concludes, import, that "by that name it shall and may be lawful for the Bank to do and execute all and singular the acts, matters, and things, which to them it shall and may appertain to do." But this contains no grant of either, since the inquiry, at last, must be into the incidents of such an institution, and, as incidents, they needed not these words to sustain them; nor could those words give any more force to the right. So that, at last, we are referred to the mere fact of its corporate existence, for the basis of either of the actions, or either of the powers here insisted on, as bringing this cause within the constitutional definition. Having a legal [894] existence as an incorporated banking institution, it has a right to security in its possessions, and to the performance of its contracts; but that right will be precisely the same, if incorporated by a State law, or even, as was held in the case of Terrett v. Taylor, if having a common law corporate existence. The common law, or the State law, is referred to by the law of the United States, as the source of these incidents, when it speaks of the acts which are appurtenant to it; and I know of no other law that can define them, or confer them as incidents. Suppose a naturalization act passed, which, after specifying the terms and conditions upon which an alien shall become a citizen, proceeds to declare, "that, as a citizen, he shall lawfully do and execute all and singular the acts, matters, and things, which to `a citizen,' or `to him as a citizen,' it shall and may appertain to do," would not these words be a mere nullity? His new existence, and the relations with the society into which he is introduced, that grow out of that connexion, give him the right to defend his property or his existence, (as in this case,) and to enter into and enforce those contracts which, as an alien, he would have been precluded from. He was no more a citizen, without an act of Congress, than this was a Bank. Finally, after the most attentive consideration of this cause, I cannot help thinking, that this idea of taking jurisdiction upon an hypothesis, or even of assuming original, unlimited jurisdiction, of all questions arising under a law of the United States, involves some striking inconsistencies. A Court may take cognizance of a question [895] in a cause, and enter a judgment upon it, and yet not have jurisdiction of the cause itself. Such are all questions of jurisdiction, of which every Court, however limited its jurisdiction, must have cognizance in every cause brought before it. So, also, I see not why, upon the same principle, a law expressly violating the constitution, may not be made the groundwork of a transfer of jurisdiction. Cases may arise, and would arise, under such a law; and if the simple existence, or possibility of such a case, is a sufficient ground of jurisdiction, and that ground sufficient to transfer the whole case to the federal judiciary, the least that can be said of it is, that it was not a case within the mischief intended to be obviated by the constitution. I shall say no more on this subject, but proceed to one which also acts forcibly on my judgment in forming my opinion in this cause.

I will not undertake to define the limits within which the discretion of the Legislature of the Union may range, in the adoption of measures for executing their constitutional powers. It is very possible, that in the choice of means as "proper and necessary" to carry their powers into effect, they may have assumed a latitude not foreseen at the adoption of the constitution. For example, in order to collect a stamp duty, they have exercised a power over the general law of contracts; in order to secure debts due the United States, they have controlled the State laws of estates of deceased persons and of insolvents' estates; in the distributions and the powers of individuals themselves, when insolvent, in the assignment of their [896] own estates; in the exercise of various powers, they have taken jurisdiction over crimes which the State laws took cognizance of; and all this, being within the range of their discretion, is aloof from judicial control, while unaffectedly exercised for the purposes of the constitution. Nor, indeed, is there much to be alarmed at in it, while the same people who govern the States, can, where they will, control the Legislature of the United States.

Yet, certainly, there is one limit to this chain of implied powers, which must lie beyond the reach of legislative discretion. No one branch of the general government can new model the constitutional structure of the other.

Much stress was laid, in the argument, upon the necessity of giving co-ordinate extent to the several departments of a government; but it was altogether unnecessary to bring this consideration into the present case. As a ground of policy, this is not its proper place; and as a ground of construction, it must be needless, when applied to a constitution in which the judicial power so very far transcends both the others, in its acknowledged limits.

The principle is, that every government should possess the means of protecting itself; that is, of construing and enforcing its own laws. But this is not the half of the extent of the judicial power of the Union. Its most interesting province, is to enforce the equal administration of laws, and systems of laws, over which the legislative power can exercise no control. And thus, the judicial power is distributed into the two [897] classes: 1. That which is defined by the circumstances of the case; and, 2. That which depends upon the circumstances of the person. On the first, I have endeavoured to show, that the end is adequately effected by the provisions of the 25th section of the Judiciary Act, and, practically, can be exercised in no other way. But with regard to the second class, the argument turns against the United States; and every reason that may be urged in favour of eking out the jurisdiction in the first class of cases, reacts forcibly to confine the jurisdiction strictly within its constitutional limits, as to the second class. When the alien, or the citizen of another State, or the grants of another State, are implicated, the State Courts open their tribunals to the judiciary of the United States, and recognize their power as co-ordinate. Their citizens, their territory, their laws, all are subjected to a power quite foreign to the States, and judicial power is literally poured out upon the Courts of the Union, without stint.

How interesting, then, is it to the States, that the number of those persons who claim the privilege of coming into the Courts of the United States should be strictly limited! Cases, since they arise out of laws, &c.; of the United States, must be very limited in number; but persons may bring into the Courts of the United States any question and every question, and, if this law be correctly construed, for any, the very smallest possible amount.

But if the plain dictates of our senses be relied on, what state of facts have we exhibited here? [898] Making a person, makes a case; and thus, a government which cannot exercise jurisdiction unless an alien or citizen of another State be a party, makes a party which is neither alien nor citizen, and then claims jurisdiction because it has made a case. If this be true, why not make every citizen a corporation sole, and thus bring them all into the Courts of the United States quo minus? Nay, it is still worse, for there is not only an evasion of the constitution implied in this doctrine, but a positive power to violate it. Suppose every individual of this corporation were citizens of Ohio, or, as applicable to the other case, were citizens of Georgia, the United States could not give any one of them, individually, the right to sue a citizen of the same State in the Courts of the United States; then, on what principle could that right be communicated to them in a body? But the question is equally unanswerable, if any single member of the corporation is of the same State with the defendant, as has been repeatedly adjudged.

One of the counsel who argued this cause in behalf of the Bank, has denominated it a bundle of faculties. This is very true; but those faculties are substituted for the organization of a natural person; and it is perfectly certain, that when it comes into this Court, it must be treated as a person. It is altogether inadmissible, to refine away the principles of jurisprudence, so as to consider it in any other light than that of a person. As such, it sues out a writ, declares, pleads, takes judgment, and levies an execution. If it is not a [899] person, it has no standing in this Court; it must, therefore, abandon this suit, or be subjected to personal disabilities. Gentlemen have a right to take what ground here they please, to sustain this action; but it is perfectly clear to me, that the act of Congress was intended to vest this right as a personal right, or not at all. Let any one look through this act, and notice the unrestricted latitude that has been assumed in vesting the right to sue both by and against this Bank, and he will see, that either there is no general right to sue given in the seventh section, now relied on, or that it is given under the general power granted to pass all laws necessary to carry the powers of the general government into execution. The proviso to the 17th section is a remarkable proof of this. It puts the limits of judicial power altogether out of view. If Congress, in legislating on this subject, did intend such a grant as is here contended for, it must be presumed that they did not advert to the consideration, that granting to an individual a right to sue, was enlarging the jurisdiction of the Court. It never can be supposed, that they meant to assume the power of adding to the number of persons who might constitutionally become suitors in the Courts of the United States. But every difficulty vanishes, when we limit the meaning of the language of the act, by a reference to the context. In fact, a general power to bring actions in the Courts of the United States, is so peculiarly and explicitly personal on the face of the constitution, that it is hard to perceive how Congress could have for a moment lost sight of the restrictions [900] imposed, in this respect, upon the judicial power.

Nor had the Bank any idea that this power was vested in it, upon the ground that every possible case in which it might be involved in litigation, came within the constitutional definition of cases arising under laws, &c.; of the United States. In its averments, those on which it claims jurisdiction, it simply takes two grounds: 1. That it was incorporated by an act of Congress; 2. That the right to sue was given it by an act of Congress. But there is no averment, that the cause of action was a case arising under a law of the United States. It well knew, that it was a case emphatically arising out of an act of the State of Ohio, operating upon the domicil of the Bank, which, although purchased in right of an existence metaphysically given it by Congress, was acquired and held according to the laws of Ohio, acting upon its own territory. Technically, these averments cover only two grounds; they affirm, 1. That the Bank, being incorporated by Congress, had, therefore, a right to sue; 2. That being incorporated, and having the right to sue conferred upon it by an act of Congress, therefore, it could maintain this action. But yet neither, nor both of these, could give the right, unless in one of the cases defined in the constitution, which case is not the subject of an averment. I would not willingly place the case on the ground of mere technicality; and, therefore, only make the observation to show, that the ground assumed in argument, is an after-thought. I believe that, until this argument, the [901] ground now made was never thought of; and I am at a loss to conceive how it is possible to maintain the position, that all possible cases in which this Bank shall sue or be sued, come within the description now contended for. Take, for instance, a trespass or a fraud committed by the Bank, and suit brought by the injured party, in what sense could they be said to be cases arising under a law of the United States? Or, take the case of ejectment, suppose to recover part of the premises of the banking house in Philadelphia, and not a question raised in the suit, but what arises under the territorial laws of the country, and what circumstances characterize that as a case of the proper description to give this Court jurisdiction? If this cause of action arises under a statute, why is not the statute referred to, and the provision particularly relied on, if there is any other than what the averments specify?

Various instances have been cited and relied on, in which this right of suit in the Courts of the United States has been given to particular officers of the United States. But on these I would remark, that it is not logical to cite as proofs, the exercise of this right, in instances which may themselves be the subject of constitutional questions. It cannot be intended to surprise this Court into the recognition of the constitutionality of the laws so cited. But there is a stronger objection; no such instance is in point, until it be shown that Congress has authorized such officers to bring their private contracts and private controversies into the Courts of the United States. In all the [902] cases cited, the individual is acting distinctly as the organ of government; but let them take the character of a mere contractor, a factor, a broker, a common carrier, and then let laws authorizing them to sue in the Courts of the United States be passed, and I will acknowledge the cases to be in point; though I will still dispute the principle, that a repetition of error can convert an act into law or truth. The distinction is a clear one between all these cases and the Bank. The latter is a mere agent or attorney, in some instances; in others, and especially in the cases now before the Court, it is a private person, acting on its own account, not clothed with an official character at all. But the acts of public officers are the acts of government; and emphatically so, in suits by the Postmaster-General; the money to be recovered being the property of the United States, it may be considered that they are parties to the suit, just as those States are to the suits by or against their Attorney-General, where he is by law authorized to bring and defend suits in his own name officially. When the United States are parties, the grant of jurisdiction is general. But, there is express law also for every contract that the Postmaster enters into, or it will be in vain for him to bring his suit in his own name or otherwise. It would be in vain for him to rely simply on his being made Postmaster under an act of Congress; in which point alone, there would seem to exist any analogy between his case and that of the Bank.

As to the instance of the action given under the patent law, it has been before remarked, that so [903] entirely is its existence blended with an act of Congress, that to prosecute it, it is indispensable that the act should be set forth as the ground of action. I rather think it an unfortunate quotation, since it presents a happy illustration of what we are to understand by those cases arising under a law of Congress, which in their nature admit of an exercise of original jurisdiction. The plaintiff must recover, must count upon the act of Congress; the constitutional characteristic appears on the record before the defendant is called to answer; and the repeal of the statute before judgment, puts an end to his right altogether. Various such cases may be cited. But how the act of Congress is to be introduced into an action of trespass, ejectment, or slander, before the defendant is called to plead, I cannot imagine.

Upon the whole, I feel compelled to dissent from the Court, on the point of jurisdiction; and this renders it unnecessary for me to express my sentiments on the residue of the points in the cause.

Decree affirmed, except as to interest on the amount of the specie in the hands of the defendant, Sullivan.

[1] 1 Mad. 154, 155.

APPEAL FROM THE CIRCUIT COURT OF THE UNITED STATES FOR THE WESTERN DISTRICT OF KENTUCKY.

[151] Mr. Henry Lane Stone for appellant.

Mr. Lewis McQuown and Mr. Clarence U. McElroy for appellees.

By leave of court, Mr. L.A. Shaver, in behalf of The Interstate Commerce Commission, submitted a brief as amicus curioe.

MR. JUSTICE MOODY, after making the foregoing statement, delivered the opinion of the court.

Two questions of law were raised by the demurrer to the bill, were brought here by appeal, and have been argued before us. They are, first, whether that part of the act of Congress of June 29, 1906 (34 Stat. 584), which forbids the giving of free passes or the collection of any different compensation for transportation of passengers than that specified in the tariff filed, makes it unlawful to perform a contract for transportation of persons, who in good faith, before the passage of the act, had accepted such contract in satisfaction of a valid cause of action against the railroad; and, second, whether the statute, if it should be construed to render such a contract unlawful, is in [152] violation of the Fifth Amendment of the Constitution of the United States. We do not deem it necessary, however, to consider either of these questions, because, in our opinion, the court below was without jurisdiction of the cause. Neither party has questioned that jurisdiction, but it is the duty of this court to see to it that the jurisdiction of the Circuit Court, which is defined and limited by statute, is not exceeded. This duty we have frequently performed of our own motion. Mansfield, &c.; Railway Company v. Swan, 111 U.S. 379, 382; King Bridge Company v. Otoe County, 120 U.S. 225; Blacklock v. Small, 127 U.S. 96, 105; Cameron v. Hodges, 127 U.S. 322, 326; Metcalf v. Watertown, 128 U.S. 586, 587; Continental National Bank v. Buford, 191 U.S. 119.

There was no diversity of citizenship and it is not and cannot be suggested that there was any ground of jurisdiction, except that the case was a "suit . . . arising under the Constitution and laws of the United States." Act of August 13, 1888, c. 866, 25 Stat. 433, 434. It is the settled interpretation of these words, as used in this statute, conferring jurisdiction, that a suit arises under the Constitution and laws of the United States only when the plaintiff's statement of his own cause of action shows that it is based upon those laws or that Constitution. It is not enough that the plaintiff alleges some anticipated defense to his cause of action and asserts that the defense is invalidated by some provision of the Constitution of the United States. Although such allegations show that very likely, in the course of the litigation, a question under the Constitution would arise, they do not show that the suit, that is, the plaintiff's original cause of action, arises under the Constitution. In Tennessee v. Union & Planters' Bank, 152 U.S. 454, the plaintiff, the State of Tennessee, brought suit in the Circuit Court of the United States to recover from the defendant certain taxes alleged to be due under the laws of the State. The plaintiff alleged that the defendant claimed an immunity from the taxation by virtue of its charter, and that therefore the tax was void, because in violation of the provision of the Constitution of the United [153] States, which forbids any State from passing a law impairing the obligation of contracts. The cause was held to be beyond the jurisdiction of the Circuit Court, the court saying, by Mr. Justice Gray (p. 464), "a suggestion of one party, that the other will or may set up a claim under the Constitution or laws of the United States, does not make the suit one arising under that Constitution or those laws." Again, in Boston & Montana Consolidated Copper & Silver Mining Company v. Montana Ore Purchasing Company, 188 U.S. 632, the plaintiff brought suit in the Circuit Court of the United States for the conversion of copper ore and for an injunction against its continuance. The plaintiff then alleged, for the purpose of showing jurisdiction, in substance, that the defendant would set up in defense certain laws of the United States. The cause was held to be beyond the jurisdiction of the Circuit Court, the court saying, by Mr. Justice Peckham (pp. 638, 639).

"It would be wholly unnecessary and improper in order to prove complainant's cause of action to go into any matters of defence which the defendants might possibly set up and then attempt to reply to such defence, and thus, if possible, to show that a Federal question might or probably would arise in the course of the trial of the case. To allege such defence and then make an answer to it before the defendant has the opportunity to itself plead or prove its own defence is inconsistent with any known rule of pleading so far as we are aware, and is improper.

"The rule is a reasonable and just one that the complainant in the first instance shall be confined to a statement of its cause of action, leaving to the defendant to set up in his answer what his defence is and, if anything more than a denial of complainant's cause of action, imposing upon the defendant the burden of proving such defence.

"Conforming itself to that rule the complainant would not, in the assertion or proof of its cause of action, bring up a single Federal question. The presentation of its cause of action would not show that it was one arising under the Constitution or laws of the United States.

[154] "The only way in which it might be claimed that a Federal question was presented would be in the complainant's statement of what the defence of defendants would be and complainant's answer to such defence. Under these circumstances the case is brought within the rule laid down in Tennessee v. Union & Planters' Bank, 152 U.S. 454. That case has been cited and approved many times since, . . ."

The interpretation of the act which we have stated was first announced in Metcalf v. Watertown, 128 U.S. 586, and has since been repeated and applied in Colorado Central Consolidated Mining Company v. Turck, 150 U.S. 138, 142; Tennessee v. Union & Planters' Bank, 152 U.S. 454, 459; Chappell v. Waterworth, 155 U.S. 102, 107; Postal Telegraph Cable Company v. Alabama, 155 U.S. 482, 487; Oregon Short Line & Utah Northern Railway Company v. Skottowe, 162 U.S. 490, 494; Walker v. Collins, 167 U.S. 57, 59; Muse v. Arlington Hotel Company, 168 U.S. 430, 436; Galveston &c.; Railway v. Texas, 170 U.S. 226, 236; Third Street & Suburban Railway Company v. Lewis, 173 U.S. 457, 460; Florida Central & Peninsular Railroad Company v. Bell, 176 U.S. 321, 327; Houston & Texas Central Railroad Company v. Texas, 177 U.S. 66, 78; Arkansas v. Kansas & Texas Coal Company & San Francisco Railroad, 183 U.S. 185, 188; Vicksburg Waterworks Company v. Vicksburg, 185 U.S. 65, 68; Boston & Montana Consolidated Copper & Silver Mining Company v. Montana Ore Purchasing Company, 188 U.S. 632, 639; Minnesota v. Northern Securities Company, 194 U.S. 48, 63; Joy v. City of St. Louis, 201 U.S. 332, 340; Devine v. Los Angeles, 202 U.S. 313, 334. The application of this rule to the case at bar is decisive against the jurisdiction of the Circuit Court.

It is ordered that the

Judgment be reversed and the case remitted to the Circuit Court with instructions to dismiss the suit for want of jurisdiction.

CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE FEDERAL CIRCUIT

[828] Scalia, J., delivered the opinion of the Court, in which Rehnquist, C. J., and Kennedy, Souter, Thomas, and Breyer, JJ., joined, and in which Stevens, J., joined as to Parts I and II—A. Stevens, J., filed an opinion concurring in part and concurring in the judgment, post, p. 834. Ginsburg, J., filed an opinion concurring in the judgment, in which O'Connor, J., joined, post, p. 839.

James W. Dabney argued the cause for petitioner. With him on the brief were Paul Izzo, Timothy P. Gallogly, Arthur R. Miller, Marcia H. Sundeen, and Carol M. Wilhelm.

Peter W. Gowdey argued the cause for respondent. With him on the brief were Christopher P. Murphy, Janine A. Carlan, Kenneth W. Starr, and Daryl L. Joseffer.^[1]

Justice Scalia, delivered the opinion of the Court.

In this case, we address whether the Court of Appeals for the Federal Circuit has appellate jurisdiction over a case in which the complaint does not allege a claim arising under federal patent law, but the answer contains a patent-law counterclaim.

I

Respondent, Vornado Air Circulation Systems, Inc., is a manufacturer of patented fans and heaters. In late 1992, [828] respondent sued a competitor, Duracraft Corp., claiming that Duracraft's use of a "spiral grill design" in its fans infringed respondent's trade dress. The Court of Appeals for the Tenth Circuit found for Duracraft, holding that Vornado had no protectable trade-dress rights in the grill design. See Vornado Air Circulation Systems, Inc. v. Duracraft Corp., 58 F. 3d 1498 (1995) (Vornado I).

Nevertheless, on November 26, 1999, respondent lodged a complaint with the United States International Trade Commission against petitioner, The Holmes Group, Inc., claiming that petitioner's sale of fans and heaters with a spiral grill design infringed respondent's patent and the same trade dress held unprotectable in Vornado I. Several weeks later, petitioner filed this action against respondent in the United States District Court for the District of Kansas, seeking, inter alia, a declaratory judgment that its products did not infringe respondent's trade dress and an injunction restraining respondent from accusing it of trade-dress infringement in promotional materials. Respondent's answer asserted a compulsory counterclaim alleging patent infringement.

The District Court granted petitioner the declaratory judgment and injunction it sought. 93 F. Supp. 2d 1140 (Kan. 2000). The court explained that the collateralestoppel effect of Vornado I precluded respondent from relitigating its claim of trade-dress rights in the spiral grill design. It rejected respondent's contention that an intervening Federal Circuit case, Midwest Industries, Inc. v. Karavan Trailers, Inc., 175 F. 3d 1356 (1999), which disagreed with the Tenth Circuit's reasoning in Vornado I, constituted a change in the law of trade dress that warranted relitigation of respondent's trade-dress claim. The court also stayed all proceedings related to respondent's counterclaim, adding that the counterclaim would be dismissed if the declaratory judgment and injunction entered in favor of petitioner were affirmed on appeal.

[829] Respondent appealed to the Court of Appeals for the Federal Circuit. Notwithstanding petitioner's challenge to its jurisdiction, the Federal Circuit vacated the District Court's judgment, 13 Fed. Appx. 961 (2001), and remanded for consideration of whether the "change in the law" exception to collateral estoppel applied in light of TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U. S. 23 (2001), a case decided after the District Court's judgment which resolved a Circuit split involving Vornado I and Midwest Industries. We granted certiorari to consider whether the Federal Circuit properly asserted jurisdiction over the appeal. 534 U. S. 1016 (2001).

II

Congress vested the Federal Circuit with exclusive jurisdiction over "an appeal from a final decision of a district court of the United States . . . if the jurisdiction of that court was based, in whole or in part, on [28 U. S. C. § ] 1338 . . . ." 28 U. S. C. § 1295(a)(1) (emphasis added). Section 1338(a), in turn, provides in relevant part that "[t]he district courts shall have original jurisdiction of any civil action arising under any Act of Congress relating to patents . . . ." Thus, the Federal Circuit's jurisdiction is fixed with reference to that of the district court, and turns on whether the action arises under federal patent law.^[2]

Section 1338(a) uses the same operative language as 28 U. S. C. § 1331, the statute conferring general federalquestion jurisdiction, which gives the district courts "original jurisdiction of all civil actions arising under the Constitution, laws, or treaties of the United States." (Emphasis added.) We said in Christianson v. Colt Industries Operat- [830] ing Corp., 486 U. S. 800, 808 (1988), that "[l]inguistic consistency" requires us to apply the same test to determine whether a case arises under § 1338(a) as under § 1331.

The well-pleaded-complaint rule has long governed whether a case "arises under" federal law for purposes of § 1331.^[3] See, e. g., Phillips Petroleum Co. v. Texaco Inc., 415 U. S. 125, 127-128 (1974) (per curiam). As "appropriately adapted to § 1338(a)," the well-pleaded-complaint rule provides that whether a case "arises under" patent law "must be determined from what necessarily appears in the plaintiff's statement of his own claim in the bill or declaration . . . ." Christianson, 486 U. S., at 809 (internal quotation marks omitted). The plaintiff's well-pleaded complaint must "establis[h] either that federal patent law creates the cause of action or that the plaintiff's right to relief necessarily depends on resolution of a substantial question of federal patent law . . . ." Ibid. Here, it is undisputed that petitioner's well-pleaded complaint did not assert any claim arising under federal patent law. The Federal Circuit therefore erred in asserting jurisdiction over this appeal.

A

Respondent argues that the well-pleaded-complaint rule, properly understood, allows a counterclaim to serve as the basis for a district court's "arising under" jurisdiction. We disagree.

[831] Admittedly, our prior cases have only required us to address whether a federal defense, rather than a federal counterclaim, can establish "arising under" jurisdiction. Nevertheless, those cases were decided on the principle that federal jurisdiction generally exists "only when a federal question is presented on the face of the plaintiff's properly pleaded complaint." Caterpillar Inc. v. Williams, 482 U. S. 386, 392 (1987) (emphasis added). As we said in The Fair v. Kohler Die & Specialty Co., 228 U. S. 22, 25 (1913), whether a case arises under federal patent law "cannot depend upon the answer." Moreover, we have declined to adopt proposals that "the answer as well as the complaint . . . be consulted before a determination [is] made whether the case `ar[ises] under' federal law . . . ." Franchise Tax Bd. of Cal. v. Construction Laborers Vacation Trust for Southern Cal., 463 U. S. 1, 10-11, n. 9 (1983) (citing American Law Institute, Study of the Division of Jurisdiction Between State and Federal Courts § 1312, pp. 188-194 (1969)). It follows that a counterclaim—which appears as part of the defendant's answer, not as part of the plaintiff's complaint—cannot serve as the basis for "arising under" jurisdiction. See, e. g., In re Adams, 809 F. 2d 1187, 1188, n. 1 (CA5 1987); FDIC v. Elefant, 790 F. 2d 661, 667 (CA7 1986); Takeda v. Northwestern National Life Ins. Co., 765 F. 2d 815, 822 (CA9 1985); 14B C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure § 3722, pp. 402-414 (3d ed. 1998).

Allowing a counterclaim to establish "arising under" jurisdiction would also contravene the longstanding policies underlying our precedents. First, since the plaintiff is "the master of the complaint," the well-pleaded-complaint rule enables him, "by eschewing claims based on federal law, . . . to have the cause heard in state court." Caterpillar Inc., supra, at 398-399. The rule proposed by respondent, in contrast, would leave acceptance or rejection of a state forum to the master of the counterclaim. It would allow a [832] defendant to remove a case brought in state court under state law, thereby defeating a plaintiff's choice of forum, simply by raising a federal counterclaim. Second, conferring this power upon the defendant would radically expand the class of removable cases, contrary to the "[d]ue regard for the rightful independence of state governments" that our cases addressing removal require. See Shamrock Oil & Gas Corp. v. Sheets, 313 U. S. 100, 109 (1941) (internal quotation marks omitted). And finally, allowing responsive pleadings by the defendant to establish "arising under" jurisdiction would undermine the clarity and ease of administration of the well-pleaded-complaint doctrine, which serves as a "quick rule of thumb" for resolving jurisdictional conflicts. See Franchise Tax Bd., supra, at 11.

For these reasons, we decline to transform the longstanding well-pleaded-complaint rule into the "well-pleadedcomplaint-or-counterclaim rule" urged by respondent.

B

Respondent argues, in the alternative, that even if a counterclaim generally cannot establish the original "arising under" jurisdiction of a district court, we should interpret the phrase "arising under" differently in ascertaining the Federal Circuit's jurisdiction. In respondent's view, effectuating Congress's goal of "promoting the uniformity of patent law," Brief for Respondent 21, requires us to interpret §§ 1295(a)(1) and 1338(a) to confer exclusive appellate jurisdiction on the Federal Circuit whenever a patent-law counterclaim is raised.^[4]

[833] We do not think this option is available. Our task here is not to determine what would further Congress's goal of ensuring patent-law uniformity, but to determine what the words of the statute must fairly be understood to mean. It would be difficult enough to give "arising under" the meaning urged by respondent if that phrase appeared in § 1295(a)(1)—the jurisdiction-conferring statute—itself. Cf. Economic Stabilization Act of 1970, § 211(b)(2), 85 Stat. 749 (providing the Temporary Emergency Court of Appeals with exclusive jurisdiction over appeals "in cases and controversies arising under this title"). Even then the phrase would not be some neologism that might justify our adverting to the general purpose of the legislation, but rather a term familiar to all law students as invoking the wellpleaded-complaint rule. Cf. Coastal States Marketing, Inc. v. New England Petroleum Corp., 604 F. 2d 179, 183 (CA2 1979) ("The use of the phrase `cases and controversies arising under' . . . is strong evidence that Congress intended to borrow the body of decisional law that has developed under 28 U. S. C. § 1331 and other grants of jurisdiction to the district courts over cases `arising under' various regulatory statutes"). But the present case is even weaker than that, since § 1295(a)(1) does not itself use the term, but rather refers to jurisdiction under § 1338, where it is well established that "arising under any Act of Congress relating to patents" invokes, specifically, the well-pleaded-complaint rule. It would be an unprecedented feat of interpretive necromancy to say that § 1338(a)'s "arising under" language means one thing (the well-pleaded-complaint rule) in its own right, [834] but something quite different (respondent's complaint-orcounterclaim rule) when referred to by § 1295(a)(1).^[5]

* * *

Not all cases involving a patent-law claim fall within the Federal Circuit's jurisdiction. By limiting the Federal Circuit's jurisdiction to cases in which district courts would have jurisdiction under § 1338, Congress referred to a wellestablished body of law that requires courts to consider whether a patent-law claim appears on the face of the plaintiff's well-pleaded complaint. Because petitioner's complaint did not include any claim based on patent law, we vacate the judgment of the Federal Circuit and remand the case with instructions to transfer the case to the Court of Appeals for the Tenth Circuit. See 28 U. S. C. § 1631.

It is so ordered.

Justice Stevens, concurring in part and concurring in the judgment.

The Court correctly holds that the exclusive jurisdiction of the Court of Appeals for the Federal Circuit in patent [835] cases is "fixed with reference to that of the district court," ante, at 829. It is important to note the general rule, however, that the jurisdiction of the court of appeals is not "fixed" until the notice of appeal is filed. See Griggs v. Provident Consumer Discount Co., 459 U. S. 56, 58-59 (1982) (per curiam) ("The filing of a notice of appeal is an event of jurisdictional significance—it confers jurisdiction on the court of appeals and divests the district court of its control over those aspects of the case involved in the appeal").

Thus, if a case began as an antitrust case, but an amendment to the complaint added a patent claim that was pending or was decided when the appeal is taken, the jurisdiction of the district court would have been based "in part" on 28 U. S. C. § 1338(a), and therefore § 1295(a)(1) would grant the Federal Circuit jurisdiction over the appeal. Conversely, if the only patent count in a multicount complaint was voluntarily dismissed in advance of trial, it would seem equally clear that the appeal should be taken to the appropriate regional court of appeals rather than to the Federal Circuit. See Christianson v. Colt Industries Operating Corp., 486 U. S. 800, 823-824 (1988) (Stevens, J., concurring). Any other approach "would enable an unscrupulous plaintiff to manipulate appellate court jurisdiction by the timing of the amendments to its complaint." Id., at 824. To the extent that the Court's opinion might be read as endorsing a contrary result by reason of its reliance on cases involving the removal jurisdiction of the district court, I do not agree with it.

I also do not agree with the Court's statement that an interpretation of the "in whole or in part" language of § 1295(a)(1) to encompass patent claims alleged in a compulsory counterclaim providing an independent basis for the district court's jurisdiction would be a "neologism" that would involve "an unprecedented feat of interpretive necromancy," ante, at 833. For there is well-reasoned precedent supporting precisely that conclusion. See Aerojet-General Corp. v. [836] Machine Tool Works, Oerlikon-Buehrle Ltd., 895 F. 2d 736, 742-743 (CA Fed. 1990) (en banc) (opinion of Markey, C. J., for a unanimous court) (citing, e. g., Rengo Co. v. Molins Machine Co., 657 F. 2d 535, 539 (CA3 1981); Dale Electronics, Inc. v. R. C. L. Electronics, Inc., 488 F. 2d 382, 390 (CA1 1973); Pioche Mines Consol., Inc. v. Fidelity-Philadelphia Trust Co., 206 F. 2d 336, 336-337 (CA9 1953); Lion Mfg. Corp. v. Chicago Flexible Shaft Co., 106 F. 2d 930, 933 (CA7 1939)).^[6] I am nevertheless persuaded that a correct interpretation of § 1295(a)(1) limits the Federal Circuit's exclusive jurisdiction to those cases in which the patent claim is alleged in either the original complaint or an amended pleading filed by the [837] plaintiff. In my judgment, each of the three policies that the Court has identified as supporting the "well-pleadedcomplaint" rule governing district court jurisdiction, ante, at 831-832, points in the same direction with respect to appellate jurisdiction.

First, the interest in preserving the plaintiff's choice of forum includes not only the court that will conduct the trial but the appellate court as well. A plaintiff who has a legitimate interest in litigating in a circuit whose precedents support its theory of the case might omit a patent claim in order to avoid review in the Federal Circuit. In some cases that interest would be defeated by a rule that allowed a patent counterclaim to determine the appellate forum.

Second, although I doubt that a rule that enabled the counterclaimant to be the occasional master of the appellate forum "would radically expand" the number of cases heard by the Federal Circuit, ante, at 832, we must recognize that the exclusive jurisdiction of the Federal Circuit defined in § 1295(a)(1) does not comprise claims arising under the trademark and copyright laws, which are included in the district court's grant of jurisdiction under § 1338(a).^[7] As the instant litigation demonstrates, claims sounding in these other areas of intellectual property law are not infrequently bound up with patent counterclaims. The potential number of cases in which a counterclaim might direct to the Federal Circuit appeals that Congress specifically chose not to place within its exclusive jurisdiction is therefore significant.

Third, the interest in maintaining clarity and simplicity in rules governing appellate jurisdiction will be served by limiting [838] the number of pleadings that will mandate review in the Federal Circuit. In his opinion in Aerojet, Chief Judge Markey merely held that a counterclaim for patent infringement that was "compulsory" and not "frivolous" or "insubstantial" sufficed to establish jurisdiction; he made a point of noting that there was no assertion in the case that the patent counterclaim at issue had been filed "to manipulate the jurisdiction of [the Federal Circuit]." 895 F. 2d, at 738. The text of the statute, however, would not seem to distinguish between that counterclaim and those that are permissive, insubstantial, or manipulative, and there is very good reason not to make the choice of appellate forum turn on such distinctions. Requiring assessment of a defendant's motive in raising a patent counterclaim or the counterclaim's relative strength wastes judicial resources by inviting "unhappy interactions between jurisdiction and the merits." Kennedy v. Wright, 851 F. 2d 963, 968 (CA7 1988).

There is, of course, a countervailing interest in directing appeals in patent cases to the specialized court that was created, in part, to promote uniformity in the development of this area of the law. But we have already decided that the Federal Circuit does not have exclusive jurisdiction over all cases raising patent issues.^[8]Christianson, 486 U. S., at [839] 811-812. Necessarily, therefore, other circuits will have some role to play in the development of this area of the law. An occasional conflict in decisions may be useful in identifying questions that merit this Court's attention. Moreover, occasional decisions by courts with broader jurisdiction will provide an antidote to the risk that the specialized court may develop an institutional bias.^[9]

In sum, I concur in the Court's judgment and join Parts I and II—A of its opinion.

Justice Ginsburg, with whom Justice O'Connor joins, concurring in the judgment.

For reasons stated by Chief Judge Markey, writing for a unanimous en banc Federal Circuit in Aerojet-General Corp. v. Machine Tool Works, Oerlikon-Buehrle Ltd., 895 F. 2d 736 (1990), I conclude that, when the claim stated in a compulsory counterclaim "aris[es] under" federal patent law and is adjudicated on the merits by a federal district court, the Federal Circuit has exclusive appellate jurisdiction over that adjudication and other determinations made in the same case. See id., at 741-744 (distinguishing Christianson v. Colt Industries Operating Corp., 486 U. S. 800 (1988), in which this Court affirmed the jurisdictional decision of the Federal Circuit; in discussing the "well-pleaded complaint rule," the Federal Circuit observed that a patent infringement [840] counterclaim, unlike a patent issue raised only as a defense, has as its own, independent jurisdictional base 28 U. S. C. § 1338, i. e., such a claim discretely "arises under the patent laws").

The question now before this Court bears not at all on a plaintiff's choice of trial forum. The sole question presented here concerns Congress' allocation of adjudicatory authority among the federal courts of appeals. At that appellate level, Congress sought to eliminate forum shopping and to advance uniformity in the interpretation and application of federal patent law. See generally Dreyfuss, The Federal Circuit: A Case Study in Specialized Courts, 64 N. Y. U. L. Rev. 1, 30-37 (1989).

The Court's opinion dwells on district court authority. See ante, at 829-832. But, all agree, Congress left that authority entirely untouched. I would attend, instead, to the unique context at issue, and give effect to Congress' endeavor to grant the Federal Circuit exclusive appellate jurisdiction at least over district court adjudications of patent claims. See Dreyfuss, supra, at 36.

In the instant case, however, no patent claim was actually adjudicated. For that sole reason, I join the Court's judgment.

[1] David W. Long filed a brief for the Patent, Trademark, and Copyright Section of the Bar Association of the District of Columbia as amicus curiae.

[2] Like Christianson v. Colt Industries Operating Corp., 486 U. S. 800, 814-815 (1988), this case does not call upon us to decide whether the Federal Circuit's jurisdiction is fixed with reference to the complaint as initially filed or whether an actual or constructive amendment to the complaint raising a patent-law claim can provide the foundation for the Federal Circuit's jurisdiction.

[3]The well-pleaded-complaint rule also governs whether a case is removable from state to federal court pursuant to 28 U. S. C. § 1441(a), which provides in relevant part:

"Except as otherwise expressly provided by Act of Congress, any civil action brought in a State court of which the district courts of the United States have original jurisdiction, may be removed by the defendant or the defendants, to the district court of the United States for the district and division embracing the place where such action is pending."

See Franchise Tax Bd. of Cal. v. Construction Laborers Vacation Trust for Southern Cal., 463 U. S. 1 (1983).

[4] Echoing a variant of this argument, Justice Ginsburg contends that "giv[ing] effect" to Congress's intention "to eliminate forum shopping and to advance uniformity in . . . patent law" requires that the Federal Circuit have exclusive jurisdiction whenever a patent claim was "actually adjudicated." Post, at 840 (opinion concurring in judgment).We rejected precisely this argument in Christianson, viz., the suggestion that the Federal Circuit's jurisdiction is "fixed `by reference to the case actually litigated.' " 486 U. S., at 813 (quoting Brief for Respondent in Christianson v. Colt Industries Operating Corp., O. T. 1987, No. 87-499, p. 31). We held that the Federal Circuit's jurisdiction, like that of the district court, "is determined by reference to the well-pleaded complaint, not the well-tried case." 486 U. S., at 814.

[5] Although Justice Stevens agrees that a correct interpretation of § 1295(a)(1) does not allow a patent-law counterclaim to serve as the basis for the Federal Circuit's jurisdiction, he nevertheless quibbles that "there is well-reasoned precedent" supporting the contrary conclusion. See post, at 835 (opinion concurring in part and concurring in judgment). There is not. The cases relied upon by Justice Stevens and by the court in Aerojet-General Corp. v. Machine Tool Works, Oerlikon-Buehrle Ltd., 895 F. 2d 736 (CA Fed. 1990), simply address whether a district court can retain jurisdiction over a counterclaim if the complaint (or a claim therein) is dismissed or if a jurisdictional defect in the complaint is identified. They do not even mention the well-pleaded-complaint rule that the statutory phrase "arising under" invokes. Nor do any of these cases interpret § 1295(a)(1) or another statute conferring appellate jurisdiction with reference to the jurisdiction of the district court. Thus, the cases relied upon by Justice Stevens have no bearing on whether the phrase "arising under" can be interpreted differently in ascertaining the jurisdiction of the Federal Circuit than that of the district court.

[6] The Court dismisses the cases cited in Aerojet, a unanimous opinion for an en banc Federal Circuit, as having "no bearing" on this case because they do not parse the term "arising under" or interpret 28 U. S. C. § 1295(a)(1). Ante, at 834, n. 4. But surely it is not a "quibbl[e]" to acknowledge them as supporting the Aerojetcourt's conclusion that the jurisdiction of the district court can be based on a patent counterclaim, thereby satisfying the "in whole or in part" requirement of § 1295(a)(1).

In any event, the assertion that only the power of black magic could give "arising under" a different meaning with respect to appellate jurisdiction is belied by case law involving the Temporary Emergency Court of Appeals (TECA), which had exclusive jurisdiction over appeals in cases "arising under" the Economic Stabilization Act of 1970 (ESA), § 211(b)(2), 85 Stat. 749. Most courts departed from the traditional understanding of "arising under" and interpreted the statute to grant TECA appellate jurisdiction over ESA issues, including those raised as a defense. Courts nevertheless interpreted the statute's identical language respecting the district courts to grant traditional "arising under" jurisdiction. See Coastal States Marketing, Inc. v. New England Petroleum Corp., 604 F. 2d 179, 185-186 (CA2 1979) ("It must be candidly recognized that according the TECA some form of `issue' jurisdiction places on the phrase, `cases and controversies arising under,' . . . a construction that differs from the meaning associated with these words in other jurisdictional statutes, and differs even from the grant of jurisdiction to the district courts in [the ESA]"). Thus, although I am in agreement with the Court's ultimate decision not to determine appellate jurisdiction by reference to the defendant's patent counterclaim, I find it unnecessary and inappropriate to slight the contrary reasoning of the Court of Appeals.

[7] The statute grants the Federal Circuit "exclusive jurisdiction . . . if the jurisdiction of [the district] court was based, in whole or in part, on [28 U. S. C.] section 1338 .. . , except that a case involving a claim arising under any Act of Congress relating to copyrights, exclusive rights in mask works, or trademarks and no other claims under section 1338(a) shall be governed" by provisions relating to appeals to the regional courts of appeals. 28 U. S. C. § 1295(a)(1).

[8] In explicit contrast with the TECA, see n. 1, supra, the Federal Circuit was granted appellate jurisdiction over cases involving patent law claims, not issues. See Christianson v. Colt Industries Operating Corp., 486 U. S. 800,820-821, n. 1 (1988) (Stevens, J.,concurring) (quoting H. R. Rep. No. 97-312, p. 41 (1981)) ("Cases will be within the jurisdiction of the Court of Appeals for the Federal Circuit in the same sense that cases are said to `arise under' federal law for purposes of federal question jurisdiction. Contrast, Coastal States Marketing, Inc. v.New England Petroleum Corp., 604 F. 2d 179 (2d Cir., 1979) [Temporary Emergency Court of Appeals properly has jurisdiction over issues, not claims,arising under the Economic Stabilization Act]" (internal quotation marks omitted)).

Considerations of convenience to the parties and the courts support Congress' decision to determine the Federal Circuit's appellate jurisdiction based on the claims alleged in the well-pleaded complaint rather than the issues resolved by the district court's judgment. If, for example, the district court's judgment rests on multiple grounds, directing the appeal is a relatively straightforward matter by reference to the complaint. As Judge Easterbrook explains in Kennedy v.Wright, 851 F. 2d 963 (CA7 1988), fixing appellate jurisdiction with respect to the complaint also ensures that a case that has been appealed and remanded will return to the same appellate court if there is a subsequent appeal. Id., at 968 (describing the risk of "a game of jurisdictional ping-pong" if subsequent appeals are directed based on the grounds for decision rather than the pleadings).

[9] See Dreyfuss, The Federal Circuit: A Case Study in Specialized Courts, 64 N. Y. U. L. Rev. 1, 25-30, 54 (1989) (evaluating criticism that the Federal Circuit demonstrates a greater pro-patent bias than regional circuits).

CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE TENTH CIRCUIT.

[668] Charles L. Black argued the cause for petitioners. With him on the brief were W. P. Z. German, Alvin F. Molony, Hawley C. Kerr, Donald Campbell, Ray S. Fellows, Dan Moody, Walace Hawkins, Earl A. Brown and Raymond M. Myers.

Harry D. Turner argued the cause for respondent. With him on the brief were Don Emery, Rayburn L. Foster, George L. Sneed, S. E. Floren, Jr. and Eugene O. Monnett.

[669] MR. JUSTICE FRANKFURTER delivered the opinion of the Court.

In 1945, Michigan-Wisconsin Pipe Line Company sought from the Federal Power Commission a certificate of public convenience and necessity, required by § 7 (c) of the Natural Gas Act, 52 Stat. 825, as amended, 15 U. S. C. § 717f (c), for the construction and operation of a pipe line to carry natural gas from Texas to Michigan and Wisconsin. A prerequisite for such a certificate is adequate reserves of gas. To obtain these reserves Michigan-Wisconsin entered into an agreement with Phillips Petroleum Company on December 11, 1945, whereby the latter undertook to make available gas from the Hugoton Gas Field, sprawling over Kansas, Oklahoma and Texas, which it produced or purchased from others. Phillips had contracted with petitioners, Skelly Oil Company, Stanolind Oil and Gas Company, and Magnolia Petroleum Company, to purchase gas produced by them in the Hugoton Field for resale to Michigan-Wisconsin. Each contract provided that "in the event Michigan-Wisconsin Pipe Line Company shall fail to secure from the Federal Power Commission on or before [October 1, 1946] a certificate of public convenience and necessity for the construction and operation of its pipe line, Seller [a petitioner] shall have the right to terminate this contract by written notice to Buyer [Phillips] delivered to Buyer at any time after December 1, 1946, but before the issuance of such certificate." The legal significance of this provision is at the core of this litigation.

The Federal Power Commission, in response to the application of Michigan-Wisconsin, on November 30, 1946, ordered that "A certificate of public convenience and necessity be and it is hereby issued to applicant [Michigan-Wisconsin], upon the terms and conditions of this order," listing among the conditions that there be no transportation [670] or sale of natural gas by means of the sanctioned facilities until all necessary authorizations were obtained from the State of Wisconsin and the communities proposed to be served, that Michigan-Wisconsin should have the approval of the Securities and Exchange Commission for its plan of financing, that the applicant should file for the approval of the Commission a schedule of reasonable rates, and that the sanctioned facilities should not be used for the transportation of gas to Detroit and Ann Arbor except with due regard for the rights and duties of Panhandle Eastern Pipe Line Company, which had intervened before the Federal Power Commission, in its established service for resale in these areas, such rights and duties to be set forth in a supplemental order. It was also provided that Michigan-Wisconsin should have fifteen days from the issue of the supplemental order to notify the Commission whether the certificate "as herein issued is acceptable to it." Finally, the Commission's order provided that for purposes of computing the time within which applications for rehearing could be filed, "the date of issuance of this order shall be deemed to be the date of issuance of the opinions, or of the supplemental order referred to herein, whichever may be the later." 5 F. P. C. 953, 954, 956.

News of the Commission's action was released on November 30, 1946, but the actual content of the order was not made public until December 2, 1946. Petitioners severally, on December 2, 1946, gave notice to Phillips of termination of their contracts on the ground that Michigan-Wisconsin had not received a certificate of public convenience and necessity. Thereupon Michigan-Wisconsin and Phillips brought suit against petitioners in the District Court for the Northern District of Oklahoma. Alleging that a certificate of public convenience and necessity, "within the meaning of said Natural Gas Act and said contracts" had been issued prior to petitioners' attempt [671] at termination of the contracts, they invoked the Federal Declaratory Judgment Act for a declaration that the contracts were still "in effect and binding upon the parties thereto." Motions by petitioners to have Michigan-Wisconsin dropped as a party plaintiff were sustained, but motions to dismiss the complaint for want of jurisdiction were denied. The case then went to the merits, and the District Court decreed that the contracts between Phillips and petitioners had not been "effectively terminated and that each of such contracts remain [sic] in full force and effect." The Court of Appeals for the Tenth Circuit affirmed, 174 F. 2d 89, and we brought the case here, 338 U. S. 846, because it raises in sharp form the question whether a suit like this "arises under the Constitution, laws or treaties of the United States," 28 U. S. C. § 1331, so as to enable District Courts to give declaratory relief under the Declaratory Judgment Act. 48 Stat. 955, as amended, now 28 U. S. C. § 2201.

"[T]he operation of the Declaratory Judgment Act is procedural only." Aetna Life Ins. Co. v. Haworth, 300 U. S. 227, 240. Congress enlarged the range of remedies available in the federal courts but did not extend their jurisdiction. When concerned as we are with the power of the inferior federal courts to entertain litigation within the restricted area to which the Constitution and Acts of Congress confine them, "jurisdiction" means the kinds of issues which give right of entrance to federal courts. Jurisdiction in this sense was not altered by the Declaratory Judgment Act. Prior to that Act, a federal court would entertain a suit on a contract only if the plaintiff asked for an immediately enforceable remedy like money damages or an injunction, but such relief could only be given if the requisites of jurisdiction, in the sense of a federal right or diversity, provided foundation for resort to the federal courts. The Declaratory Judgment Act allowed relief to be given by way of recognizing the plaintiff's [672] right even though no immediate enforcement of it was asked. But the requirements of jurisdiction—the limited subject matters which alone Congress had authorized the District Courts to adjudicate—were not impliedly repealed or modified. See Great Lakes Dredge & Dock Co. v. Huffman, 319 U. S. 293, 300; Colegrove v. Green, 328 U. S. 549, 551-52.

If Phillips sought damages from petitioners or specific performance of their contracts, it could not bring suit in a United States District Court on the theory that it was asserting a federal right. And for the simple reason that such a suit would "arise" under the State law governing the contracts. Whatever federal claim Phillips may be able to urge would in any event be injected into the case only in anticipation of a defense to be asserted by petitioners. "Not every question of federal law emerging in a suit is proof that a federal law is the basis of the suit." Gully v. First National Bank, 299 U. S. 109, 115; compare 28 U. S. C. § 1257, with 28 U. S. C. § 1331. Ever since Metcalf v. Watertown, 128 U. S. 586, 589, it has been settled doctrine that where a suit is brought in the federal courts "upon the sole ground that the determination of the suit depends upon some question of a Federal nature, it must appear, at the outset, from the declaration or the bill of the party suing, that the suit is of that character." But "a suggestion of one party, that the other will or may set up a claim under the Constitution or laws of the United States, does not make the suit one arising under that Constitution or those laws." Tennessee v. Union & Planters' Bank, 152 U. S. 454, 464. The plaintiff's claim itself must present a federal question "unaided by anything alleged in anticipation of avoidance of defenses which it is thought the defendant may interpose." Taylor v. Anderson, 234 U. S. 74, 75-76; Louisville & Nashville R. Co. v. Mottley, 211 U. S. 149, 152.

[673] These decisions reflect the current of jurisdictional legislation since the Act of March 3, 1875, 18 Stat. 470, first entrusted to the lower federal courts wide jurisdiction in cases "arising under this Constitution, the Laws of the United States, and Treaties." U. S. Const. Art. III, § 2. "The change is in accordance with the general policy of these acts, manifest upon their face, and often recognized by this court, to contract the jurisdiction of the Circuit Courts [which became the District Courts] of the United States." Tennessee v. Union & Planters' Bank, supra at 462. See also Arkansas v. Kansas & Texas Coal Co., 183 U. S. 185, 188, and Gully v. First National Bank, supra at 112-14. With exceptions not now relevant Congress has narrowed the opportunities for entrance into the federal courts, and this Court has been more careful than in earlier days in enforcing these jurisdictional limitations. See Gully v. First National Bank, supra at 113.

To be observant of these restrictions is not to indulge in formalism or sterile technicality. It would turn into the federal courts a vast current of litigation indubitably arising under State law, in the sense that the right to be vindicated was State-created, if a suit for a declaration of rights could be brought into the federal courts merely because an anticipated defense derived from federal law. Not only would this unduly swell the volume of litigation in the District Courts but it would also embarrass those courts—and this Court on potential review—in that matters of local law may often be involved, and the District Courts may either have to decide doubtful questions of State law or hold cases pending disposition of such State issues by State courts. To sanction suits for declaratory relief as within the jurisdiction of the District Courts merely because, as in this case, artful pleading anticipates a defense based on federal law would contravene the whole trend of jurisdictional legislation [674] by Congress, disregard the effective functioning of the federal judicial system and distort the limited procedural purpose of the Declaratory Judgment Act. See Developments in the Law—Declaratory Judgments—1941-1949, 62 Harv. L. Rev. 787, 802-03 (1949). Since the matter in controversy as to which Phillips asked for a declaratory judgment is not one that "arises under the . . . laws . . . of the United States" and since as to Skelly and Stanolind jurisdiction cannot be sustained on the score of diversity of citizenship, the proceedings against them should have been dismissed.

As to Magnolia, a Texas corporation, a different situation is presented. Since Phillips was a Delaware corporation, there is diversity of citizenship. Magnolia had qualified to do business in Oklahoma and appointed an agent for service of process in accordance with the prevailing Oklahoma statute. Okla. Stat. Ann. tit. 18, § 452 (1937). Magnolia claimed that the subject matter of this proceeding did not arise in Oklahoma within the meaning of its consent to be sued. This contention was rejected below, and we do not reexamine the local law as applied by the lower courts. Under the doctrine of Neirbo Co. v. Bethlehem Shipbuilding Corp., 308 U. S. 165, venue was properly laid in Oklahoma; that the declaratory remedy which may be given by the federal courts may not be available in the State courts is immaterial.

Therefore, in the case of Magnolia we must reach the merits. They relate to two matters: (1) the clause in the contract with Phillips permitting its termination at any time after December 1, 1946, but before the "issuance" of "a certificate of public convenience and necessity" by the Federal Power Commission; and (2) whether this provision was satisfied by Magnolia's notice of termination of December 2, 1946, despite the Commission's order of November 30, 1946. The phraseology "certificate of [675] public convenience and necessity" in the contract is identic with the phrase in § 7 (c) of the Natural Gas Act. The Court of Appeals equated the term of the contract with that in the statute and in effect deemed its problem to be the proper construction of what constitutes the "issuance" of a "certificate of public convenience and necessity" within the meaning of § 7 (c). So viewing the matter, the court held that the order of November 30, 1946, satisfied the requirement of the contract, and that therefore a certificate of public convenience and necessity had been issued within the terminal period of the contract, and that its termination was not timely.

It will be recalled that the order of November 30, 1946, had three parts: (A) it stated that "A certificate of public convenience and necessity be and it is hereby issued to applicant [Michigan-Wisconsin]"; (B) it imposed certain conditions upon the grant, some of which were to be set forth in a supplemental order; and (C) it said that "For the purpose of computing the time within which applications for rehearing may be filed, the date of issuance of this order shall be deemed to be the date of issuance of the opinions, or of the supplemental order referred to herein, whichever may be the later." 5 F. P. C. at 954, 956. The course of reasoning by which the Court of Appeals concluded that the order of November 30, 1946, satisfied the statutory requirement for a certificate of public convenience and necessity can be briefly summarized. It relied on the grammatical argument that the Commission used the present tense in its order and subsequently referred to it as an order "issuing a certificate of public convenience and necessity," e. g., 6 F. P. C. 1, 37; the conditional nature of the order was not deemed to impair its efficacy since § 7 (e) of the Natural Gas Act authorized the Commission "to attach to the issuance of the certificate and to the exercise of the rights granted there-under such reasonable terms and conditions as the public [676] convenience and necessity may require"; and the provision of the order connecting the date of the order's issuance with the time defined for securing a rehearing was thought relevant only to the supplemental order.

We are not persuaded now to rest decision on the analysis of the Court of Appeals which led to its conclusion. We need not linger long on the merely grammatical argument of that court; it is given more weight than it can bear. Of course, the Commission has considerable administrative discretion to decide when an order may fairly be deemed to have been "issued." Section 16 of the Act provides that "Orders of the Commission shall be effective on the date and in the manner which the Commission shall prescribe." But surely a certificate cannot be said to have been issued for purposes of defining rights and the seeking of reconsideration by an aggrieved person if its substance is merely in the bosom of the Commission. Knowledge of the substance must to some extent be made manifest. Here the content of the order of November 30, 1946, was not made public until December 2, 1946, the date of the termination notice.

The Commission itself in its rule for computing rehearing time distinguishes between "adoption" of an order and its "issuance."^[1] However, as a matter of [677] usage, the Commission has referred to an order as having "issued" a certificate on a particular date when in fact the date was that of "adoption." See, e. g., Arkansas Louisiana Gas Co., 5 F. P. C. 813, 897; cf. Pacific Gas & Elec. Co., 5 F. P. C. 824, 901. Finally, the restriction of the Court of Appeals of the rehearing provision of Part C to the supplemental order finds no support on the face of the order of November 30, 1946. There is nothing to indicate that Part C was not to apply to the entire order for purposes of § 19 of the Act, which allows a rehearing by a party aggrieved "within thirty days after the issuance of such order" and makes such rehearing a prerequisite to judicial review. See 6 F. P. C. 323.

Since the requirements of the Natural Gas Act for the issuance of "a certificate of public convenience and necessity" may be distributive in scope, varying with the different contexts in which the question must be examined, this is not the occasion to decide that these requirements have a single uniform content. Whether the statutory requirement here was satisfied is not a question of fact, the finding of which by the District Court is to be respected unless clearly erroneous. The District Court merely found that the content of the piece of paper dated November 30, 1946, was that day agreed upon in executive session of the Commission and that that fact was made known. But this leaves untouched the legal significance of this action of the Commission, and the Court ought not now in darkness to pronounce on this question.

We are not restricted to disposition of the controversy on so truncated a treatment of the issues that underlie the record. Considering the fact that so to dispose of the case [678] would involve determination of an important problem concerning a regulatory statute with implications of public importance that private litigants naturally enough do not wholly represent and that on these matters neither the courts below nor this Court had the benefit of the experience and illumination of the agency entrusted with the enforcement of the Act,^[2] the due administration of justice requires that we should exercise our discretionary power in reviewing cases to "require such further proceedings to be had as may be just under the circumstances." 28 U. S. C. § 2106; Honeyman v. Hanan, 300 U. S. 14, 25. Accordingly, we think that the proper disposition requires that we vacate the judgment as to Magnolia and remand the case in order that the Court of Appeals either itself or by sending the case back to the District Court can further explore, through ways that may be appropriate, the issues which have been laid bare. See Kennedy v. Silas Mason Co., 334 U. S. 249.

The impact of the litigation both here and below was on the proper construction of § 7 (c). Even though the language of the contract may be identic with that of § 7 (c), this language in the contract may have a scope independent of the proper construction of § 7 (c). The same words, in different settings, may not mean the same thing. Compare opinion of Mr. Justice Holmes in Towne v. Eisner, 245 U. S. 418, with his dissent in Eisner v. Macomber, 252 U. S. 189, 219. Parties do not necessarily endow statutory language in a contract with the scope of the statute, particularly when the same term may have variant meanings for different applications of the statute. See Standard Oil Co. v. Johnson, 316 U. S. [679] 481, 483. Of course the statutory meaning in the context of the entire Natural Gas Act may not be irrelevant. In remanding the case we do not mean to foreclose this line of inquiry.^[3]

In respect to Magnolia, the judgment of the Court of Appeals is vacated and the cause remanded for further proceedings not inconsistent with this opinion. As to Skelly and Stanolind, we reverse the judgment with directions that the cause be dismissed.

It is so ordered.

MR. JUSTICE BLACK agrees with the Court of Appeals and would affirm its judgment.

MR. JUSTICE DOUGLAS took no part in the consideration or disposition of this case.

MR. CHIEF JUSTICE VINSON, with whom MR. JUSTICE BURTON joins, dissenting in part.

I concur in that part of the Court's judgment that directs dismissal of the cause as to Skelly and Stanolind. I have real doubts as to whether there is a federal question here at all, even though interpretation of the contract between private parties requires an interpretation of a federal statute and the action of a federal regulatory [680] body. But the Court finds it unnecessary to reach that question because it holds that the federal question, if any, is not a part of the plaintiff's claim and that jurisdiction does not, therefore, attach. While this result is not a necessary one, I am not prepared to dissent from it at this time.

But I am forced to dissent from the vacation and remand of the cause in respect to Magnolia. I think that, as to this petitioner, the judgment of the Court of Appeals should be affirmed. The Court decides that the Court of Appeals erred in holding that the Federal Power Commission had issued a certificate of public convenience and necessity to Michigan-Wisconsin Pipe Line Company on November 30, 1946, despite the fact that on that date the Commission adopted an order stating that "A certificate of public convenience and necessity be and it is hereby issued to Applicant, upon the terms and conditions of this order, . . . ." This disregard for what the District Court found to be the Commission's express intention is based upon two alternative grounds. First, it is suggested that while the order issuing the certificate was "adopted" on November 30, it was not "issued" until December 2. Second, it is said that Part C of the November 30 order, which concerned the date of issuance of the order for purposes of applications for rehearing, precludes a finding that a certificate was issued on November 30. Neither of these grounds, in my judgment, supports the Court's conclusion.

As to the first, which was not argued here nor in the Court of Appeals, it is true that the Commission's rules provide that an order is not to be deemed "issued" until the full text is mimeographed and mailed to the parties to the proceeding. This usually follows within two or three days after the order is "adopted." The only purpose of the postponement of the date of issuance of the [681] order, so far as we are informed, is to postpone the running of the 30-day period for applications for rehearing until the full text is available to the parties who have standing to ask for rehearing.

But the Commission uniformly refers to the date of adoption of the order as the date upon which the certificate of public convenience and necessity was "issued."^[4] It did so in this case, when, on March 12, 1947, it issued a supplemental order referring to its "order of November 30, 1946, issuing a certificate of public convenience and necessity." Furthermore, the District Court found as a fact that

"On November 30, a Saturday, the Commission in executive session made an order granting, with conditions, a certificate of public convenience and necessity to the Michigan-Wisconsin Pipe Line Company. During this session as the members of the Commission came to agreement as to the wording of the order, Mr. Fuquay, the secretary of the Commission, prepared the order in full and exact text. The secretary was directed by the Commission to release the order immediately."

Following adjournment on that day, the secretary sent a telegram to the parties to the proceeding, informing them that the "Commission today . . . adopted Opinion and Order, in Docket No. G-669, issuing certificate, with conditions, to Michigan Wisconsin Pipe Line Company." On the same day, releases to the press were made announcing the action taken by the Commission.

Skelly, Stanolind and Magnolia were not parties to this proceeding. It may very well be that the date of issuance of the order granting the certificate is December [682] 2 or some later date—for purposes of rehearing upon application of the parties. But I think there is no question that the certificate, as distinguished from the order, was issued on November 30. That is the Commission's view, as indicated by its supplemental order. The fact that it takes a few days to get its orders mimeographed and the Commission has adopted a rule that, in fairness to the parties, the time for rehearing shall not begin to run until such orders, in full text, are available, does not mean that the issuance of the certificate is also held in abeyance until that time.

The second argument requires but short answer. Part C provides that

"For the purpose of computing the time within which applications for rehearing may be filed, the date of issuance of this order shall be deemed to be the date of issuance of the opinions, or of the supplemental order referred to herein, whichever may be the later."

The paragraph means just what it says. I do not understand the Court to hold that the Commission cannot thus postpone the running of the time for rehearing. Computation of that time, as I have indicated, has no necessary relation to the date of issuance of the certificate.

I think that the Commission intended to and did issue a certificate of public convenience and necessity to Michigan-Wisconsin Pipe Line Company on November 30, 1946, whatever the date of its order, for purposes of computation of time for rehearing. The crucial clause of the contract refers to "the issuance of such certificate [of public convenience and necessity]." By their inclusion of a provision dependent upon the action of a federal agency, it is obvious that the parties intended that the contract should be construed with reference to the effective [683] date of agency action under the statutes and the practices of the Commission. The District Court so concluded.^[5] I can see no reason, therefore, to remand the cause for further proceedings. In my view, effective agency action was taken on November 30, 1946. As to Magnolia, I would affirm the judgment of the Court of Appeals.

[1]Rule 13 (b) of the Commission's Rules of Practice and Procedure provides: "In computing any period of time involving the date of the issuance of an order by the Commission, the day of issuance of an order shall be the day the Office of the Secretary mails or delivers copies of the order (full text) to the parties or their attorneys of record, or makes such copies public, whichever be the earlier. . . . The day of issuance of an order may or may not be the day of its adoption by the Commission." 18 C. F. R. § 1.13 (b). A deposition taken of the Secretary of the Commission gave light on this point. The Commission's previous rule on rehearing time is in 18 C. F. R. Cum. Supp. § 50.75.

Rule 13 (c) provides: "Orders of the Commission shall be effective as of the dates of issuance unless otherwise specifically provided in the orders." 18 C. F. R. § 1.13 (c). This provision may be of significance if the effectiveness of a certificate is an issue in proceedings under § 20 or § 21 of the Act. The Court of Appeals did not discuss the bearing of these rules upon this case.

[2] The significance of the conditions in qualifying what is formally called a "certificate" in the order of November 30, 1946, is precisely one of those matters upon which Commission practice and experience may shed helpful light.

[3]In its conclusions of law, the District Court stated: "The certificate issued by the Commission to Michigan-Wisconsin on November 30, 1946, although containing terms and conditions, was and is a certificate issued under the requirements of the Natural Gas Act and one that is provided for by that act. A consideration of the contracts between plaintiff and defendants, together with the contract between plaintiff and Michigan-Wisconsin, compels a conclusion that such certificate was one within the contemplation of the parties and satisfied the terms of the contracts."

The context suggests that in the second sentence the District Court may still have been focusing upon statutory meaning.

[4] See, e. g., Arkansas Louisiana Gas Co., 5 F. P. C. 813, 897; Pacific Gas & Elec. Co., 5 F. P. C. 824, 901.

[5] The District Court stated as one of its conclusions of law: "The certificate issued by the Commission to Michigan-Wisconsin on November 30, 1946, although containing terms and conditions, was and is a certificate issued under the requirements of the Natural Gas Act and one that is provided for by that act. A consideration of the contracts between plaintiff and defendants, together with the contract between plaintiff and Michigan-Wisconsin, compels a conclusion that such certificate was one within the contemplation of the parties and satisfied the terms of the contracts."

[767] Willie D. Bright, pro per.

Philip R. Placier, Charles P. McCarthy, Thelen, Marrin, Johnson & Bridges, San Francisco, Cal., for defendants-appellees.

[768] Before WRIGHT, KENNEDY and BEEZER, Circuit Judges.

BEEZER, Circuit Judge:

Plaintiff brought an action against his employer in a California state court. He alleged that his employer had breached his employment contract by paying him less than the contract required. The employer removed the case to federal district court on the ground that plaintiff had artfully pleaded what was in reality a challenge to the employer's compliance with federal law requiring the withholding of federal income tax. The district court denied plaintiff's motion to remand, dismissed the action, and awarded attorney's fees to the employer. We affirm and impose sanctions.

BACKGROUND

On November 7, 1983, Willie D. Bright entered into an employment contract with Bechtel Petroleum, Inc. pursuant to a collective bargaining agreement. On January 3, 1984, Bright provided to Bechtel a federal Employee's Withholding Allowance Certificate (W-4 Form) on which he claimed an exemption from the withholding of federal income tax.

In accordance with federal internal revenue regulations, Bechtel forwarded Bright's W-4 Form to the Internal Revenue Service (IRS). On March 14, 1984, the IRS issued a directive to Bechtel declaring Bright's W-4 Form invalid and ordering Bechtel to begin withholding federal income tax. Pursuant to a California state income tax regulation, Bechtel also began withholding state income tax. In 1984, a total of $2,486.18 in federal tax and $736.00 in state tax was withheld from Bright's wages.

On June 1, 1984, Bright filed an action against Bechtel, and certain individual agents, in California Superior Court alleging a breach of contract. Bechtel removed the case to the United States District Court for the Northern District of California on the ground that the true basis of the complaint was that Bechtel had complied with federal income tax laws in withholding tax from Bright's wages. Bechtel also moved for dismissal of the action, and for an award of costs and attorney's fees on the ground the complaint was frivolous and filed in bad faith.

Bright moved to remand the case back to state court, arguing that he desired to litigate only the question of Bechtel's withholding of state income tax from his wages.

On December 4, 1984, the district court denied Bright's motion to remand, granted the motion to dismiss, and awarded attorney's fees in the amount of $3300 to Bechtel.

On December 19, 1984, Bright filed a motion asking the court for "clarification" of the grounds of its judgment. Before the district court had ruled on this motion, Bright filed a notice of appeal from the judgment. The court subsequently denied Bright's motion. Bright also filed a notice of appeal from that order. The two appeals have been consolidated.

ANALYSIS

A. Standard of Review

A question of federal subject matter jurisdiction, such as that permitting removal of a case from state to federal court, is reviewable de novo. See Mobil Oil Corp. v. City of Long Beach, 772 F.2d 534, 538 (9th Cir.1985). This court reviews de novo a district court's ruling on a motion to dismiss for failure to state a claim upon which relief can be granted. Trerice v. Pedersen, 769 F.2d 1398, 1400 (9th Cir.1985).

B. Removal of the Case — Federal Question Jurisdiction

Bright's complaint alleges that Bechtel breached the employment contract by "issuing [769] to plaintiff paychecks that were in amounts less than contracted and agreed for." The complaint alleges a cause of action based upon "breach of contract and a gross violation of Christian principles."

Bechtel petitioned to remove the case to federal district court on the ground that the "breach" Bright alleges is based upon Bechtel's withholding of income tax pursuant to federal law. Bechtel contends the action falls within the district court's original jurisdiction because it "arises under" federal internal revenue statutes.^[1]

An action may "arise under" a law of the United States if the plaintiff's right to relief necessarily turns on construction of federal law. Franchise Tax Board v. Construction Laborers Vacation Trust, 463 U.S. 1, 9, 103 S.Ct. 2841, 2846, 77 L.Ed.2d 420 (1983); Mobil Oil Corp. v. City of Long Beach, 772 F.2d 534, 539 (9th Cir.1985). Under the "well-pleaded complaint" rule, the federal question, which invokes federal jurisdiction, must appear from the complaint and not from any federal defense the defendant might raise to defeat the claim. Franchise Tax Board, 463 U.S. at 10, 103 S.Ct. at 2846; Mobil Oil Corp., 772 F.2d at 538-39.

Although the plaintiff is generally considered the "master of his complaint" and is free to choose the forum for his action, this principle is not without limitation. Schroeder v. Trans World Airlines, Inc., 702 F.2d 189, 190 (9th Cir.1983).^[2] A plaintiff will not be allowed to conceal the true nature of a complaint through "artful pleading." Id. at 191; see also Olguin v. Inspiration Consolidated Copper Co., 740 F.2d 1468, 1472 (9th Cir.1984).

Bechtel, through background information in its petition for removal and supporting affidavits,^[3] demonstrated that Bright, despite "artfully pleading" his action as a breach of contract, in fact is challenging federal income tax withholding laws and regulations. Bright had written two letters threatening a lawsuit for breach of contract against Bechtel for complying with the IRS directive to withhold income tax from his wages. Both letters criticized the IRS directive as illegal and cited federal internal revenue statutes. The second letter expressly referred to a decision in this circuit, Stonecipher v. Bray, 653 F.2d 398 (9th Cir.1981), cert. denied, 454 U.S. 1145, 102 S.Ct. 1006, 71 L.Ed.2d 297 (1982), which dismissed a suit, similar to the instant action, as frivolous.

In addition, affidavits were presented to the district court showing that Bright's complaint contains language virtually identical to that in the complaint of another employee who had filed an action against Bechtel in state court.^[4] That action had been removed to the Northern District on the same ground, that the alleged breach of contract derived from the company's withholding of federal income tax from a [770] paycheck, and had been dismissed as frivolous shortly before Bright filed the complaint in this action.

Once the removal petition had been filed, Bright asserted for the first time that he intended only to litigate a claim involving the withholding of California state income tax. Thus, Bright argues, no federal question is raised. Bright's assertion must be regarded as disingenuous. The complaint simply does not state a claim limited to a state tax issue. It is clear under the circumstances that Bright has raised this new argument as another attempt to evade the jurisdiction of the Northern District, which has been acting swiftly to dismiss similar suits as frivolous.

Accordingly, the removal of the action to federal court was proper. See Chandler v. Perini Power Constructors, Inc., 520 F.Supp. 1152, 1155 (D.N.H.1981).

C. Dismissal of Federal Tax Withholding Claim

Under 26 U.S.C. § 3402, an employer has a mandatory duty to withhold federal income tax from an employee's wages where required by applicable regulations. See Maxfield v. United States Postal Service, 752 F.2d 433, 434 (9th Cir.1984); Chandler v. Perini Power Constructors, Inc., 520 F.Supp. 1152, 1153 (D.N.H.1981). The actions taken by Bechtel, in withholding federal income tax from Bright's wages, were fully in accordance with federal internal revenue regulations.^[5] This circuit has repeatedly held that an employer is not liable to an employee for complying with its legal duty to withhold tax. Maxfield, 752 F.2d at 434; see Stonecipher v. Bray, 653 F.2d 398, 403 (9th Cir.1981) (employer has not breached employment contract by withholding taxes from wages and paying employee the balance), cert. denied, 454 U.S. 1145, 102 S.Ct. 1006, 71 L.Ed.2d 297 (1982); Callow v. Amerace Corp., 681 F.2d 1242, 1243 (9th Cir.1982) (affirming dismissal of action against employer seeking to recover withheld federal and state income tax).

The Internal Revenue Code, 26 U.S.C. § 3403, expressly provides that an employer is liable to the IRS for the payment of tax withheld, and "shall not be liable to any person for the amount of any such payment." Thus, suits by employees against employers for tax withheld are "statutorily barred." Chandler, 520 F.Supp. at 1156.

In addition, we have held that actions, such as this, are barred by the Anti-Injunction Act, 26 U.S.C. § 7421(a), which prohibits suits "for the purpose of restraining the assessment or collection of any tax." Maxfield, 752 F.2d at 434. A lawsuit, such as in the instant case, can be viewed as one to restrain collection (through withholding) of federal income tax. See id.; Chandler, 520 F.Supp. at 1155.

The district court properly dismissed Bright's claim involving withholding of federal income tax.

D. Dismissal of State Tax Withholding Claim

Bright's complaint also raises a breach of contract claim against Bechtel based on the withholding of state income tax. Bechtel withheld state income tax in compliance with a California state regulation, 22 Cal.Admin.Code § 4340-1(b), providing that a federal determination that a withholding exemption claim is invalid is also effective for state withholding purposes.

As the district court failed to remand any part of the case to state court, and granted a final dismissal of the case, we must assume [771] the district court also dismissed the claim involving withholding of state tax.

The district court had proper jurisdiction over the claim for breach of contract involving state tax withholding. Once the federal court acquires jurisdiction of a case on removal, it also acquires jurisdiction over pendent state law claims. Salveson v. Western States BankCard Ass'n, 731 F.2d 1423, 1430 (9th Cir.1984). Dismissal of the federal claim does not deprive a federal court of the power to adjudicate the remaining pendent state claims. Rosado v. Wyman, 397 U.S. 397, 403-04, 90 S.Ct. 1207, 1213, 25 L.Ed.2d 442 (1970); Anderson v. Allstate Ins. Co., 630 F.2d 677, 681 (9th Cir.1980).

The claims involving withholding of state tax and of federal tax arose out of a "common nucleus of operative fact." See United Mine Workers of America v. Gibbs, 383 U.S. 715, 725, 86 S.Ct. 1130, 1138, 16 L.Ed.2d 218 (1966). Bechtel's withholding of both taxes was based upon the single IRS directive. The California regulation makes the IRS determination effective for state purposes as well. Although the funds withheld were allocated between the IRS and the California Franchise Tax Board, the act of withholding was a single transaction.

Furthermore, the same issue, concerning whether an employer can be held liable in breach of contract for complying with government regulations, is raised in the claim involving state tax withholding. To conserve judicial resources, it was appropriate for the district court to decide this issue. It was not an abuse of discretion to refuse to remand this claim to state court.

Bright argues that it is inappropriate for a federal court to consider state tax issues. It is true that the principle of comity generally bars challenges to the validity of state tax systems in federal court.^[6]See Miller v. City of Los Angeles, 755 F.2d 1390, 1391 (9th Cir.), cert. denied, 474 U.S. 995, 106 S.Ct. 408, 88 L.Ed.2d 359 (1985). However, the California state income taxation system is not at issue in this action. Bright does contend that the regulation requiring an employer to withhold state tax if the IRS determines an exemption claim is invalid, 22 Cal.Admin.Code § 4340-1(b), violates the California constitution and state statutes which purportedly prohibit such a delegation of state taxation authority to the federal government. This is simply not relevant. Even if the regulation is invalid, it was justifiable for Bechtel to rely upon it and comply with its command. The only issue raised in this action is whether an employer can be held liable under California law for obeying a California regulation.

One California court has recently considered a claim similar to that brought by Bright in the instant case. In Otworth v. Southern Pacific Transp. Co., 166 Cal.App.3d 452, 167 Cal.App.3d 102E, 212 Cal.Rptr. 743, 744-45 (1985), the plaintiff alleged conversion, breach of contract, violations of constitutional rights, and unjust enrichment against his employer for withholding federal and state income tax. The funds had been turned over to the IRS and California Franchise Tax Board. Id., 212 Cal.Rptr. at 745. The court affirmed dismissal of these claims and imposed sanctions for frivolous appeal. Id., 212 Cal.Rptr. at 748-49.

The court appeared to view the state tax withholding issue as part and parcel of the federal tax withholding dispute. Consequently, the court's conclusion that the plaintiff failed to state a claim for breach of contract would apply to both the withholding of federal and state tax. The court held that, in the absence of a provision requiring the employer to refrain from withholding, an employer "discharges its [772] contractual obligations when it withholds taxes from the employee's wages and pays the employee the balance." Id., 212 Cal.Rptr. at 747.

Even if a technical breach of contract were established, Bright could not prevail in this action. Under Cal.Civ.Code § 1511(1), no liability exists for breach of a contract whose performance has been made impossible by operation of law. Baird v. Wendt Enterprises, Inc., 248 Cal.App.2d 52, 56 Cal.Rptr. 118, 120 (1967); see generally 1 Witkin, Summary of California Law, "Contracts," § 607, at 517 (8th ed. 1973). A party is not required to violate the law to avoid liability for breach of contract.^[7]See Northrop Corp. v. Triad Financial Establishment, 593 F.Supp. 928, 937 (C.D.Cal.1984) (interpreting California law).

The claim involving withholding of state tax was properly dismissed.

E. Award of Attorney's Fees

The district court did not abuse its discretion in awarding attorney's fees. See Callow v. Amerace Corp., 681 F.2d 1242, 1243 (9th Cir.1982). The action is plainly frivolous, brought in bad faith, and for the purposes of harassment.^[8]

The district court was made aware of another recent judgment in the Northern District, dismissing a nearly identically worded complaint by another Bechtel employee, which found that there was an organized campaign of "tax protest" lawsuits, aimed at "thwarting federal tax regulations by inundating employers with frivolous suits in the hope that the volume and cost of litigation involved in defending these actions will deter employers from withholding taxes." McFarland v. Bechtel, 586 F.Supp. 907 (N.D.Cal.1984).

Under these circumstances, the award of attorney's fees was "an appropriate deterrent to future frivolous suits." Callow, 681 F.2d at 1243.

F. Denial of Motion for "Clarification"

The district court's final judgment was entered on December 4, 1984. On December 19, 1984, Bright filed a motion for "clarification" of the judgment. Even if this court liberally construes Bright's motion as a motion under Federal Rule of Civil Procedure 52(b) to amend the findings or under Rule 59(e) to alter or amend the judgment, the motion was not made within 10 days after entry of the judgment as required by the rules.

Even had the motion been timely, it was properly denied. Rule 52(a) provides that findings of fact and conclusions of law are unnecessary in decisions on motions.

G. Attorney's Fees and Costs on Appeal

"When a judgment is affirmed, this court has authority to award fees and double costs on appeal in its discretion pursuant to 28 U.S.C. § 1912. Federal Rule of Appellate Procedure 38 provides for the same remedy if the court finds an appeal is frivolous." Maxfield v. United States Postal Service, 752 F.2d 433, 434-35 (9th Cir.1984) This court has imposed such sanctions previously in cases involving similar frivolous claims against employers for withholding income tax. Maxfield, 752 F.2d at 435; Stonecipher v. Bray, 653 F.2d 398, 403 (9th Cir.1981), cert. denied, 454 U.S. 1145, 102 S.Ct. 1006, 71 L.Ed.2d 297 (1982).

An award of attorney's fees is particularly warranted in this case as Bechtel [773] should not be forced to bear the expense of this action and appeal simply because it properly complied with federal and state law. Bechtel is an innocent bystander caught in the crossfire of Bright's war against the income taxation system. Bechtel should therefore be awarded reasonable attorney's fees and double costs on appeal. See Maxfield, 752 F.2d at 435.

In this case, Bechtel shall have 15 days following the filing of this decision within which to submit a declaration for attorneys' fees calculated in accordance with the factors set forth in Kerr v. Screen Extras Guild, Inc., 526 F.2d 67, 70 (9th Cir.1975), cert. denied, 425 U.S. 951, 96 S.Ct. 1726, 48 L.Ed.2d 195 (1976).

AFFIRMED.

[1] Bechtel argues that removal of the case was proper under 28 U.S.C. § 1441(a) & (b), as the action was within the original jurisdiction of the district court pursuant to 28 U.S.C. § 1331 (original jurisdiction of all civil actions "arising under the Constitution, laws, or treaties of the United States") and 28 U.S.C. § 1340 (original jurisdiction of any civil action arising under federal statutes providing for internal revenue). See Chandler v. Perini Power Constructors, Inc., 520 F.Supp. 1152, 1154 n. 3 (D.N.H.1981) (upholding removal pursuant to these statutes of a case alleging breach of an employment agreement, on the ground that the suit actually challenged employer's withholding of federal income tax).

[2] In Schroeder, the plaintiffs attempted to frame their action as a breach of agreement under state law. This court looked through to the substance of the action and found it actually involved a labor dispute governed by federal law. Id. at 190-91. As in the instant case, the plaintiffs had attempted to avoid application of federal law by relying on state law to articulate their claims.

[3] It is proper to use the petition for removal to clarify the action presented in the complaint in determining whether it raises a federal question. Schroeder, 702 F.2d at 191; Olguin, 740 F.2d at 1473.

[4] For example, both complaints framed the cause of action as "breach of contract and a gross violation of Christian principles." This is hardly coincidental.

[5] Under 26 C.F.R. § 31.3402(f)(2)-1(g)(1) & (2), an employer is required to forward an employee's W-4 Form to the IRS, if the employee claims an exemption from withholding of tax while earning more than $200 per week. Under 26 C.F.R. § 31.3402(f)(2)-1(g)(5)(v), if the IRS issues a directive instructing an employer to disregard an employee's withholding exemption W-4 Form as defective, the employer must withhold amounts from the employee's wages in accordance with the IRS notice.

[6] In addition, the Tax Injunction Act, 28 U.S.C. § 1341, deprives federal district courts of subject matter jurisdiction over actions for injunctive relief from state taxation schemes. Marvin F. Poer & Co. v. Counties of Alameda, 725 F.2d 1234, 1236 (9th Cir.1984). This court also refuses to recognize jurisdiction over actions for refunds of state taxes. Id.

[7] An administrative regulation, of course, has the force of law. And government regulations cannot be varied or evaded by private contract. Alpha Beta Food Markets, Inc. v. Retail Clerks Union Local 770, 45 Cal.2d 764, 291 P.2d 433, cert. denied, 350 U.S. 996, 76 S.Ct. 547, 100 L.Ed. 861 (1956).

[8] As a further example of the frivolous and bad faith nature of this action, Bright's complaint seeks more than one hundred million dollars in damages from Bechtel for withholding approximately $3222 in taxes.

CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT

[805] Frank C. Woodside III argued the cause for petitioner. With him on the briefs was Christine L. McBroom.

Stanley M. Chesley argued the cause and filed a brief for respondents.

JUSTICE STEVENS delivered the opinion of the Court.

The question presented is whether the incorporation of a federal standard in a state-law private action, when Congress has intended that there not be a federal private action for violations of that federal standard, makes the action one "arising under the Constitution, laws, or treaties of the United States," 28 U. S. C. § 1331.

I

The Thompson respondents are residents of Canada and the MacTavishes reside in Scotland. They filed virtually identical complaints against petitioner, a corporation, that manufactures and distributes the drug Bendectin. The complaints were filed in the Court of Common Pleas in Hamilton County, Ohio. Each complaint alleged that a child was born with multiple deformities as a result of the mother's ingestion of Bendectin during pregnancy. In five of the six counts, the recovery of substantial damages was requested on common-law theories of negligence, breach of warranty, strict liability, fraud, and gross negligence. In Count IV, respondents alleged that the drug Bendectin was "misbranded" in violation of the Federal Food, Drug, and Cosmetic Act (FDCA), 52 Stat. 1040, as amended, 21 U. S. C. § 301 et seq. (1982 ed. and Supp. III), because its labeling did not provide adequate [806] warning that its use was potentially dangerous. Paragraph 26 alleged that the violation of the FDCA "in the promotion" of Bendectin "constitutes a rebuttable presumption of negligence." Paragraph 27 alleged that the "violation of said federal statutes directly and proximately caused the injuries suffered" by the two infants. App. 22, 32.

Petitioner filed a timely petition for removal from the state court to the Federal District Court alleging that the action was "founded, in part, on an alleged claim arising under the laws of the United States."^[1] After removal, the two cases were consolidated. Respondents filed a motion to remand to the state forum on the ground that the federal court lacked subject-matter jurisdiction. Relying on our decision in Smith v. Kansas City Title & Trust Co., 255 U. S. 180 (1921), the District Court held that Count IV of the complaint alleged a cause of action arising under federal law and denied the motion to remand. It then granted petitioner's motion to dismiss on forum non conveniens grounds.

The Court of Appeals for the Sixth Circuit reversed. 766 F. 2d 1005 (1985). After quoting one sentence from the concluding paragraph in our recent opinion in Franchise Tax Board v. Construction Laborers Vacation Trust, 463 U. S. 1 (1983),^[2] and noting "that the FDCA does not create or imply [807] a private right of action for individuals injured as a result of violations of the Act," it explained:

"Federal question jurisdiction would, thus, exist only if plaintiffs' right to relief depended necessarily on a substantial question of federal law. Plaintiffs' causes of action referred to the FDCA merely as one available criterion for determining whether Merrell Dow was negligent. Because the jury could find negligence on the part of Merrell Dow without finding a violation of the FDCA, the plaintiffs' causes of action did not depend necessarily upon a question of federal law. Consequently, the causes of action did not arise under federal law and, therefore, were improperly removed to federal court." 766 F. 2d, at 1006.

We granted certiorari, 474 U. S. 1004 (1985), and we now affirm.

II

Article III of the Constitution gives the federal courts power to hear cases "arising under" federal statutes.^[3] That grant of power, however, is not self-executing, and it was not until the Judiciary Act of 1875 that Congress gave the federal courts general federal-question jurisdiction.^[4] Although the constitutional meaning of "arising under" may extend to all cases in which a federal question is "an ingredient" of the action, Osborn v. Bank of the United States, 9 Wheat. 738, 823 (1824), we have long construed the statutory grant of federal-question jurisdiction as conferring a more limited power. [808] Verlinden B. V. v. Central Bank of Nigeria, 461 U. S. 480, 494-495 (1983); Romero v. International Terminal Operating Co., 358 U. S. 354, 379 (1959).

Under our longstanding interpretation of the current statutory scheme, the question whether a claim "arises under" federal law must be determined by reference to the "well-pleaded complaint." Franchise Tax Board, 463 U. S., at 9-10. A defense that raises a federal question is inadequate to confer federal jurisdiction. Louisville & Nashville R. Co. v. Mottley, 211 U. S. 149 (1908). Since a defendant may remove a case only if the claim could have been brought in federal court, 28 U. S. C. § 1441(b), moreover, the question for removal jurisdiction must also be determined by reference to the "well-pleaded complaint."

As was true in Franchise Tax Board, supra, the propriety of the removal in this case thus turns on whether the case falls within the original "federal question" jurisdiction of the federal courts. There is no "single, precise definition" of that concept; rather, "the phrase `arising under' masks a welter of issues regarding the interrelation of federal and state authority and the proper management of the federal judicial system." Id., at 8.

This much, however, is clear. The "vast majority" of cases that come within this grant of jurisdiction are covered by Justice Holmes' statement that a " `suit arises under the law that creates the cause of action.' " Id., at 8-9, quoting American Well Works Co. v. Layne & Bowler Co., 241 U. S. 257, 260 (1916). Thus, the vast majority of cases brought under the general federal-question jurisdiction of the federal courts are those in which federal law creates the cause of action.

We have, however, also noted that a case may arise under federal law "where the vindication of a right under state law necessarily turned on some construction of federal law." [809] Franchise Tax Board, 463 U. S., at 9.^[5] Our actual holding in Franchise Tax Board demonstrates that this statement must be read with caution; the central issue presented in that case turned on the meaning of the Employee Retirement Income Security Act of 1974, 29 U. S. C. § 1001 et seq. (1982 ed. and Supp. III), but we nevertheless concluded that federal jurisdiction was lacking.

This case does not pose a federal question of the first kind; respondents do not allege that federal law creates any of the causes of action that they have asserted.^[6] This case thus poses what Justice Frankfurter called the "litigation-provoking problem," Textile Workers v. Lincoln Mills, 353 [810] U. S. 448, 470 (1957) (dissenting opinion) — the presence of a federal issue in a state-created cause of action.

In undertaking this inquiry into whether jurisdiction may lie for the presence of a federal issue in a nonfederal cause of action, it is, of course, appropriate to begin by referring to our understanding of the statute conferring federal-question jurisdiction. We have consistently emphasized that, in exploring the outer reaches of § 1331, determinations about federal jurisdiction require sensitive judgments about congressional intent, judicial power, and the federal system. "If the history of the interpretation of judiciary legislation teaches us anything, it teaches the duty to reject treating such statutes as a wooden set of self-sufficient words. . . . The Act of 1875 is broadly phrased, but it has been continuously construed and limited in the light of the history that produced it, the demands of reason and coherence, and the dictates of sound judicial policy which have emerged from the Act's function as a provision in the mosaic of federal judiciary legislation." Romero v. International Terminal Operating Co., 358 U. S., at 379. In Franchise Tax Board, we forcefully reiterated this need for prudence and restraint in the jurisdictional inquiry: "We have always interpreted what Skelly Oil [Co. v. Phillips Petroleum Co., 339 U. S. 667, 673 (1950)] called `the current of jurisdictional legislation since the Act of March 3, 1875' . . . with an eye to practicality and necessity." 463 U. S., at 20.

In this case, both parties agree with the Court of Appeals' conclusion that there is no federal cause of action for FDCA violations. For purposes of our decision, we assume that this is a correct interpretation of the FDCA. Thus, as the case comes to us, it is appropriate to assume that, under the settled framework for evaluating whether a federal cause of action lies, some combination of the following factors is present: (1) the plaintiffs are not part of the class for whose special benefit the statute was passed; (2) the indicia of legislative [811] intent reveal no congressional purpose to provide a private cause of action; (3) a federal cause of action would not further the underlying purposes of the legislative scheme; and (4) the respondents' cause of action is a subject traditionally relegated to state law.^[7] In short, Congress did not intend a private federal remedy for violations of the statute that it enacted.

This is the first case in which we have reviewed this type of jurisdictional claim in light of these factors. That this is so is not surprising. The development of our framework for determining whether a private cause of action exists has proceeded only in the last 11 years, and its inception represented a significant change in our approach to congressional silence on the provision of federal remedies.^[8]

The recent character of that development does not, however, diminish its importance. Indeed, the very reasons for the development of the modern implied remedy doctrine — the "increased complexity of federal legislation and the increased volume of federal litigation," as well as "the desirability of a more careful scrutiny of legislative intent," Merrill Lynch, Pierce, Fenner & Smith, Inc. v. Curran, 456 U. S. 353, 377 (1982) (footnote omitted) — are precisely the kind of considerations that should inform the concern for "practicality and necessity" that Franchise Tax Board advised for the construction of § 1331 when jurisdiction is asserted [812] because of the presence of a federal issue in a state cause of action.

The significance of the necessary assumption that there is no federal private cause of action thus cannot be overstated. For the ultimate import of such a conclusion, as we have repeatedly emphasized, is that it would flout congressional intent to provide a private federal remedy for the violation of the federal statute.^[9] We think it would similarly flout, or at least undermine, congressional intent to conclude that the federal courts might nevertheless exercise federal-question jurisdiction and provide remedies for violations of that federal statute solely because the violation of the federal statute is said to be a "rebuttable presumption" or a "proximate cause" under state law, rather than a federal action under federal law.^[10]

[813] III

Petitioner advances three arguments to support its position that, even in the face of this congressional preclusion of a federal cause of action for a violation of the federal statute, federal-question jurisdiction may lie for the violation of the federal statute as an element of a state cause of action.

First, petitioner contends that the case represents a straightforward application of the statement in Franchise Tax Board that federal-question jurisdiction is appropriate when "it appears that some substantial, disputed question of federal law is a necessary element of one of the well-pleaded state claims." 463 U. S., at 13. Franchise Tax Board, however, did not purport to disturb the long-settled understanding that the mere presence of a federal issue in a state cause of action does not automatically confer federal-question jurisdiction.^[11] Indeed, in determining that federal-question jurisdiction was not appropriate in the case before us, we stressed Justice Cardozo's emphasis on principled, pragmatic distinctions: " `What is needed is something of that common-sense accommodation of judgment to kaleidoscopic situations which characterizes the law in its treatment of causation . . . a selective process which picks the substantial causes out of the web [814] and lays the other ones aside.' " Id., at 20-21 (quoting Gully v. First National Bank, 299 U. S. 109, 117-118 (1936)).

Far from creating some kind of automatic test, Franchise Tax Board thus candidly recognized the need for careful judgments about the exercise of federal judicial power in an area of uncertain jurisdiction. Given the significance of the assumed congressional determination to preclude federal private remedies, the presence of the federal issue as an element of the state tort is not the kind of adjudication for which jurisdiction would serve congressional purposes and the federal system. This conclusion is fully consistent with the very sentence relied on so heavily by petitioner. We simply conclude that the congressional determination that there should be no federal remedy for the violation of this federal statute is tantamount to a congressional conclusion that the presence of a claimed violation of the statute as an element of a state cause of action is insufficiently "substantial" to confer federal-question jurisdiction.^[12]

[815] Second, petitioner contends that there is a powerful federal interest in seeing that the federal statute is given uniform interpretations, and that federal review is the best way of insuring such uniformity. In addition to the significance of the congressional decision to preclude a federal remedy, we do [816] not agree with petitioner's characterization of the federal interest and its implications for federal-question jurisdiction. To the extent that petitioner is arguing that state use and interpretation of the FDCA pose a threat to the order and stability of the FDCA regime, petitioner should be arguing, not that federal courts should be able to review and enforce state FDCA-based causes of action as an aspect of federal-question jurisdiction, but that the FDCA pre-empts state-court jurisdiction over the issue in dispute.^[13] Petitioner's concern about the uniformity of interpretation, moreover, is considerably mitigated by the fact that, even if there is no original district court jurisdiction for these kinds of action, this Court retains power to review the decision of a federal issue in a state cause of action.^[14]

Finally, petitioner argues that, whatever the general rule, there are special circumstances that justify federal-question jurisdiction in this case. Petitioner emphasizes that it is unclear whether the FDCA applies to sales in Canada and Scotland; there is, therefore, a special reason for having a federal [817] court answer the novel federal question relating to the extra-territorial meaning of the Act. We reject this argument. We do not believe the question whether a particular claim arises under federal law depends on the novelty of the federal issue. Although it is true that federal jurisdiction cannot be based on a frivolous or insubstantial federal question, "the interrelation of federal and state authority and the proper management of the federal judicial system," Franchise Tax Board, 463 U. S., at 8, would be ill served by a rule that made the existence of federal-question jurisdiction depend on the district court's case-by-case appraisal of the novelty of the federal question asserted as an element of the state tort. The novelty of an FDCA issue is not sufficient to give it status as a federal cause of action; nor should it be sufficient to give a state-based FDCA claim status as a jurisdiction-triggering federal question.^[15]

IV

We conclude that a complaint alleging a violation of a federal statute as an element of a state cause of action, when Congress has determined that there should be no private, federal cause of action for the violation, does not state a claim "arising under the Constitution, laws, or treaties of the United States." 28 U. S. C. § 1331.

The judgment of the Court of Appeals is affirmed.

It is so ordered.

[818] JUSTICE BRENNAN, with whom JUSTICE WHITE, JUSTICE MARSHALL, and JUSTICE BLACKMUN join, dissenting.

Article III, § 2, of the Constitution provides that the federal judicial power shall extend to "all Cases, in Law and Equity, arising under this Constitution, the Laws of the United States, and Treaties made, or which shall be made, under their Authority." We have long recognized the great breadth of this grant of jurisdiction, holding that there is federal jurisdiction whenever a federal question is an "ingredient" of the action, Osborn v. Bank of the United States, 9 Wheat. 738, 823 (1824), and suggesting that there may even be jurisdiction simply because a case involves "potential federal questions," Textile Workers v. Lincoln Mills, 353 U. S. 448, 471 (1957) (Frankfurter, J., dissenting); see also Osborn, supra, at 824; Martin v. Hunter's Lessee, 1 Wheat. 304 (1816); Pacific Railroad Removal Cases, 115 U. S. 1 (1885); Verlinden B. V. v. Central Bank of Nigeria, 461 U. S. 480, 492-493 (1983).

Title 28 U. S. C. § 1331 provides, in language that parrots the language of Article III, that the district courts shall have original jurisdiction "of all civil actions arising under the Constitution, laws, or treaties of the United States." Although this language suggests that Congress intended in § 1331 to confer upon federal courts the full breadth of permissible "federal question" jurisdiction (an inference that is supported by the contemporary evidence, see Franchise Tax Board v. Construction Laborers Vacation Trust, 463 U. S. 1, 8, n. 8 (1983); Forrester, The Nature of a "Federal Question," 16 Tulane L. Rev. 362, 374-376 (1942); Shapiro, Jurisdiction and Discretion, 60 N. Y. U. L. Rev. 543, 568 (1985)), § 1331 has been construed more narrowly than its constitutional counterpart. See Verlinden B. V., supra, at 494-495; Romero v. International Terminal Operating Co., 358 U. S. 354, 379 (1959). Nonetheless, given the language of the statute and its close relation to the constitutional grant of federal-question jurisdiction, limitations on federal-question jurisdiction under § 1331 must be justified by careful consideration of the reasons [819] underlying the grant of jurisdiction and the need for federal review. Ibid. I believe that the limitation on federal jurisdiction recognized by the Court today is inconsistent with the purposes of § 1331. Therefore, I respectfully dissent.

I

While the majority of cases covered by § 1331 may well be described by Justice Holmes' adage that "[a] suit arises under the law that creates the cause of action," American Well Works Co. v. Layne & Bowler Co., 241 U. S. 257, 260 (1916), it is firmly settled that there may be federal-question jurisdiction even though both the right asserted and the remedy sought by the plaintiff are state created. See C. Wright, Federal Courts § 17, pp. 95-96 (4th ed. 1983) (hereinafter Wright); M. Redish, Federal Jurisdiction: Tensions in the Allocation of Judicial Power 64-71 (1980) (hereinafter Redish). The rule as to such cases was stated in what Judge Friendly described as "[t]he path-breaking opinion" in Smith v. Kansas City Title & Trust Co., 255 U. S. 180 (1921). T. B. Harms Co. v. Eliscu, 339 F. 2d 823, 827 (CA2 1964). In Smith, a shareholder of the defendant corporation brought suit in the federal court to enjoin the defendant from investing corporate funds in bonds issued under the authority of the Federal Farm Loan Act. The plaintiff alleged that Missouri law imposed a fiduciary duty on the corporation to invest only in bonds that were authorized by a valid law and argued that, because the Farm Loan Act was unconstitutional, the defendant could not purchase bonds issued under its authority. Although the cause of action was wholly state created, the Court held that there was original federal jurisdiction over the case:

"The general rule is that where it appears from the bill or statement of the plaintiff that the right to relief depends upon the construction or application of the Constitution or laws of the United States, and that such federal claim is not merely colorable, and rests upon a reasonable foundation, the District Court has jurisdiction [820] under [the statute granting federal question jurisdiction]." 255 U. S., at 199.

The continuing vitality of Smith is beyond challenge. We have cited it approvingly on numerous occasions, and reaffirmed its holding several times — most recently just three Terms ago by a unanimous Court in Franchise Tax Board v. Construction Laborers Vacation Trust, supra, at 9. See American Bank & Trust Co. v. Federal Reserve Bank of Atlanta, 256 U. S. 350, 357 (1921); Bell v. Hood, 327 U. S. 678, 685 (1946); Association of Westinghouse Salaried Employees v. Westinghouse Electric Corp., 348 U. S. 437, 450, and n. 18 (1955) (plurality opinion); Machinists v. Central Airlines, Inc., 372 U. S. 682, 696 (1963); Duke Power Co. v. Carolina Environmental Study Group, Inc., 438 U. S. 59, 70 (1978). See also Ashwander v. TVA, 297 U. S. 288, 356 (1936) (separate opinion of McReynolds, J.); Textile Workers v. Lincoln Mills, supra, at 470 (Frankfurter, J., dissenting); Wheeldin v. Wheeler, 373 U. S. 647, 659 (1963) (BRENNAN, J., dissenting). Cf. Gully v. First National Bank, 299 U. S. 109, 112 (1936) ("To bring a case within [§ 1331], a right or immunity created by the Constitution or laws of the United States must be an element, and an essential one, of the plaintiff's cause of action"). Moreover, in addition to Judge Friendly's authoritative opinion in T. B. Harms Co. v. Eliscu, supra, at 827, Smith has been widely cited and followed in the lower federal courts. See, e. g., Hanes Corp. v. Millard, 174 U. S. App. D. C. 253, 263, n. 8, 531 F. 2d 585, 595, n. 8 (1976); Mungin v. Florida East Coast R. Co., 416 F. 2d 1169, 1176-1177 (CA5 1969); Ivy Broadcasting Co. v. American Tel. & Tel. Co., 391 F. 2d 486, 492 (CA2 1968); Warrington Sewer Co. v. Tracy, 463 F. 2d 771, 772 (CA3 1972) (per curiam); New York by Abrams v. Citibank, N. A., 537 F. Supp. 1192, 1196 (SDNY 1982); Kravitz v. Homeowners Warranty Corp., 542 F. Supp. 317, 319 (ED Pa. 1982). See also Stone & Webster Engineering Corp. v. Ilsley, 690 F. 2d 323 (CA2 1982); Christopher v. Cavallo, 662 F. 2d 1082 (CA4 1981); Mountain Fuel Supply Co. v. Johnson Oil Co., 586 F. 2d 1375 (CA10 1978), [821] cert. denied, 441 U. S. 952 (1979); Garrett v. Time-D. C., Inc., 502 F. 2d 627 (CA9 1974), cert. denied, 421 U. S. 913 (1975); Sweeney v. Abramovitz, 449 F. Supp. 213 (Conn. 1978). Furthermore, the principle of the Smith case has been recognized and endorsed by most commentators as well. Redish 67, 69; American Law Institute, Study of the Division of Jurisdiction Between State and Federal Courts 178 (1969) (hereinafter ALI); Wright § 17, at 96; P. Bator, P. Mishkin, D. Shapiro, & H. Wechsler, Hart & Wechsler's The Federal Courts and the Federal System 889 (2d ed., 1973); Mishkin, The Federal "Question" in the District Courts, 53 Colum. L. Rev. 157, 166 (1953); Wechsler, Federal Jurisdiction and the Revision of the Judicial Code, 13 Law & Contemp. Prob. 216, 225 (1948).^[16]

[822] There is, to my mind, no question that there is federal jurisdiction over the respondents' fourth cause of action under the rule set forth in Smith and reaffirmed in Franchise Tax [823] Board. Respondents pleaded that petitioner's labeling of the drug Bendectin constituted "misbranding" in violation of §§ 201 and 502(f)(2) and (j) of the Federal Food, Drug, and Cosmetic Act (FDCA), 52 Stat. 1040, as amended, 21 U. S. C. § 301 et seq. (1982 ed. and Supp. III), and that this violation "directly and proximately caused" their injuries. App. 21-22 (Thompson complaint), 31-32 (MacTavish complaint). Respondents asserted in the complaint that this violation established petitioner's negligence per se and entitled them to recover damages without more. Ibid. No other basis for finding petitioner negligent was asserted in connection with this claim. As pleaded, then, respondents' "right to relief depend[ed] upon the construction or application of the Constitution or laws of the United States." Smith, 255 U. S., at 199; see also Franchise Tax Board, 463 U. S., at 28 (there is federal jurisdiction under § 1331 where the plaintiff's right to relief "necessarily depends" upon resolution of a federal question).^[17] Furthermore, although petitioner disputes its liability under the FDCA, it concedes that respondents' claim that petitioner violated the FDCA is "colorable, and rests upon a reasonable foundation." Smith, supra, at 199.^[18] [824] Of course, since petitioner must make this concession to prevail in this Court, it need not be accepted at face value. However, independent examination of respondents' claim substantiates the conclusion that it is neither frivolous nor meritless. As stated in the complaint, a drug is "misbranded" under the FDCA if "the labeling or advertising fails to reveal facts material . . . with respect to consequences which may result from the use of the article to which the labeling or advertising relates . . . ." 21 U. S. C. § 321(n). Obviously, the possibility that a mother's ingestion of Bendectin during pregnancy could produce malformed children is material. Petitioner's principal defense is that the Act does not govern the branding of drugs that are sold in foreign countries. It is certainly not immediately obvious whether this argument is correct. Thus, the statutory question is one which "discloses a need for determining the meaning or application of [the FDCA]," T. B. Harms Co. v. Eliscu, 339 F. 2d, at 827, and the claim raised by the fourth cause of action is one "arising under" federal law within the meaning of § 1331.

II

The Court apparently does not disagree with any of this — except, of course, for the conclusion. According to the Court, if we assume that Congress did not intend that there be a private federal cause of action under a particular federal law (and, presumably, a fortiori if Congress' decision not to create a private remedy is express), we must also assume that Congress did not intend that there be federal jurisdiction over a state cause of action that is determined by that federal law. Therefore, assuming — only because the parties [825] have made a similar assumption — that there is no private cause of action under the FDCA,^[19] the Court holds that there is no federal jurisdiction over the plaintiffs' claim:

"The significance of the necessary assumption that there is no federal private cause of action thus cannot be overstated. For the ultimate import of such a conclusion, as we have repeatedly emphasized, is that it would flout congressional intent to provide a private federal remedy for the violation of the federal statute. We think it would similarly flout, or at least undermine, congressional intent to conclude that the federal courts might nevertheless exercise federal-question jurisdiction and provide remedies for violations of that federal statute solely because the violation of the federal statute is said to be a `rebuttable presumption' or a `proximate cause' under state law, rather than a federal action under federal law." Ante, at 812 (footnotes omitted).

The Court nowhere explains the basis for this conclusion. Yet it is hardly self-evident. Why should the fact that Congress chose not to create a private federal remedy mean that Congress would not want there to be federal jurisdiction to adjudicate a state claim that imposes liability for violating the federal law? Clearly, the decision not to provide a private federal remedy should not affect federal jurisdiction unless the reasons Congress withholds a federal remedy are also reasons for withholding federal jurisdiction. Thus, it is necessary [826] to examine the reasons for Congress' decisions to grant or withhold both federal jurisdiction and private remedies, something the Court has not done.

A

In the early days of our Republic, Congress was content to leave the task of interpreting and applying federal laws in the first instance to the state courts; with one short-lived exception,^[20] Congress did not grant the inferior federal courts original jurisdiction over cases arising under federal law until 1875. Judiciary Act of 1875, ch. 137, § 1, 18 Stat. 470. The reasons Congress found it necessary to add this jurisdiction to the district courts are well known. First, Congress recognized "the importance, and even necessity of uniformity of decisions throughout the whole United States, upon all subjects within the purview of the constitution." Martin v. Hunter's Lessee, 1 Wheat., at 347-348 (Story, J.) (emphasis in original). See also, Comment, Federal Preemption, Removal Jurisdiction, and the Well-Pleaded Complaint Rule, 51 U. Chi. L. Rev. 634, 636 (1984) (hereinafter Comment); D. Currie, Federal Courts 160 (3d ed. 1982) (hereinafter Currie). Concededly, because federal jurisdiction is not always exclusive and because federal courts may disagree with one another, absolute uniformity has not been obtained even under § 1331. However, while perfect uniformity may not have been achieved, experience indicates that the availability of a federal forum in federal-question cases has done much to advance that goal. This, in fact, was the conclusion of the American Law Institute's Study of the Division of Jurisdiction Between State and Federal Courts. ALI 164-168.

In addition, § 1331 has provided for adjudication in a forum that specializes in federal law and that is therefore more likely to apply that law correctly. Because federal-question [827] cases constitute the basic grist for federal tribunals, "[t]he federal courts have acquired a considerable expertness in the interpretation and application of federal law." Id., at 164-165. By contrast, "it is apparent that federal question cases must form a very small part of the business of [state] courts." Id., at 165. As a result, the federal courts are comparatively more skilled at interpreting and applying federal law, and are much more likely correctly to divine Congress' intent in enacting legislation.^[21] See ibid.; Redish 71; Currie 160; Comment 636; Hornstein, Federalism, Judicial Power and the "Arising Under" Jurisdiction of the Federal Courts: A Hierarchical Analysis, 56 Ind. L. J. 563, 564-565 (1981).

These reasons for having original federal-question jurisdiction explain why cases like this one and Smith — i. e., cases where the cause of action is a creature of state law, but an [828] essential element of the claim is federal — "arise under" federal law within the meaning of § 1331. Congress passes laws in order to shape behavior; a federal law expresses Congress' determination that there is a federal interest in having individuals or other entities conform their actions to a particular norm established by that law. Because all laws are imprecise to some degree, disputes inevitably arise over what specifically Congress intended to require or permit. It is the duty of courts to interpret these laws and apply them in such a way that the congressional purpose is realized. As noted above, Congress granted the district courts power to hear cases "arising under" federal law in order to enhance the likelihood that federal laws would be interpreted more correctly and applied more uniformly. In other words, Congress determined that the availability of a federal forum to adjudicate cases involving federal questions would make it more likely that federal laws would shape behavior in the way that Congress intended.

By making federal law an essential element of a state-law claim, the State places the federal law into a context where it will operate to shape behavior: the threat of liability will force individuals to conform their conduct to interpretations of the federal law made by courts adjudicating the state-law claim. It will not matter to an individual found liable whether the officer who arrives at his door to execute judgment is wearing a state or a federal uniform; all he cares about is the fact that a sanction is being imposed — and may be imposed again in the future — because he failed to comply with the federal law. Consequently, the possibility that the federal law will be incorrectly interpreted in the context of adjudicating the state-law claim implicates the concerns that led Congress to grant the district courts power to adjudicate cases involving federal questions in precisely the same way as if it was federal law that "created" the cause of action. It therefore follows that there is federal jurisdiction under § 1331.

[829] B

The only remaining question is whether the assumption that Congress decided not to create a private cause of action alters this analysis in a way that makes it inappropriate to exercise original federal jurisdiction. According to the Court, "the very reasons for the development of the modern implied remedy doctrine" support the conclusion that, where the legislative history of a particular law shows (whether expressly or by inference) that Congress intended that there be no private federal remedy, it must also mean that Congress would not want federal courts to exercise jurisdiction over a state-law claim making violations of that federal law actionable. Ante, at 811. These reasons are " `the increased complexity of federal legislation,' " " `the increased volume of federal litigation,' " and " `the desirability of a more careful scrutiny of legislative intent.' " Ibid. (quoting Merrill Lynch, Pierce, Fenner & Smith, Inc. v. Curran, 456 U. S. 353, 377 (1982)).

These reasons simply do not justify the Court's holding. Given the relative expertise of the federal courts in interpreting federal law, supra, at 826-827, the increased complexity of federal legislation argues rather strongly in favor of recognizing federal jurisdiction. And, while the increased volume of litigation may appropriately be considered in connection with reasoned arguments that justify limiting the reach of § 1331, I do not believe that the day has yet arrived when this Court may trim a statute solely because it thinks that Congress made it too broad.^[22]

[830] This leaves only the third reason: " `the desirability of a more careful scrutiny of legislative intent.' " Ante, at 811. I certainly subscribe to the proposition that the Court should consider legislative intent in determining whether or not there is jurisdiction under § 1331. But the Court has not examined the purposes underlying either the FDCA or § 1331 in reaching its conclusion that Congress' presumed decision not to provide a private federal remedy under the FDCA must be taken to withdraw federal jurisdiction over a private state remedy that imposes liability for violating the FDCA. Moreover, such an examination demonstrates not only that it is consistent with legislative intent to find that there is federal jurisdiction over such a claim, but, indeed, that it is the Court's contrary conclusion that is inconsistent with congressional intent.

The enforcement scheme established by the FDCA is typical of other, similarly broad regulatory schemes. Primary responsibility for overseeing implementation of the Act has been conferred upon a specialized administrative agency, here the Food and Drug Administration (FDA).^[23] Congress has provided the FDA with a wide-ranging arsenal of weapons to combat violations of the FDCA, including authority to obtain an ex parte court order for the seizure of goods subject to the Act, see 21 U. S. C. § 334, authority to initiate proceedings in a federal district court to enjoin continuing violations of the FDCA, see § 332, and authority to request a United States Attorney to bring criminal proceedings against violators, see § 333. See generally 1 J. O'Reilly, Food and Drug Administration, chs. 6-10 (1979 and Supp. 1985). Significantly, the FDA has no independent enforcement authority; final enforcement must come from the federal courts. [831] which have exclusive jurisdiction over actions under the FDCA. See §§ 332(a), 333, 334(a)(1). Thus, while the initial interpretive function has been delegated to an expert administrative body whose interpretations are entitled to considerable deference, final responsibility for interpreting the statute in order to carry out the legislative mandate belongs to the federal courts. Cf. Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U. S. 837, 843, n. 9 (1984) ("The judiciary is the final authority on issues of statutory construction and must reject administrative constructions which are contrary to clear congressional intent").

Given that Congress structured the FDCA so that all express remedies are provided by the federal courts, it seems rather strange to conclude that it either "flout[s]" or "undermine[s]" congressional intent for the federal courts to adjudicate a private state-law remedy that is based upon violating the FDCA. See ante, at 812. That is, assuming that a state cause of action based on the FDCA is not preempted, it is entirely consistent with the FDCA to find that it "arises under" federal law within the meaning of § 1331. Indeed, it is the Court's conclusion that such a state cause of action must be kept out of the federal courts that appears contrary to legislative intent inasmuch as the enforcement provisions of the FDCA quite clearly express a preference for having federal courts interpret the FDCA and provide remedies for its violation.

It may be that a decision by Congress not to create a private remedy is intended to preclude all private enforcement. If that is so, then a state cause of action that makes relief available to private individuals for violations of the FDCA is pre-empted. But if Congress' decision not to provide a private federal remedy does not pre-empt such a state remedy, then, in light of the FDCA's clear policy of relying on the federal courts for enforcement, it also should not foreclose federal jurisdiction over that state remedy. Both § 1331 and the enforcement provisions of the FDCA reflect Congress' strong [832] desire to utilize the federal courts to interpret and enforce the FDCA, and it is therefore at odds with both these statutes to recognize a private state-law remedy for violating the FDCA but to hold that this remedy cannot be adjudicated in the federal courts.

The Court's contrary conclusion requires inferring from Congress' decision not to create a private federal remedy that, while some private enforcement is permissible in state courts, it is "bad" if that enforcement comes from the federal courts. But that is simply illogical. Congress' decision to withhold a private right of action and to rely instead on public enforcement reflects congressional concern with obtaining more accurate implementation and more coordinated enforcement of a regulatory scheme. See National Railroad Passenger Corporation v. National Assn. of Railroad Passengers, 414 U. S. 453, 462-465 (1974); Holloway v. Bristol-Myers Corp., 158 U. S. App. D. C. 207, 218-220, 485 F. 2d 986, 997-999 (1973); Stewart & Sunstein, Public Programs and Private Rights, 95 Harv. L. Rev. 1193, 1208-1209 (1982). These reasons are closely related to the Congress' reasons for giving federal courts original federal-question jurisdiction. Thus, if anything, Congress' decision not to create a private remedy strengthens the argument in favor of finding federal jurisdiction over a state remedy that is not pre-empted.

[1] App. 36-37. The petition also alleged that the action "is between citizens of a State and citizens or subjects of a foreign state." Id.,at 36. Because petitioner is a corporation with its principal place of business in Ohio, however, the removal was not proper unless the action was founded on a claim arising under federal law. Title 28 U. S. C. § 1441(b) provides:

"(b) Any civil action of which the district courts have original jurisdiction founded on a claim or right arising under the Constitution, treaties or laws of the United States shall be removable without regard to the citizenship or residence of the parties. Any other such action shall be removable only if none of the parties in interest properly joined and served as defendants is a citizen of the State in which such action is brought."

[2] " `Under our interpretations, Congress has given the lower courts jurisdiction to hear, originally or by removal from a state court, only those cases in which a well-pleaded complaint establishes either that federal law creates the cause of action or that the plaintiff's right to relief necessarily depends on resolution of a substantial question of federal law.' " 766 F. 2d, at 1006 (quoting Franchise Tax Board, 463 U. S., at 28).

[3] See Art. III, § 2 ("The judicial Power shall extend to all Cases, in Law and Equity, arising under this Constitution, the Laws of the United States, and Treaties made, or which shall be made, under their Authority . . .").

[4] Act of Mar. 3, 1875, § 1, 18 Stat. 470. As currently codified, the statute provides: "The district courts shall have original jurisdiction of all civil actions arising under the Constitution, laws or treaties of the United States." 28 U. S. C. § 1331.

[5] The case most frequently cited for that proposition is Smith v. Kansas City Title & Trust Co.,255 U. S. 180 (1921). In that case the Court upheld federal jurisdiction of a shareholder's bill to enjoin the corporation from purchasing bonds issued by the federal land banks under the authority of the Federal Farm Loan Act on the ground that the federal statute that authorized the issuance of the bonds was unconstitutional. The Court stated:

"The general rule is that where it appears from the bill or statement of the plaintiff that the right to relief depends upon the construction or application of the Constitution or laws of the United States, and that such federal claim is not merely colorable, and rests upon a reasonable foundation, the District Court has jurisdiction under this provision." Id., at 199.

The effect of this view, expressed over Justice Holmes' vigorous dissent, on his American Well Works formulation has been often noted. See, e. g., Franchise Tax Board, 463 U. S., at 9 ("[I]t is well settled that Justice Holmes' test is more useful for describing the vast majority of cases that come within the district courts' original jurisdiction than it is for describing which cases are beyond district court jurisdiction"); T. B. Harms Co. v. Eliscu, 339 F. 2d 823, 827 (CA2 1964) (Friendly, J.) ("It has come to be realized that Mr. Justice Holmes' formula is more useful for inclusion than for the exclusion for which it was intended").

[6] Jurisdiction may not be sustained on a theory that the plaintiff has not advanced. See Healy v. Sea Gull Specialty Co., 237 U. S. 479, 480 (1915) ("[T]he plaintiff is absolute master of what jurisdiction he will appeal to"); The Fair v. Kohler Die & Specialty Co., 228 U. S. 22, 25 (1913) ("[T]he party who brings a suit is master to decide what law he will rely upon"). See also United States v. Mottaz, 476 U. S. 834, 850 (1986).

[7] See California v. Sierra Club, 451 U. S. 287, 293 (1981); Cannon v. University of Chicago, 441 U. S. 677, 689-709 (1979); Cort v. Ash, 422 U. S. 66, 78 (1975).

[8] See Merrill Lynch, Pierce, Fenner & Smith, Inc. v. Curran, 456 U. S. 353, 377 (1982) ("In 1975 the Court unanimously decided to modify its approach to the question whether a federal statute includes a private right of action"). Cf. Middlesex County Sewerage Authority v. National Sea Clammers Assn., 453 U. S. 1, 25 (1981) (STEVENS, J., concurring in judgment in part and dissenting in part) ("In 1975, in Cort v. Ash, 422 U. S. 66, the Court cut back on the simple common-law presumption by fashioning a four-factor formula that led to the denial of relief in that case").

[9] See, e. g., Daily Income Fund, Inc. v. Fox, 464 U. S. 523, 535-536 (1984) ("In evaluating such a claim, our focus must be on the intent of Congress when it enacted the statute in question"); Middlesex County Sewerage Authority v. National Sea Clammers Assn., 453 U. S., at 13 ("The key to the inquiry is the intent of the Legislature"); Texas Industries, Inc. v. Radcliff Materials, Inc., 451 U. S. 630, 639 (1981) ("Our focus, as it is in any case involving the implication of a right of action, is on the intent of Congress"); California v. Sierra Club, 451 U. S., at 293 ("[T]he ultimate issue is whether Congress intended to create a private right of action"); Northwest Airlines, Inc. v. Transport Workers, 451 U. S. 77, 91 (1981) ("The ultimate question in cases such as this is whether Congress intended to create the private remedy"); Transamerica Mortgage Advisors, Inc. v. Lewis, 444 U. S. 11, 15 (1979) ("The question whether a statute creates a cause of action, either expressly or by implication, is basically a matter of statutory construction"); Touche Ross & Co. v. Redington, 442 U. S. 560, 568 (1979) ("The question of the existence of a statutory cause of action is, of course, one of statutory construction").

[10] When we conclude that Congress has decided not to provide a particular federal remedy, we are not free to "supplement" that decision in a way that makes it "meaningless." Cf. Mobil Oil Corp. v. Higginbotham, 436 U. S. 618, 625 (1978) (When Congress "does speak directly to a question, the courts are not free to `supplement' Congress' answer so thoroughly that the Act becomes meaningless"). See also California v. Sierra Club, 451 U. S., at 297 ("The federal judiciary will not engraft a remedy on a statute, no matter how salutary, that Congress did not intend to provide").

[11] See, e. g., Textile Workers v. Lincoln Mills, 353 U. S. 448, 470 (1957) (Frankfurter, J., dissenting) (defining inquiry as "the degree to which federal law must be in the forefront of the case and not collateral, peripheral or remote"); Gully v. First National Bank, 299 U. S. 109, 115 (1936) ("Not every question of federal law emerging in a suit is proof that a federal law is the basis of the suit"); id., at 118 ("If we follow the ascent far enough, countless claims of right can be discovered to have their source or their operative limits in the provisions of a federal statute or in the Constitution itself with its circumambient restrictions upon legislative power. To set bounds to the pursuit, the courts have formulated the distinction between controversies that are basic and those that are collateral, between disputes that are necessary and those that are merely possible. We shall be lost in a maze if we put that compass by").

[12] Several commentators have suggested that our § 1331 decisions can best be understood as an evaluation of the nature of the federal interest at stake. See, e. g., Shapiro, Jurisdiction and Discretion, 60 N. Y. U. L. Rev. 543, 568 (1985); C. Wright, Federal Courts 96 (4th ed. 1983); Cohen, The Broken Compass: The Requirement That a Case Arise "Directly" Under Federal Law, 115 U. Pa. L. Rev. 890, 916 (1967). Cf. Kravitz v. Homeowners Warranty Corp.,542 F. Supp. 317, 320 (ED Pa. 1982) (Pollak, J.) ("I cannot identify any compelling reasons of federal judicial policy for embracing a case of this kind as a federal question case. The essential Pennsylvania elements of plaintiffs' suit for rescission would be more appropriately dealt with by a Court of Common Pleas than by this court; and, with respect to the lesser-included issue of federal law, Pennsylvania's courts are fully competent to interpret the Magnuson-Moss Warranty Act and the relevant F. T. C. regulations, subject to review by the United States Supreme Court").

Focusing on the nature of the federal interest, moreover, suggests that the widely perceived "irreconcilable" conflict between the finding of federal jurisdiction in Smith v. Kansas City Title & Trust Co., 255 U. S. 180 (1921), and the finding of no jurisdiction in Moore v. Chesapeake & Ohio R. Co., 291 U. S. 205 (1934), see, e. g., M. Redish, Federal Jurisdiction: Tensions in the Allocation of Judicial Power 67 (1980), is far from clear. For the difference in results can be seen as manifestations of the differences in the nature of the federal issues at stake. In Smith, as the Court emphasized, the issue was the constitutionality of an important federal statute. See 255 U. S., at 201 ("It is . . . apparent that the controversy concerns the constitutional validity of an act of Congress which is directly drawn in question. The decision depends upon the determination of this issue"). In Moore, in contrast, the Court emphasized that the violation of the federal standard as an element of state tort recovery did not fundamentally change the state tort nature of the action. See 291 U. S., at 216-217 (" `The action fell within the familiar category of cases involving the duty of a master to his servant. This duty is defined by the common law, except as it may be modified by legislation. The federal statute, in the present case, touched the duty of the master at a single point and, save as provided in the statute, the right of the plaintiff to recover was left to be determined by the law of the State' ") (quoting Minneapolis, St. P. & S. S. M. R. Co. v. Popplar, 237 U. S. 369, 372 (1915)).

The importance of the nature of the federal issue in federal-question jurisdiction is highlighted by the fact that, despite the usual reliability of the Holmes test as an inclusionary principle, this Court has sometimes found that formally federal causes of action were not properly brought under federal-question jurisdiction because of the overwhelming predominance of state-law issues. See Shulthis v. McDougal, 225 U. S. 561, 569-570 (1912) ("A suit to enforce a right which takes its origin in the laws of the United States is not necessarily, or for that reason alone, one arising under those laws, for a suit does not so arise unless it really and substantially involves a dispute or controversy respecting the validity, construction or effect of such a law, upon the determination of which the result depends. This is especially so of a suit involving rights to land acquired under a law of the United States. If it were not, every suit to establish title to land in the central and western States would so arise, as all titles in those States are traceable back to those laws"); Shoshone Mining Co. v. Rutter, 177 U. S. 505, 507 (1900) ("We pointed out in the former opinion that it was well settled that a suit to enforce a right which takes its origin in the laws of the United States is not necessarily one arising under the Constitution or laws of the United States, within the meaning of the jurisdiction clauses, for if it did every action to establish title to real estate (at least in the newer States) would be such a one, as all titles in those States come from the United States or by virtue of its laws").

[13] Cf. Longshoremen v. Davis, 476 U. S. 380, 391 (1986) ("[O]ur decisions describing the nature of Garmon pre-emption and defining its boundaries have rested on a determination that in enacting the [National Labor Relations Act] Congress intended for the [National Labor Relations] Board generally to exercise exclusive jurisdiction in this area").

[14] See Moore v. Chesapeake & Ohio R. Co., 291 U. S., at 214-215 ("Questions arising in actions in state courts to recover for injuries sustained by employees in intrastate commerce and relating to the scope or construction of the Federal Safety Appliance Acts are, of course, federal questions which may appropriately be reviewed in this Court. . . . But it does not follow that a suit brought under the state statute which defines liability to employees who are injured while engaged in intrastate commerce, and brings within the purview of the statute a breach of the duty imposed by the federal statute, should be regarded as a suit arising under the laws of the United States and cognizable in the federal court in the absence of diversity of citizenship"). Cf. Franchise Tax Board, 463 U. S., at 12, n. 12 ("[T]he absence of original jurisdiction does not mean that there is no federal forum in which a pre-emption defense may be heard. If the state courts reject a claim of federal pre-emption, that decision may ultimately be reviewed on appeal by this Court").

[15] Petitioner also contends that the Court of Appeals opinion rests on a view that federal-question jurisdiction was inappropriate because, whatever the role of the federal issue in the FDCA-related count, the plaintiff could recover on other, strictly state-law claims. See 766 F. 2d, at 1006 (noting that "the jury could find negligence on the part of Merrell Dow without finding a violation of the FDCA"). To the extent that the opinion can be read to express such a view, we agree that it was erroneous. If the FDCA-related count presented a sufficient federal question, its relationship to the other, state-law claims would be determined by the ordinary principles of pendent jurisdiction described in Mine Workers v. Gibbs, 383 U. S. 715 (1966). For the reasons that we have stated, however, there is no federal-question jurisdiction even with that possible error corrected.

[16] Some commentators have argued that the result in Smith conflicts with our decision in Moore v. Chesapeake & Ohio R. Co., 291 U. S. 205 (1934). See, e. g., Greene, Hybrid State Law in the Federal Courts, 83 Harv. L. Rev. 289, 323 (1969). In Moore, the plaintiff brought an action under Kentucky's Employer Liability Act, which provided that a plaintiff could not be held responsible for contributory negligence or assumption of risk where his injury resulted from the violation of any state or federal statute enacted for the safety of employees. The plaintiff in Moorealleged that his injury was due to the defendant's failure to comply with the Federal Safety Appliance Act; therefore, an important issue in the adjudication of the state cause of action was whether the terms of the federal law had been violated. The Court could have dismissed the complaint on the ground that the federal issue would arise only in response to a defense of contributory negligence or assumption of risk, and that therefore there was no jurisdiction under the well-pleaded complaint rule. Instead, the Court held that "a suit brought under the state statute which defines liability to employees who are injured while engaged in intrastate commerce, and brings within the purview of the statute a breach of the duty imposed by the federal statute, should [not] be regarded as a suit arising under the laws of the United States and cognizable in the federal court in the absence of diversity of citizenship." 291 U. S., at 214-215.

The Court suggests that Smith and Moore may be reconciled if one views the question whether there is jurisdiction under § 1331 as turning upon "an evaluation of the nature of the federal interest at stake." Ante, at 814, n. 12 (emphasis in original). Thus, the Court explains, while in Smith the issue was the constitutionality of "an important federal statute," in Moore the federal interest was less significant in that "the violation of the federal standard as an element of state tort recovery did not fundamentally change the state tort nature of the action." Ante, at 815, n. 12.

In one sense, the Court is correct in asserting that we can reconcile Smith and Moore on the ground that the "nature" of the federal interest was more significant in Smith than in Moore. Indeed, as the Court appears to believe, ante, at 814-815, n. 12, we could reconcile many of the seemingly inconsistent results that have been reached under § 1331 with such a test. But this is so only because a test based upon an ad hoc evaluation of the importance of the federal issue is infinitely malleable: at what point does a federal interest become strong enough to create jurisdiction? What principles guide the determination whether a statute is "important" or not? Why, for instance, was the statute in Smith so "important" that direct review of a state-court decision (under our mandatory appellate jurisdiction) would have been inadequate? Would the result in Moore have been different if the federal issue had been a more important element of the tort claim? The point is that if one makes the test sufficiently vague and general, virtually any set of results can be "reconciled." However, the inevitable — and undesirable — result of a test such as that suggested in the Court's footnote 12 is that federal jurisdiction turns in every case on an appraisal of the federal issue, its importance and its relation to state-law issues. Yet it is precisely because the Court believes that federal jurisdiction would be "ill served" by such a case-by-case appraisal that it rejects petitioner's claim that the difficulty and importance of the statutory issue presented by its claim suffices to confer jurisdiction under § 1331. Ante, at 817. The Court cannot have it both ways.

My own view is in accord with those commentators who view the results in Smith and Moore as irreconcilable. See, e. g., Redish 67; D. Currie, Federal Jurisdiction in a Nutshell 109 (2d ed. 1981). That fact does not trouble me greatly, however, for I view Moore as having been a "sport" at the time it was decided and having long been in a state of innocuous desuetude. Unlike the jurisdictional holding in Smith, the jurisdictional holding in Moore has never been relied upon or even cited by this Court. Moore has similarly borne little fruit in the lower courts, leading Professor Redish to conclude after comparing the vitality of Smith and Moore that "the principle enunciated in Smith is the one widely followed by modern lower federal courts." Redish 67. Finally, as noted in text, the commentators have also preferred Smith. Supra, at 821. Moore simply has not survived the test of time; it is presently moribund, and, to the extent that it is inconsistent with the well-established rule of the Smith case, it ought to be overruled.

[17] As the Court correctly notes, the Court of Appeals erred in holding that respondents' right to relief did not depend upon the resolution of a federal question because respondents might prevail on one of their other, wholly state-law claims. The fourth cause of action presents an independent and independently sufficient claim for relief. Whether it "arises under" federal law within the meaning of § 1331 must therefore be determined without reference to any other claims, as if only that claim was asserted. If, after such consideration, it is determined that there is jurisdiction, the plaintiff may join additional state-law claims meeting the test for pendent jurisdiction set forth in Mine Workers v. Gibbs, 383 U. S. 715 (1966). See ante, at 817, n. 15.

[18] Franchise Tax Board states that the plaintiff's right to relief must necessarily depend upon resolution of a "substantial" federal question. 463 U. S., at 28. In context, however, it is clear that this was simply another way of stating that the federal question must be colorable and have a reasonable foundation. This understanding is consistent with the manner in which the Smith test has always been applied, as well as with the way we have used the concept of a "substantial" federal question in other cases concerning federal jurisdiction. See, e. g., Hagans v. Lavine, 415 U. S. 528, 536-537 (1974); Bell v. Hood, 327 U. S. 678, 682 (1946).

[19] It bears emphasizing that the Court does not hold that there is no private cause of action under the FDCA. Rather, it expressly states that "[f]or purposes of our decision, we assume that this is a correct interpretation of the FDCA." Ante, at 810. The Court simply holds petitioner to its concession that the FDCA provides no private remedy, and decides petitioner's claim on the basis of this concession. I shall do the same. Under the Court's analysis, however, if a party persuaded a court that there is a private cause of action under the FDCA, there would be federal jurisdiction under Smith and Franchise Tax Board over a state cause of action making violations of the FDCA actionable. Such jurisdiction would apparently exist even if the plaintiff did not seek the federal remedy.

[20] Congress granted original federal-question jurisdiction briefly in the Midnight Judges Act, ch. 4, § 11, 2 Stat. 92 (1801), which was repealed in 1802, Act of Mar. 8, 1802, ch. 8, § 1, 2 Stat. 132.

[21] Another reason Congress conferred original federal-question jurisdiction on the district courts was its belief that state courts are hostile to assertions of federal rights. See Hornstein, Federalism, Judicial Power and the "Arising Under" Jurisdiction of the Federal Courts: A Hierarchical Analysis, 56 Ind. L. J. 563, 564-565 (1981); Comment 636; Redish 71. Although this concern may be less compelling today than it once was, the American Law Institute reported as recently as 1969 that "it is difficult to avoid concluding that federal courts are more likely to apply federal law sympathetically and understandingly than are state courts." ALI 166. In any event, this rationale is, like the rationale based on the expertise of the federal courts, simply an expression of Congress' belief that federal courts are more likely to interpret federal law correctly.

One might argue that this Court's appellate jurisdiction over state-court judgments in cases arising under federal law can be depended upon to correct erroneous state-court decisions and to insure that federal law is interpreted and applied uniformly. However, as any experienced observer of this Court can attest, "Supreme Court review of state courts, limited by docket pressures, narrow review of the facts, the debilitating possibilities of delay, and the necessity of deferring to adequate state grounds of decision, cannot do the whole job." Currie 160. Indeed, having served on this Court for 30 years, it is clear to me that, realistically, it cannot even come close to "doing the whole job" and that § 1331 is essential if federal rights are to be adequately protected.

[22] Cf. Cohens v. Virginia, 6 Wheat. 264, 404 (1821) (Marshall, C. J.) ("It is most true that this Court will not take jurisdiction if it should not; but it is equally true, that it must take jurisdiction if it should. . . . We have no more right to decline the exercise of jurisdiction which is given, than to usurp that which is not given"). The narrow exceptions we have recognized to Chief Justice Marshall's famous dictum have all been justified by compelling judicial concerns of comity and federalism. See, e. g., Younger v. Harris, 401 U. S. 37 (1971); Burford v. Sun Oil Co., 319 U. S. 315 (1943). It would be wholly illegitimate, however, for this Court to determine that there was no jurisdiction over a class of cases simply because the Court thought that there were too many cases in the federal courts.

[23] The Federal Trade Commission retains regulatory and enforcement authority over the advertising (as opposed to the labeling) of foods, drugs, and cosmetics. See 15 U. S. C. §§ 52-55.

[1062] Jane Webre, Austin, Texas, for Petitioners.

Thomas M. Michel, Fort Worth, Texas, for Respondent.

Jane M.N. Webre, Cynthia S. Connolly, Scott, Douglass & McConnico, L.L.P., Austin, Texas, Robert S. Harrell, Charles B. Walker, Jr., Fulbright & Jaworski, L.L.P., Houston, Texas, David E. Keltner, Kelly Hart & Hallman, L.L.P., Fort Worth, Texas, for Petitioners.

Thomas M. Michel, Robley E. Sicard, Griffith, Jay & Michel, LLP, Fort Worth, Texas, Coyt Randal Johnston, Robert L. Tobey, Coyt Randal Johnston, Jr., Johnston Tobey, P.C., Dallas, Texas, Theodore F. Shiells, Shiells Law Firm P.C., Dallas, Texas, Gregory W. Carr, Carr LLP, Frisco, Texas, Daniel R. Ortiz, Charlottesville, Virginia, for Respondent.

Chief Justice ROBERTS delivered the opinion of the Court.

Federal courts have exclusive jurisdiction over cases "arising under any Act of Congress relating to patents." 28 U.S.C. § 1338(a). The question presented is whether a state law claim alleging legal malpractice in the handling of a patent case must be brought in federal court.

I

In the early 1990s, respondent Vernon Minton developed a computer program and telecommunications network designed to facilitate securities trading. In March 1995, he leased the system — known as the Texas Computer Exchange Network, or TEXCEN — to R.M. Stark & Co., a securities brokerage. A little over a year later, he applied for a patent for an interactive securities trading system that was based substantially on TEXCEN. The U.S. Patent and Trademark Office issued the patent in January 2000.

Patent in hand, Minton filed a patent infringement suit in Federal District Court against the National Association of Securities Dealers, Inc. (NASD) and the NASDAQ Stock Market, Inc. He was represented by Jerry Gunn and the other petitioners. NASD and NASDAQ moved for summary judgment on the ground that Minton's patent was invalid under the "on sale" bar, 35 U.S.C. § 102(b). That provision specifies that an inventor is not entitled to a patent if "the invention was ... on sale in [the United States], more than one year prior to the date of the application," and Minton had leased TEXCEN to Stark more than one year prior to filing his patent application. Rejecting Minton's argument that there were differences between TEXCEN and the patented system that precluded application of the on-sale bar, the District Court granted the summary judgment motion and declared Minton's patent invalid. Minton v. National Assn. of Securities Dealers, Inc., 226 F.Supp.2d 845, 873, 883-884 (E.D.Tex. 2002).

Minton then filed a motion for reconsideration in the District Court, arguing for the first time that the lease agreement with Stark was part of ongoing testing of TEXCEN and therefore fell within the [1063] "experimental use" exception to the on-sale bar. See generally Pfaff v. Wells Electronics, Inc., 525 U.S. 55, 64, 119 S.Ct. 304, 142 L.Ed.2d 261 (1998) (describing the exception). The District Court denied the motion. Minton v. National Assn. of Securities Dealers, Inc., No. 9:00-cv-00019 (ED Tex., July 15, 2002).

Minton appealed to the U.S. Court of Appeals for the Federal Circuit. That court affirmed, concluding that the District Court had appropriately held Minton's experimental-use argument waived. See Minton v. National Assn. of Securities Dealers, Inc., 336 F.3d 1373, 1379-1380 (C.A.Fed.2003).

Minton, convinced that his attorneys' failure to raise the experimental-use argument earlier had cost him the lawsuit and led to invalidation of his patent, brought this malpractice action in Texas state court. His former lawyers defended on the ground that the lease to Stark was not, in fact, for an experimental use, and that therefore Minton's patent infringement claims would have failed even if the experimental-use argument had been timely raised. The trial court agreed, holding that Minton had put forward "less than a scintilla of proof" that the lease had been for an experimental purpose. App. 213. It accordingly granted summary judgment to Gunn and the other lawyer defendants.

On appeal, Minton raised a new argument: Because his legal malpractice claim was based on an alleged error in a patent case, it "aris[es] under" federal patent law for purposes of 28 U.S.C. § 1338(a). And because, under § 1338(a), "[n]o State court shall have jurisdiction over any claim for relief arising under any Act of Congress relating to patents," the Texas court — where Minton had originally brought his malpractice claim — lacked subject matter jurisdiction to decide the case. Accordingly, Minton argued, the trial court's order should be vacated and the case dismissed, leaving Minton free to start over in the Federal District Court.

A divided panel of the Court of Appeals of Texas rejected Minton's argument. Applying the test we articulated in Grable & Sons Metal Products, Inc. v. Darue Engineering & Mfg., 545 U.S. 308, 314, 125 S.Ct. 2363, 162 L.Ed.2d 257 (2005), it held that the federal interests implicated by Minton's state law claim were not sufficiently substantial to trigger § 1338 "arising under" jurisdiction. It also held that finding exclusive federal jurisdiction over state legal malpractice actions would, contrary to Grable's commands, disturb the balance of federal and state judicial responsibilities. Proceeding to the merits of Minton's malpractice claim, the Court of Appeals affirmed the trial court's determination that Minton had failed to establish experimental use and that arguments on that ground therefore would not have saved his infringement suit.

The Supreme Court of Texas reversed, relying heavily on a pair of cases from the U.S. Court of Appeals for the Federal Circuit. 355 S.W.3d 634, 641-642 (2011) (discussing Air Measurement Technologies, Inc. v. Akin Gump Strauss Hauer & Feld, L.L. P., 504 F.3d 1262 (2007); Immunocept, LLC v. Fulbright & Jaworski, LLP, 504 F.3d 1281 (2007)). The Court concluded that Minton's claim involved "a substantial federal issue" within the meaning of Grable "because the success of Minton's malpractice claim is reliant upon the viability of the experimental use exception as a defense to the on-sale bar." 355 S.W.3d, at 644. Adjudication of Minton's claim in federal court was consistent with the appropriate balance between federal and state judicial responsibilities, it held, because "the federal government and patent litigants have an interest in the uniform application of patent law by courts [1064] well-versed in that subject matter." Id., at 646 (citing Immunocept, supra, at 1285-1286; Air Measurement Technologies, supra, at 1272).

Justice Guzman, joined by Justices Medina and Willett, dissented. The dissenting justices would have held that the federal issue was neither substantial nor disputed, and that maintaining the proper balance of responsibility between state and federal courts precluded relegating state legal malpractice claims to federal court.

We granted certiorari. 568 U.S. ___, 133 S.Ct. 420, 184 L.Ed.2d 251 (2012).

II

"Federal courts are courts of limited jurisdiction," possessing "only that power authorized by Constitution and statute." Kokkonen v. Guardian Life Ins. Co. of America, 511 U.S. 375, 377, 114 S.Ct. 1673, 128 L.Ed.2d 391 (1994). There is no dispute that the Constitution permits Congress to extend federal court jurisdiction to a case such as this one, see Osborn v. Bank of United States, 9 Wheat. 738, 823-824, 6 L.Ed. 204 (1824); the question is whether Congress has done so, see Powell v. McCormack, 395 U.S. 486, 515-516, 89 S.Ct. 1944, 23 L.Ed.2d 491 (1969).

As relevant here, Congress has authorized the federal district courts to exercise original jurisdiction in "all civil actions arising under the Constitution, laws, or treaties of the United States," 28 U.S.C. § 1331, and, more particularly, over "any civil action arising under any Act of Congress relating to patents," § 1338(a). Adhering to the demands of "[l]inguistic consistency," we have interpreted the phrase "arising under" in both sections identically, applying our § 1331 and § 1338(a) precedents interchangeably. See Christianson v. Colt Industries Operating Corp., 486 U.S. 800, 808-809, 108 S.Ct. 2166, 100 L.Ed.2d 811 (1988). For cases falling within the patent-specific arising under jurisdiction of § 1338(a), however, Congress has not only provided for federal jurisdiction but also eliminated state jurisdiction, decreeing that "[n]o State court shall have jurisdiction over any claim for relief arising under any Act of Congress relating to patents." § 1338(a) (2006 ed., Supp. V). To determine whether jurisdiction was proper in the Texas courts, therefore, we must determine whether it would have been proper in a federal district court — whether, that is, the case "aris[es] under any Act of Congress relating to patents."

For statutory purposes, a case can "aris[e] under" federal law in two ways. Most directly, a case arises under federal law when federal law creates the cause of action asserted. See American Well Works Co. v. Layne & Bowler Co., 241 U.S. 257, 260, 36 S.Ct. 585, 60 L.Ed. 987 (1916) ("A suit arises under the law that creates the cause of action"). As a rule of inclusion, this "creation" test admits of only extremely rare exceptions, see, e.g., Shoshone Mining Co. v. Rutter, 177 U.S. 505, 20 S.Ct. 726, 44 L.Ed. 864 (1900), and accounts for the vast bulk of suits that arise under federal law, see Franchise Tax Bd. of Cal. v. Construction Laborers Vacation Trust for Southern Cal., 463 U.S. 1, 9, 103 S.Ct. 2841, 77 L.Ed.2d 420 (1983). Minton's original patent infringement suit against NASD and NASDAQ, for example, arose under federal law in this manner because it was authorized by 35 U.S.C. §§ 271, 281.

But even where a claim finds its origins in state rather than federal law — as Minton's legal malpractice claim indisputably does — we have identified a "special and small category" of cases in which arising under jurisdiction still lies. Empire Healthchoice Assurance, Inc. v. McVeigh, [1065] 547 U.S. 677, 699, 126 S.Ct. 2121, 165 L.Ed.2d 131 (2006). In outlining the contours of this slim category, we do not paint on a blank canvas. Unfortunately, the canvas looks like one that Jackson Pollock got to first. See 13D C. Wright, A. Miller, E. Cooper, & R. Freer, Federal Practice and Procedure § 3562, pp. 175-176 (3d ed. 2008) (reviewing general confusion on question).

In an effort to bring some order to this unruly doctrine several Terms ago, we condensed our prior cases into the following inquiry: Does the "state-law claim necessarily raise a stated federal issue, actually disputed and substantial, which a federal forum may entertain without disturbing any congressionally approved balance of federal and state judicial responsibilities"? Grable, 545 U.S., at 314, 125 S.Ct. 2363. That is, federal jurisdiction over a state law claim will lie if a federal issue is: (1) necessarily raised, (2) actually disputed, (3) substantial, and (4) capable of resolution in federal court without disrupting the federal-state balance approved by Congress. Where all four of these requirements are met, we held, jurisdiction is proper because there is a "serious federal interest in claiming the advantages thought to be inherent in a federal forum," which can be vindicated without disrupting Congress's intended division of labor between state and federal courts. Id., at 313-314, 125 S.Ct. 2363.

III

Applying Grable's inquiry here, it is clear that Minton's legal malpractice claim does not arise under federal patent law. Indeed, for the reasons we discuss, we are comfortable concluding that state legal malpractice claims based on underlying patent matters will rarely, if ever, arise under federal patent law for purposes of § 1338(a). Although such cases may necessarily raise disputed questions of patent law, those cases are by their nature unlikely to have the sort of significance for the federal system necessary to establish jurisdiction.

A

To begin, we acknowledge that resolution of a federal patent question is "necessary" to Minton's case. Under Texas law, a plaintiff alleging legal malpractice must establish four elements: (1) that the defendant attorney owed the plaintiff a duty; (2) that the attorney breached that duty; (3) that the breach was the proximate cause of the plaintiff's injury; and (4) that damages occurred. See Alexander v. Turtur & Associates, Inc., 146 S.W.3d 113, 117 (Tex.2004). In cases like this one, in which the attorney's alleged error came in failing to make a particular argument, the causation element requires a "case within a case" analysis of whether, had the argument been made, the outcome of the earlier litigation would have been different. 355 S.W.3d, at 639; see 4 R. Mallen & J. Smith, Legal Malpractice § 37:15, pp. 1509-1520 (2012). To prevail on his legal malpractice claim, therefore, Minton must show that he would have prevailed in his federal patent infringement case if only petitioners had timely made an experimental-use argument on his behalf. 355 S.W.3d, at 644. That will necessarily require application of patent law to the facts of Minton's case.

B

The federal issue is also "actually disputed" here — indeed, on the merits, it is the central point of dispute. Minton argues that the experimental-use exception properly applied to his lease to Stark, saving his patent from the on-sale bar; petitioners argue that it did not. This is just the sort of "`dispute ... respecting the [1066] ... effect of [federal] law'" that Grable envisioned. 545 U.S., at 313, 125 S.Ct. 2363 (quoting Shulthis v. McDougal, 225 U.S. 561, 569, 32 S.Ct. 704, 56 L.Ed. 1205 (1912)).

C

Minton's argument founders on Grable's next requirement, however, for the federal issue in this case is not substantial in the relevant sense. In reaching the opposite conclusion, the Supreme Court of Texas focused on the importance of the issue to the plaintiff's case and to the parties before it. 355 S.W.3d, at 644 ("because the success of Minton's malpractice claim is reliant upon the viability of the experimental use exception as a defense to the on-sale bar, we hold that it is a substantial federal issue"); see also Air Measurement Technologies, 504 F.3d, at 1272 ("the issue is substantial, for it is a necessary element of the malpractice case"). As our past cases show, however, it is not enough that the federal issue be significant to the particular parties in the immediate suit; that will always be true when the state claim "necessarily raise[s]" a disputed federal issue, as Grable separately requires. The substantiality inquiry under Grable looks instead to the importance of the issue to the federal system as a whole.

In Grable itself, for example, the Internal Revenue Service had seized property from the plaintiff and sold it to satisfy the plaintiff's federal tax delinquency. 545 U.S., at 310-311, 125 S.Ct. 2363. Five years later, the plaintiff filed a state law quiet title action against the third party that had purchased the property, alleging that the IRS had failed to comply with certain federally imposed notice requirements, so that the seizure and sale were invalid. Ibid. In holding that the case arose under federal law, we primarily focused not on the interests of the litigants themselves, but rather on the broader significance of the notice question for the Federal Government. We emphasized the Government's "strong interest" in being able to recover delinquent taxes through seizure and sale of property, which in turn "require[d] clear terms of notice to allow buyers ... to satisfy themselves that the Service has touched the bases necessary for good title." Id., at 315, 125 S.Ct. 2363. The Government's "direct interest in the availability of a federal forum to vindicate its own administrative action" made the question "an important issue of federal law that sensibly belong[ed] in a federal court." Ibid.

A second illustration of the sort of substantiality we require comes from Smith v. Kansas City Title & Trust Co., 255 U.S. 180, 41 S.Ct. 243, 65 L.Ed. 577 (1921), which Grable described as "[t]he classic example" of a state claim arising under federal law. 545 U.S., at 312, 125 S.Ct. 2363. In Smith, the plaintiff argued that the defendant bank could not purchase certain bonds issued by the Federal Government because the Government had acted unconstitutionally in issuing them. 255 U.S., at 198, 41 S.Ct. 243. We held that the case arose under federal law, because the "decision depends upon the determination" of "the constitutional validity of an act of Congress which is directly drawn in question." Id., at 201, 41 S.Ct. 243. Again, the relevant point was not the importance of the question to the parties alone but rather the importance more generally of a determination that the Government "securities were issued under an unconstitutional law, and hence of no validity." Ibid.; see also Merrell Dow Pharmaceuticals Inc. v. Thompson, 478 U.S. 804, 814, n. 12, 106 S.Ct. 3229, 92 L.Ed.2d 650 (1986).

Here, the federal issue carries no such significance. Because of the backward-looking [1067] nature of a legal malpractice claim, the question is posed in a merely hypothetical sense: If Minton's lawyers had raised a timely experimental-use argument, would the result in the patent infringement proceeding have been different? No matter how the state courts resolve that hypothetical "case within a case," it will not change the real-world result of the prior federal patent litigation. Minton's patent will remain invalid.

Nor will allowing state courts to resolve these cases undermine "the development of a uniform body of [patent] law." Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 162, 109 S.Ct. 971, 103 L.Ed.2d 118 (1989). Congress ensured such uniformity by vesting exclusive jurisdiction over actual patent cases in the federal district courts and exclusive appellate jurisdiction in the Federal Circuit. See 28 U.S.C. §§ 1338(a), 1295(a)(1). In resolving the nonhypothetical patent questions those cases present, the federal courts are of course not bound by state court case-within-a-case patent rulings. See Tafflin v. Levitt, 493 U.S. 455, 465, 110 S.Ct. 792, 107 L.Ed.2d 887 (1990). In any event, the state court case-within-a-case inquiry asks what would have happened in the prior federal proceeding if a particular argument had been made. In answering that question, state courts can be expected to hew closely to the pertinent federal precedents. It is those precedents, after all, that would have applied had the argument been made. Cf. ibid. ("State courts adjudicating civil RICO claims will ... be guided by federal court interpretations of the relevant federal criminal statutes, just as federal courts sitting in diversity are guided by state court interpretations of state law").

As for more novel questions of patent law that may arise for the first time in a state court "case within a case," they will at some point be decided by a federal court in the context of an actual patent case, with review in the Federal Circuit. If the question arises frequently, it will soon be resolved within the federal system, laying to rest any contrary state court precedent; if it does not arise frequently, it is unlikely to implicate substantial federal interests. The present case is "poles apart from Grable," in which a state court's resolution of the federal question "would be controlling in numerous other cases." Empire Healthchoice Assurance, Inc., 547 U.S., at 700, 126 S.Ct. 2121.

Minton also suggests that state courts' answers to hypothetical patent questions can sometimes have real-world effect on other patents through issue preclusion. Brief for Respondent 33-36. Minton, for example, has filed what is known as a "continuation patent" application related to his original patent. See 35 U.S.C. § 120; 4A D. Chisum, Patents § 13.03 (2005) (describing continuation applications). He argues that, in evaluating this separate application, the patent examiner could be bound by the Texas trial court's interpretation of the scope of Minton's original patent. See Brief for Respondent 35-36. It is unclear whether this is true. The Patent and Trademark Office's Manual of Patent Examining Procedure provides that res judicata is a proper ground for rejecting a patent "only when the earlier decision was a decision of the Board of Appeals" or certain federal reviewing courts, giving no indication that state court decisions would have preclusive effect. See Dept. of Commerce, Patent and Trademark Office, Manual of Patent Examining Procedure § 706.03(w), p. 700-79 (rev. 8th ed. 2012); 35 U.S.C. §§ 134(a), 141, 145; Reply Brief 9-10. In fact, Minton has not identified any case finding such preclusive effect based on a state court decision. But even assuming that a state court's case-within-a-case adjudication may be preclusive under [1068] some circumstances, the result would be limited to the parties and patents that had been before the state court. Such "fact-bound and situation-specific" effects are not sufficient to establish federal arising under jurisdiction. Empire Healthchoice Assurance, Inc., supra, at 701, 126 S.Ct. 2121.

Nor can we accept the suggestion that the federal courts' greater familiarity with patent law means that legal malpractice cases like this one belong in federal court. See Air Measurement Technologies, 504 F.3d, at 1272 ("The litigants will also benefit from federal judges who have experience in claim construction and infringement matters"); 355 S.W.3d, at 646 ("patent litigants have an interest in the uniform application of patent law by courts well-versed in that subject matter"). It is true that a similar interest was among those we considered in Grable, 545 U.S., at 314, 125 S.Ct. 2363. But the possibility that a state court will incorrectly resolve a state claim is not, by itself, enough to trigger the federal courts' exclusive patent jurisdiction, even if the potential error finds its root in a misunderstanding of patent law.

There is no doubt that resolution of a patent issue in the context of a state legal malpractice action can be vitally important to the particular parties in that case. But something more, demonstrating that the question is significant to the federal system as a whole, is needed. That is missing here.

D

It follows from the foregoing that Grable's fourth requirement is also not met. That requirement is concerned with the appropriate "balance of federal and state judicial responsibilities." Ibid. We have already explained the absence of a substantial federal issue within the meaning of Grable. The States, on the other hand, have "a special responsibility for maintaining standards among members of the licensed professions." Ohralik v. Ohio State Bar Assn., 436 U.S. 447, 460, 98 S.Ct. 1912, 56 L.Ed.2d 444 (1978). Their "interest ... in regulating lawyers is especially great since lawyers are essential to the primary governmental function of administering justice, and have historically been officers of the courts." Goldfarb v. Virginia State Bar, 421 U.S. 773, 792, 95 S.Ct. 2004, 44 L.Ed.2d 572 (1975) (internal quotation marks omitted). We have no reason to suppose that Congress — in establishing exclusive federal jurisdiction over patent cases — meant to bar from state courts state legal malpractice claims simply because they require resolution of a hypothetical patent issue.

* * *

As we recognized a century ago, "[t]he Federal courts have exclusive jurisdiction of all cases arising under the patent laws, but not of all questions in which a patent may be the subject-matter of the controversy." New Marshall Engine Co. v. Marshall Engine Co., 223 U.S. 473, 478, 32 S.Ct. 238, 56 L.Ed. 513 (1912). In this case, although the state courts must answer a question of patent law to resolve Minton's legal malpractice claim, their answer will have no broader effects. It will not stand as binding precedent for any future patent claim; it will not even affect the validity of Minton's patent. Accordingly, there is no "serious federal interest in claiming the advantages thought to be inherent in a federal forum," Grable, supra, at 313, 125 S.Ct. 2363. Section 1338(a) does not deprive the state courts of subject matter jurisdiction.

The judgment of the Supreme Court of Texas is reversed, and the case is remanded [1069] for further proceedings not inconsistent with this opinion.

It is so ordered.

CERTIORARI TO THE CIRCUIT COURT OF APPEALS FOR THE SEVENTH CIRCUIT.

[206] Mr. Edward Davidson, with whom Mr. John P. Bramhall was on the brief, for petitioner.

Mr. Albert H. Cole for respondent.

[207] MR. CHIEF JUSTICE HUGHES delivered the opinion of the Court.

Petitioner brought this action in the District Court of the United States for the Northern District of Indiana, Fort Wayne Division, to recover for injuries which he sustained on November 29, 1930, in the course of his employment by respondent, an interstate carrier, in its yard [208] at Russell, Kentucky. In his complaint he set forth two "paragraphs" or counts, both being for the same injuries. In the first paragraph, petitioner alleged that at the time of the injuries he was employed in interstate commerce and that he brought the action under the Acts of Congress known as the Federal Employers' Liability Act^[1] and the Safety Appliance Acts,^[2] and the rules and orders which the Interstate Commerce Commission had promulgated under the latter.^[3] In the second paragraph, he alleged that at the time of the injuries he was employed in intrastate commerce and he invoked the Safety Appliance Acts enacted by the Congress, and the rules and orders of the Interstate Commerce Commission thereunder, and the Employers' Liability Act of Kentucky. The provisions of the laws of Kentucky which were alleged to govern the rights of the parties at the time and place in question were set forth.^[4] In each count petitioner stated that the injuries were received while he was engaged as a switchman in attempting to uncouple certain freight cars and were due to a defective uncoupling lever.

Objections to the jurisdiction of the District Court as to each count were raised by plea in abatement. They were overruled and petitioner had a general verdict. The judgment, entered accordingly, was reversed by the Circuit Court of Appeals upon the ground that the District Court was without jurisdiction to entertain the case upon [209] either count. 64 F. (2d) 472. This Court granted certiorari.

Distinct questions are presented with respect to each count and they will be considered separately.

First. By the first paragraph, the jurisdiction of the Federal court was rested upon the sole ground that the injury had been sustained during petitioner's employment in interstate commerce and that the cause of action arose under the pertinent Federal legislation. To support the jurisdiction of the District Court for the Northern District of Indiana, the complaint alleged that respondent was engaged in business in that district at the time of the commencement of the action. Respondent's challenge to the jurisdiction was upon the grounds (1) that at the time of the injuries petitioner was not employed in interstate commerce and hence the action would not lie under the Federal Employers' Liability Act, and (2) that respondent was a corporation organized under the laws of Virginia and an inhabitant of the Eastern District of Virginia, and hence, so far as the action rested upon the Safety Appliance Acts of Congress, and the rules and orders of the Interstate Commerce Commission, it could not be brought in a Federal court in any district other than the Eastern District of Virginia. Jud. Code, § 51; 28 U.S.C. § 112.

Petitioner's demurrer to the plea in abatement as to the first cause of action was sustained by the trial court. That court pointed out that the plea did not deny that respondent was doing business within the Northern District of Indiana and that the pleading, in substance, went to the merits. The Circuit Court of Appeals took a different view, holding that so far as petitioner relied upon a violation of the Safety Appliance Acts, the action must be brought in the district of respondent's residence. In reversing the judgment, the Circuit Court of Appeals remanded [210] the cause with instructions to grant permission to petitioner to amend his first paragraph to conform exclusively to the theory of a violation of the Federal Employers' Liability Act.

This ruling of the appellate court cannot be sustained. The jurisdiction of the District Court is to be determined by the allegations of the complaint. Mosher v. Phoenix, 287 U.S. 29, 30; Levering & Garrigues Co. v. Morrin, 289 U.S. 103, 105. These allegations clearly set forth, in the first paragraph, a cause of action under the Federal Employers' Liability Act. Every essential ingredient of such a cause of action was appropriately alleged. The Federal Employers' Liability Act expressly recognized that in an action brought under its provisions the question of a violation of the Safety Appliance Acts might be presented and determined. This is the unmistakable effect of the provisions that, in such an action, the employee shall not be held "to have been guilty of contributory negligence," or "to have assumed the risks of his employment" in any case "where the violation by such common carrier of any statute enacted for the safety of employees contributed to the injury or death of such employee." Act of April 22, 1908, §§ 3, 4, 45 U.S.C. §§ 53, 54. By the phrase "any statute enacted for the safety of employees" the Congress evidently intended to embrace its Safety Appliance Acts. Seaboard Air Line Ry. v. Horton, 233 U.S. 492, 503. This Court has said that the statutes are in pari materia and that "where the Employers' Liability Act refers to `any defect or insufficiency, due to its negligence, in its cars, engines, appliances,' etc., it clearly is the legislative intent to treat a violation of the Safety Appliance Act as `negligence' — what is sometimes called negligence per se." San Antonio & Aransas Pass Ry. Co. v. Wagner, 241 U.S. 476, 484. Where an employee of an interstate carrier sustains injuries while employed in the interstate commerce of the carrier, his action [211] may thus be brought under the Federal Employers' Liability Act in connection with the Safety Appliance Acts.^[5]

Under the Federal Employers' Liability Act an action may be brought "in a District Court of the United States, in the district of the residence of the defendant, or in which the cause of action arose, or in which the defendant shall be doing business at the time of commencing such action." 45 U.S.C. § 56. It follows that, upon the allegations of the complaint, the action on the claim set forth in the first paragraph was properly brought in the District Court for the Northern District of Indiana where respondent was doing business when the action was begun.

Second. In the second paragraph of the complaint, which treated the injuries as received in intrastate commerce, diversity of citizenship was alleged; that petitioner was a citizen of Indiana, and a resident of the city of Fort Wayne in that State, and that respondent was a citizen of Virginia doing business in Indiana. The plea in abatement, admitting respondent's citizenship in Virginia, denied that petitioner was a resident of Fort Wayne or of the Northern District of Indiana, or was a citizen of that State, and alleged that as the cause of action set forth in the second paragraph arose under the Federal Safety Appliance Acts, the action could not be brought [212] in any district other than the Eastern District of Virginia. The District Court took evidence on the issue of fact, found that the petitioner was a citizen of Indiana and a resident of Fort Wayne, and overruled the plea. The Circuit Court of Appeals held that the District Court of the Northern District of Indiana was without jurisdiction, in the view that the second count attempted to set forth a cause of action "under the Federal Safety Appliance Act as well as under the statutes of Kentucky" and hence that jurisdiction did not rest solely on diversity of citizenship. Jud. Code, § 51, 28 U.S.C. § 112. In remanding the cause, the Circuit Court of Appeals directed that petitioner be allowed to amend the second paragraph of his complaint so as to conform exclusively to the theory of a violation of the Kentucky statute.

While invoking, in the second count, the Safety Appliance Acts, petitioner fully set forth and relied upon the laws of the State of Kentucky where the cause of action arose. In relation to injuries received in that State in intrastate commerce, aside from the particular bearing of the Federal Safety Appliance Acts, the liability of respondent was determined by the laws of Kentucky. Slater v. Mexican National R. Co., 194 U.S. 120, 126; Cuba R. Co. v. Crosby, 222 U.S. 473, 478; Young v. Masci, 289 U.S. 253, 258; Ormsby v. Chase, 290 U.S. 387. The statute of Kentucky, in prescribing the liability of common carriers for negligence causing injuries to employees while engaged in intrastate commerce, reproduced in substance, and with almost literal exactness, the corresponding provisions of the Federal Employers' Liability Act as to injuries received in interstate commerce. Ky. Acts, 1918, c. 52, §§ 1-3, p. 153; Carroll's Ky. Statutes, 1930, §§ 820 b-1, 820 b-2, 820 b-3. The Kentucky Act provided that no employee should be held "to have been guilty of contributory negligence" or "to have assumed the risk of his employment" in any case "where the violation by [213] such common carrier of any statute, state or federal, enacted for the safety of employees contributed to the injury or death of such employee." Id. The Kentucky legislature read into its statute the provisions of statutes both state and federal which were enacted for the safety of employees, and the Federal Safety Appliance Acts were manifestly embraced in this description. Louisville & Nashville R. Co. v. Layton, 243 U.S. 617, 619. Thus, the second count of the complaint, in invoking the Federal Safety Appliance Acts, while declaring on the Kentucky Employers' Liability Act, cannot be regarded as setting up a claim which lay outside the purview of the state statute. As in the analogous case under the Federal Employers' Liability Act, a violation of the acts for the safety of employees was to constitute negligence per se in applying the state statute and was to furnish the ground for precluding the defense of contributory negligence as well as that of assumption of risk.

The Circuit Court of Appeals took the view that if it were assumed that the second count was based exclusively upon the Kentucky statute, that statute and the federal requirements could not be considered as being in pari materia because the latter applied only to interstate commerce, and that, if the petitioner were permitted to establish the negligence required by the state statute by showing the violation of the federal requirements the court would thereby be placed "in the anomalous position of extending the benefits of the Safety Appliance Act to intrastate commerce."

This is an erroneous view. The original Safety Appliance Act of March 2, 1893, 27 Stat. 531, did not embrace all cars on the lines of interstate carriers but only those engaged in interstate commerce. Brinkmeier v. Missouri Pacific Ry. Co., 224 U.S. 268. By the amending Act of March 2, 1903, 32 Stat. 943, the scope of the statute was enlarged so as to include all cars "used on any railroad [214] engaged in interstate commerce." The statute as amended was intended to embrace all locomotives, cars, and similar vehicles used on any railroad which is a highway of interstate commerce. Southern Ry. Co. v. United States, 222 U.S. 20. With respect to such vehicles, the duty to protect employees by the prescribed safety appliances exists even though the vehicles and the employee injured through the failure to provide such protection are at the time engaged in intrastate commerce. Texas & Pacific Ry. Co. v. Rigsby, 241 U.S. 33; Louisville & Nashville R. Co. v. Layton, supra. The Federal Act in its application to such a case is thus in pari materia with the statute of Kentucky which prescribes the liability of carriers for injuries to employees while employed in intrastate commerce and which, in effect, reads into the provisions of the statute the requirements of the Federal Act for the safety of employees. There appears to be no anomaly in enforcing the state law with this defined content.

The Federal Safety Appliance Acts prescribed duties, and injured employees are entitled to recover for injuries sustained through the breach of these duties. Johnson v. Southern Pacific Co., 196 U.S. 1; St. Louis, I.M. & S. Ry. Co. v. Taylor, 210 U.S. 281; Texas & Pacific Ry. Co. v. Rigsby, supra. Questions arising in actions in state courts to recover for injuries sustained by employees in intrastate commerce and relating to the scope or construction of the Federal Safety Appliance Acts are, of course, federal questions which may appropriately be reviewed in this Court. St. Louis, I.M. & S. Ry. Co. v. Taylor, supra; Louisville & Nashville R. Co. v. Layton, supra. But it does not follow that a suit brought under the state statute which defines liability to employees who are injured while engaged in intrastate commerce, and brings within the purview of the statute a breach of the duty imposed by the federal statute, should be regarded as a suit arising under the laws of the United States and cognizable in the [215] federal court in the absence of diversity of citizenship. The Federal Safety Appliance Acts, while prescribing absolute duties, and thus creating correlative rights in favor of injured employees, did not attempt to lay down rules governing actions for enforcing these rights. The original Act of 1893 made no provision for suits, except for penalties. That Act did impliedly recognize the employee's right of action by providing in § 8 that he should not be deemed to have assumed the risk of injury occasioned by the breach of duty. But the Act made no provision as to the place of suit or the time within which it should be brought, or as to the right to recover, or as to those who should be the beneficiaries of recovery, in case of the death of the employee. While dealing with assumption of risk, the statute did not affect the defense of contributory negligence and hence that defense was still available according to the applicable state law. Schlemmer v. Buffalo, R. & P. Ry. Co., 220 U.S. 590; Minneapolis, St. P. & S.S.M. Ry. Co. v. Popplar, 237 U.S. 369, 371, 372. In these respects the amended Act of 1903 made no change, notwithstanding the enlargement of the scope of the statutory requirements. The Act of 1910, by a proviso in § 4 relating to penalties (36 Stat. 299), provided that nothing in that section should "be construed to relieve such carrier from liability in any remedial action for the death or injury of any railroad employee" caused by the use of the prohibited equipment.

The Safety Appliance Acts having prescribed the duty in this fashion, the right to recover damages sustained by the injured employee through the breach of duty sprang from the principle of the common law (Texas & Pacific R. Co. v. Rigsby, supra, at pp. 39, 40^[6]) and was left to be enforced accordingly, or, in case of the death of [216] the injured employee, according to the applicable statute.^[7]St. Louis, I.M. & S. Ry. Co. v. Taylor, supra, at p. 285; Minneapolis, St. P. & S.S.M. Ry. Co. v. Popplar, supra. When the Federal Employers' Liability Act was enacted, it drew to itself the right of action for injuries or death of the employees within its purview who were engaged in interstate commerce, including those cases in which injuries were due to a violation of the Safety Appliance Acts. Such an action must be brought as prescribed in the Federal Employers' Liability Act, and if brought in the state court, it cannot be removed to the federal court, although violation of the Safety Appliance Acts is involved. See St. Joseph & G.I. Ry. Co. v. Moore, 243 U.S. 311. With respect to injuries sustained in intrastate commerce, nothing in the Safety Appliance Acts precluded the State from incorporating in its legislation applicable to local transportation the paramount duty which the Safety Appliance Acts imposed as to the equipment of cars used on interstate railroads. As this Court said in Minneapolis, St. P. & S.S.M. Ry. Co. v. Popplar, supra, as to an action for injuries sustained in intrastate commerce: "The action fell within the familiar category of cases involving the duty of a master to his servant. This duty is defined by the common law, except as it may be modified by legislation. The federal statute, in the present case, touched the duty of the master at a single point and, save as provided in the statute, the right of the [217] plaintiff to recover was left to be determined by the law of the State."

We are of the opinion that the second paragraph of the complaint set forth a cause of action under the Kentucky statute and, as to this cause of action, the suit is not to be regarded as one arising under the laws of the United States. In view of the diversity of citizenship and the residence of petitioner, the District Court of the Northern District of Indiana had jurisdiction.

As the Circuit Court of Appeals did not consider any questions save those relating to the jurisdiction of the District Court, the judgment of the Circuit Court of Appeals will be reversed and the cause remanded to that court with directions to consider such other questions as may be presented by the appeal.

Reversed.

[1] Act of April 22, 1908, c. 149, 35 Stat. 65, 45 U.S.C. §§ 51 et seq.; Act of April 5, 1910, c. 143, 36 Stat. 291, 45 U.S.C. § 56.

[2] Acts of March 2, 1893, c. 196, 27 Stat. 531, 45 U.S.C. §§ 1 et seq.; April 1, 1896, c. 87, 29 Stat. 85, 45 U.S.C. § 6; March 2, 1903, c. 976, 32 Stat. 943, 45 U.S.C. §§ 8, 9, 10; April 14, 1910, c. 160, 36 Stat. 298, 45 U.S.C. §§ 11 et seq.

[3] Order of March 13, 1911; Roberts' Federal Liabilities of Carriers, Vol. 2, pp. 2010, 2016.

[4] Ky. Acts, 1918, c. 52, §§ 1-3, p. 153; Carroll's Ky. Statutes, §§ 820 b-1, 820 b-2, 820 b-3.

[5] See Southern Ry. Co. v. Crockett, 234 U.S. 725, 727; St. Louis & San Francisco R. Co. v. Conarty, 238 U.S. 243, 248; Great Northern Ry. Co. v. Otos, 239 U.S. 349, 350; San Antonio & Aransas Pass Ry. Co. v. Wagner, 241 U.S. 476, 484; Spokane & I.E.R. Co. v. Campbell, 241 U.S. 497, 498; Atlantic City R. Co. v. Parker, 242 U.S. 56, 58; St. Joseph & G.I. Ry. Co. v. Moore, 243 U.S. 311, 312; Minneapolis & St. Louis R. Co. v. Gotschall, 244 U.S. 66; Great Northern Ry. Co. v. Donaldson, 246 U.S. 121, 124; Davis v. Wolfe, 263 U.S. 239, 240; Baltimore & Ohio R. Co. v. Groeger, 266 U.S. 521, 528; Chicago Great Western R. Co. v. Schendel, 267 U.S. 287, 289; Minneapolis, St. P. & S.S.M. Ry. Co. v. Goneau, 269 U.S. 406, 407.

[6] In Texas & Pacific R. Co. v. Rigsby, 241 U.S. 33, the action was brought in the state court and was removed to the federal court upon the ground that the defendant was a federal corporation.

[7] In St. Louis, I.M. & S. Ry. Co. v. Taylor, 210 U.S. 281, 285, the Court said: "The accident by which the plaintiff's intestate lost his life occurred in the Indian Territory, where, contrary to the doctrine of the common law, a right of action for death exists. The cause of action arose under the laws of the Territory, and was enforced in the courts of Arkansas." The question whether the action was triable in those courts was held not to present a federal question, but the question as to the interpretation of the Safety Appliance Act of 1893 did present the federal question which was reviewed by this Court.

APPEAL FROM THE DISTRICT COURT OF THE UNITED STATES FOR THE WESTERN DISTRICT OF MISSOURI.

[181] Mr. William Marshall Bullitt for appellant.

Mr. Frank Hagerman for appellant.

Mr. Charles E. Hughes for Federal Land Bank of Wichita, Kansas, appellee.

Mr. W.W. Willoughby filed a separate brief on behalf of the Federal Land Bank of Wichita, Kansas, appellee.

Mr. George W. Wickersham, with whom Mr. W.G. McAdoo was on the briefs, for First Joint Stock Land Bank of Chicago, appellee.

Mr. Justin D. Bowersock filed a brief on behalf of Kansas City Title & Trust Company, appellee.

The Solicitor General, Mr. W.G. McAdoo, Special Assistant to the Attorney General, and Mr. J.P. Cotton, by leave of court, filed a brief on behalf of the United States as amici curiae.^[1]

Restored to docket for reargument April 26, 1920.

[195] MR. JUSTICE DAY delivered the opinion of the court.

A bill was filed in the United States District Court for the Western Division of the Western District of Missouri by a shareholder in the Kansas City Title & Trust Company to enjoin the Company, its officers, agents and employees from investing the funds of the Company in farm loan bonds issued by Federal Land Banks or Joint Stock Land Banks under authority of the Federal Farm Loan Act of July 17, 1916, c. 245, 39 Stat. 360, as amended January 18, 1918, c. 9, 40 Stat. 431.

The relief was sought on the ground that these acts were beyond the constitutional power of Congress. The bill avers that the Board of Directors of the Company are [196] about to invest its funds in the bonds to the amount of $10,000 in each of the classes described, and will do so unless enjoined by the court in this action. The bill avers the formation of twelve Federal Land Banks, and twenty-one Joint Stock Land Banks under the provisions of the act.

As to the Federal Land Banks, it is averred that each of them has loaned upon farm lands large amounts secured by mortgage, and, after depositing the same with the Farm Loan Registrar, has executed and issued collateral trust obligations called Farm Loan Bonds, secured by the depositing of an equivalent amount of farm mortgages and notes; and that each of said Federal Land Banks has sold, and is continuing to offer for sale, large amounts of said Farm Loan Bonds. The bill also avers that various persons in different parts of the United States have organized twenty-one Joint Stock Land Banks, the capital stock of which is subscribed for and owned by private persons; that the Joint Stock Land Banks have deposited notes and mortgages with the Farm Loan Registrar, and issued an equivalent amount of collateral trust obligations called Farm Loan Bonds, which have been sold and will be continued to be offered for sale to investors in large amounts in the markets of the country. A statement is given of the amount of deposits by the Secretary of the Treasury with the Federal Land Banks, for which the banks have issued their certificates of indebtedness bearing interest at 2% per annum. It is averred that on September 30, 1919, Federal Land Banks owned United States bonds of the par value of $4,230,805; and the Joint Stock Land Banks owned like bonds of the par value of $3,287,503 on August 31, 1919; that pursuant to the provisions of the act the Secretary of the Treasury has invested $8,892,130 of the public funds in the capital stock of the Federal Land Banks, and that on July 1, 1919, the Secretary of the Treasury on behalf of the United States held $8,265,809 of the capital stock of the Federal Land Banks; [197] that pursuant to the provisions of § 32 of the act, as amended, the Secretary of the Treasury has purchased Farm Loan Bonds issued by the Federal Land Banks of the par value of $149,775,000; that up to September 30, 1919, bonds have been issued under the act by the Federal Land Banks to the amount of $285,600,000, of which about $135,000,000 are held in the Treasury of the United States, purchased under the authority of the amendment of January 18, 1918; that up to September 30, 1919, twenty-seven Joint Stock Land Banks have been incorporated under the act, having an aggregate capital of $8,000,000, all of which has been subscribed and $7,450,000 paid in; that bonds have been issued by Joint Stock Land Banks to the amount of $41,000,000, which are now in the hands of the public; that the Secretary of the Treasury up to the time of the filing of the bill has not designated any of the Federal Land Banks nor the Joint Stock Land Banks as depositaries of public money, nor, except as stated later in the bill, has he employed them or any of them as financial agents of the Government, nor have they or any of them performed any duties as depositaries of public money, nor have they or any of them accepted any deposits or engaged in any banking business. The bill avers that during the summer of 1918 the Federal Land Banks at Wichita, St. Paul and Spokane were designated as financial agents of the Government for making seed grain loans to farmers in drought-stricken sections, the President having at the request of the Secretary of Agriculture set aside $5,000,000 for that purpose out of the $100,000,000 war funds. The three banks mentioned made upwards of 15,000 loans of that character, aggregating a sum upwards of $4,500,000, and are now engaged in collecting these loans, all of which are secured by crop liens; that these banks act in that capacity without compensation, receiving only the actual expenses incurred.

Section 27 of the act provides that Farm Loan Bonds [198] issued under the provisions of the act by Federal Land Banks or Joint Stock Land Banks shall be a lawful investment for all fiduciary and trust funds, and may be accepted as security for all public deposits. The bill avers that the defendant Trust Company is authorized to buy, invest in and sell government, state and municipal and other bonds, but it cannot buy, invest in or sell any such bonds, papers, stocks or securities which are not authorized to be issued by a valid law or which are not investment securities, but that nevertheless it is about to invest in Farm Loan Bonds; that the Trust Company has been induced to direct its officers to make the investment by reason of its reliance upon the provisions of the Farm Loan Acts, especially §§ 21, 26 and 27, by which the Farm Loan Bonds are declared to be instrumentalities of the Government of the United States, and as such with the income derived therefrom, are declared to be exempt from federal, state, municipal and local taxation, and are further declared to be lawful investments for all fiduciary and trust funds. The bill further avers that the acts by which it is attempted to authorize the bonds are wholly illegal, void and unconstitutional and of no effect because unauthorized by the Constitution of the United States.

The bill prays that the acts of Congress authorizing the creation of the banks, especially §§ 21, 26 and 27 thereof, shall be adjudged and decreed to be unconstitutional, void and of no effect, and that the issuance of the Farm Loan Bonds, and the taxation exemption feature thereof, shall be adjudged and decreed to be invalid.

The First Joint Stock Land Bank of Chicago and the Federal Land Bank of Wichita, Kansas, were allowed to intervene and became parties defendant to the suit. The Kansas City Title & Trust Company filed a motion to dismiss in the nature of a general demurrer, and upon hearing the District Court entered a decree dismissing the bill. From this decree appeal was taken to this court.

[199] No objection is made to the federal jurisdiction, either original or appellate, by the parties to this suit, but that question will be first examined. The Company is authorized to invest its funds in legal securities only. The attack upon the proposed investment in the bonds described is because of the alleged unconstitutionality of the acts of Congress undertaking to organize the banks and authorize the issue of the bonds. No other reason is set forth in the bill as a ground of objection to the proposed investment by the Board of Directors acting in the Company's behalf. As diversity of citizenship is lacking, the jurisdiction of the District Court depends upon whether the cause of action set forth arises under the Constitution or laws of the United States. Judicial Code, § 24.

The general rule is that where it appears from the bill or statement of the plaintiff that the right to relief depends upon the construction or application of the Constitution or laws of the United States, and that such federal claim is not merely colorable, and rests upon a reasonable foundation, the District Court has jurisdiction under this provision.

At an early date, considering the grant of constitutional power to confer jurisdiction upon the federal courts, Chief Justice Marshall said: "A case in law or equity consists of the right of the one party, as well as of the other, and may truly be said to arise under the Constitution or a law of the United States, whenever its correct decision depends on the construction of either," Cohens v. Virginia, 6 Wheat. 264, 379; and again, when "the title or right set up by the party may be defeated by one construction of the Constitution or law of the United States, and sustained by the opposite construction." Osborn v. Bank of the United States, 9 Wheat. 738, 822. These definitions were quoted and approved in Patton v. Brady, 184 U.S. 608, 611, citing Gold-Washing Co. v. Keyes, 96 U.S. 199, 201; Tennessee v. Davis, 100 U.S. 257; White v. Greenhow, 114 U.S. 307; Railroad Company v. Mississippi, 102 U.S. 135, 139.

[200] This characterization of a suit arising under the Constitution or laws of the United States has been followed in many decisions of this and other federal courts. See Macon Grocery Co. v. Atlantic Coast Line R.R. Co., 215 U.S. 501 506, 507; Shulthis v. McDougal, 225 U.S. 561, 569, paragraph 3. The principle was applied in Brushaber v. Union Pacific R.R. Co., 240 U.S. 1, in which a shareholder filed a bill to enjoin the defendant corporation from complying with the income tax provisions of the Tariff Act of October 3, 1913. In that case while there was diversity of citizenship, a direct appeal to this court was sustained because of the constitutional questions raised in the bill, which had been dismissed by the court below. The repugnancy of the statute to the Constitution of the United States, as well as grounds of equitable jurisdiction, were set forth in the bill, and the right to come here on direct appeal was sustained because of the averments based upon constitutional objections to the act. Reference was made to Pollock v. Farmers' Loan & Trust Co., 157 U.S. 429, where a similar shareholder's right to sue was maintained, and a direct appeal to this court from a decree of the Circuit Court was held to be authorized.

In the Brushaber Case the Chief Justice, speaking for the court, said:

"The right to prevent the corporation from returning and paying the tax was based upon many averments as to the repugnancy of the statute to the Constitution of the United States, of the peculiar relation of the corporation to the stockholders and their particular interests resulting from many of the administrative provisions of the assailed act, of the confusion, wrong and multiplicity of suits and the absence of all means of redress which would result if the corporation paid the tax and complied with the act in other respects without protest, as it was alleged it was its intention to do. To put out of the way a question of jurisdiction we at once say that in view of these averments and [201] the ruling in Pollock v. Farmers' Loan & Trust Co., 157 U.S. 429, sustaining the right of a stockholder to sue to restrain a corporation under proper averments from voluntarily paying a tax charged to be unconstitutional on the ground that to permit such a suit did not violate the prohibitions of § 3224, Rev. Stat., against enjoining the enforcement of taxes, we are of opinion that the contention here made that there was no jurisdiction of the cause since to entertain it would violate the provisions of the Revised Statutes referred to is without merit. . . .

"Aside from averments as to citizenship and residence, recitals as to the provisions of the statute and statements as to the business of the corporation contained in the first ten paragraphs of the bill advanced to sustain jurisdiction, the bill alleged twenty-one constitutional objections specified in that number of paragraphs or subdivisions. As all the grounds assert a violation of the Constitution, it follows that in a wide sense they all charge a repugnancy of the statute to the Sixteenth Amendment under the more immediate sanction of which the statute was adopted."

The jurisdiction of this court is to be determined upon the principles laid down in the cases referred to. In the instant case the averments of the bill show that the directors were proceeding to make the investments in view of the act authorizing the bonds about to be purchased, maintaining that the act authorizing them was constitutional and the bonds valid and desirable investments. The objecting shareholder avers in the bill that the securities were issued under an unconstitutional law, and hence of no validity. It is, therefore, apparent that the controversy concerns the constitutional validity of an act of Congress which is directly drawn in question. The decision depends upon the determination of this issue.

The general allegations as to the interest of the shareholder, and his right to have an injunction to prevent the purchase of the alleged unconstitutional securities by misapplication [202] of the funds of the corporation, give jurisdiction under the principles settled in Pollock v. Farmers' Loan & Trust Co., and Brushaber v. Union Pacific R.R. Co., supra. We are, therefore, of the opinion that the District Court had jurisdiction under the averments of the bill, and that a direct appeal to this court upon constitutional grounds is authorized.

We come to examine the questions presented by the attack upon the constitutionality of the legislation in question. The Federal Farm Loan Act is too lengthy to set out in full. It is entitled: "An Act To provide capital for agricultural development, to create standard forms of investment based upon farm mortgage, to equalize rates of interest upon farm loans, to furnish a market for United States bonds, to create Government depositaries and financial agents for the United States, and for other purposes."

The administration of the act is placed under the direction and control of a Federal Farm Loan Bureau established at the seat of Government in the Treasury Department, under the general supervision of the Federal Farm Loan Board, consisting of the Secretary of the Treasury and four members appointed by the President by and with the advice and consent of the Senate. The United States is divided into twelve districts for the purpose of establishing Federal Land Banks. Each of the banks must have a subscribed capital of not less than $750,000, divided into shares of $5.00 each, which may be subscribed for by any individual, firm or corporation, or by the government of any State, or of the United States. No dividends shall be paid on the stock owned by the United States, but all other stock shall share in dividend distributions without preference. The Federal Farm Loan Board is to designate five directors who shall temporarily manage the affairs of each Federal Land Bank, and who shall prepare an organization certificate which, when approved by the Federal Farm Loan Board and filed with the Farm Loan Commissioner, [203] shall operate to create the bank a body corporate. The Federal Farm Loan Board is required to open books of subscription for the capital stock of each Federal Land Bank, and, if within thirty days thereafter any part of the minimum capitalization of $750,000 of any such bank shall remain unsubscribed, it is made the duty of the Secretary of the Treasury to subscribe the balance on behalf of the United States.

The amendment of January 18, 1918, authorizes the Secretary of the Treasury to purchase bonds issued by Federal Land Banks, and provides that the temporary organization of any such bank shall be continued so long as any Farm Loan Bonds shall be held by the Treasury, and until the subscription to stock in such bank by National Farm Loan Associations shall equal the amount of the stock held by the United States Government. When these conditions are complied with a permanent organization is to take over the management of the bank consisting of a Board of Directors composed of nine members, three of whom shall be known as district directors and shall be appointed by the Farm Loan Board, who shall represent the public interest, six of whom to be known as local directors, shall be chosen by, and be representative of National Farm Loan Associations.

Federal Land Banks are empowered to invest their funds in the purchase of qualified first mortgages on farm lands situated within the Federal Land Bank District within which they are organized or acting. Loans on farm mortgages are to be made to cooperative borrowers through the organization of corporations known as National Farm Loan Associations, by persons desiring to borrow money on farm mortgage security under the terms of the act. Ten or more natural persons who are the owners of or are about to become the owners of farm land qualified as security for mortgage loans, and who desire to borrow money on farm mortgage security, may unite to form a National Farm [204] Loan Association. The manner of forming these associations, and the qualifications for membership, are set out in the act.

A loan desired by each such person must be for not more than $10,000 nor less than $100, and the aggregate of the desired loans not less than $20,000. The application for loan must be accompanied by subscriptions to stock of a Federal Land Bank equal to 5% of the aggregate sum desired on the mortgage loan. Provision is made for appraisal of the land, and report to the Federal Farm Loan Board. No persons but borrowers on farm loan mortgages shall be members or shareholders of National Farm Loan Associations.

Shareholders in Farm Loan Associations are made individually responsible for the debts of the Association to the extent of the amount of the stock owned by them respectively, in addition to the amount paid in and represented by their shares.

When any National Farm Loan Association shall desire to secure for any member a loan on first mortgage from the Federal Land Bank in its district, it must subscribe to the capital stock of the Federal Land Bank to an amount of 5% of such loan, which capital stock shall be held by the Federal Land Bank as collateral security for the payment of the loan, the Association shall be paid any dividends accruing and payable on the capital stock while it is outstanding. Such stock may, in the discretion of the directors and with the approval of the Federal Farm Loan Board, be paid off at par and retired, and shall be so retired upon the full payment of the mortgage loan. In such event, the National Farm Loan Association must pay off at par and retire the corresponding shares of its stock which were issued when the Land Bank stock so retired was issued; but it is further provided that the capital stock of the Land Bank shall not be reduced to less than 5% of the principal of the outstanding Farm [205] Loan Bonds issued by it. The shares in National Farm Loan Associations shall be of the par value of $5.00 each.

At least 25% of that part of the capital of any Federal Land Bank for which stock is outstanding in the name of National Farm Loan Associations must be held in quick assets. Not less than 5% of such capital must be invested in United States Government Bonds.

The loans which Federal Land Banks may make upon first mortgages on farm lands are provided for in § 12 of the act. By § 13 these banks are empowered, subject to the provisions of the act, to issue and sell Farm Loan Bonds of the kind described in the act, and to invest funds in their possession in qualified first mortgages on farm lands, to receive and to deposit in trust with the Farm Loan Registrar, to be held by him as collateral security for Farm Loan Bonds, first mortgages upon farm lands, and, with the approval of the Farm Loan Board, to issue and to sell their bonds secured by the deposit of first mortgages on qualified farm lands as collateral, in conformity with the provisions of § 18 of the act. By the amendment of January 18, 1918, the Secretary of the Treasury was empowered during the years 1918 and 1919, to purchase Farm Loan Bonds issued by Federal Land Banks to an amount not exceeding $100,000,000 each year, and any Federal Land Bank was authorized at any time to repurchase at par and accrued interest, for the purpose of redemption or resale, any of the bonds so purchased from it and held in the United States Treasury.

It is also provided that the bonds of any Federal Land Bank so purchased and held in the Treasury one year after the termination of the pending war shall, upon thirty days' notice from the Secretary of the Treasury, be redeemed and repurchased by such bank at par and accrued interest. By § 15 it is provided that whenever, after the act shall have been in effect for one year, it shall appear to the Federal Farm Loan Board that National Farm Loan [206] Associations have not been formed and are not likely to be formed, in any locality, because of peculiar local conditions, the Board may in its discretion authorize Federal Land Banks to make loans on farm lands through agents approved by the Board, on the terms and conditions and subject to the restrictions prescribed in that section.

The act also authorizes the incorporation of Joint Stock Land Banks, with capital provided by private subscription. They are organized by not less than ten natural persons, and are subject to the requirements of the provisions of § 4 of the act so far as applicable. The board of directors shall consist of not less than five members. Each shareholder shall have the same voting privileges as the holders of shares in National Banking Associations, and shall be held individually responsible, equally and ratably, and not one for another, for all contracts, debts, and engagements of such bank to the extent of the amount of stock owned by them at the par value thereof, in addition to the amount paid in and represented by their shares. The Joint Stock Land Bank is authorized to do business when capital stock to the amount of $250,000 has been subscribed, and one-half paid in cash, the balance remaining subject to call by the board of directors, the charter to be issued by the Federal Farm Loan Board. No bonds shall be issued until the capital stock is entirely paid up. Except as otherwise provided, Joint Stock Land Banks shall have the powers of and be subject to all the restrictions and conditions imposed on Federal Land Banks by the act, so far as such conditions or restrictions are applicable.

Federal Land Banks may issue Farm Loan Bonds up to twenty times their capital and surplus. Joint Stock Land Banks are limited to the issue of Farm Loan Bonds not in excess of fifteen times the amount of their capital and surplus. Joint Stock Land Banks can only loan on first mortgages upon land in the State where located, or in a State [207] contiguous thereto. No loan on mortgage may be made by any bank at a rate exceeding 6% per annum exclusive of amortization payments. Joint Stock Land Banks shall in no case charge a rate of interest on farm loans which shall exceed by more than 1% the rate established by the last series of Farm Loan Bonds issued by them, which rate shall not exceed 5% per annum.

Provisions for the issue of Farm Loan Bonds secured by first mortgages on farm lands or United States bonds, as collateral, are made for Federal Land Banks and Joint Stock Land Banks; in each case the issue is made subject to the approval of the Federal Farm Loan Board. The farm loan mortgages, or United States bonds, which constitute the collateral security for the bonds, must be deposited with the Farm Loan Registrar.

Section 26 of the act provides as follows: "That every Federal land bank and every national farm loan association, including the capital and reserve or surplus therein and the income derived therefrom, shall be exempt from Federal, State, municipal, and local taxation, except taxes upon real estate held, purchased, or taken by said bank or association under the provisions of section eleven and section thirteen of this Act. First mortgages executed to Federal land banks, or to joint stock land banks, and farm loan bonds issued under the provisions of this Act, shall be deemed and held to be instrumentalities of the Government of the United States, and as such they and the income derived therefrom shall be exempt from Federal, State, municipal, and local taxation.

"Nothing herein shall prevent the shares in any joint stock land bank from being included in the valuation of the personal property of the owner or holder of such shares, in assessing taxes imposed by authority of the State within which the bank is located; but such assessment and taxation shall be in manner and subject to the conditions and limitations contained in section fifty-two [208] hundred and nineteen of the Revised Statutes with reference to the shares of national banking associations.

"Nothing herein shall be construed to exempt the real property of Federal and joint stock land banks and national farm loan associations from either State, county, or municipal taxes, to the same extent, according to its value, as other real property is taxed."

Since the decision of the great cases of McCulloch v. Maryland, 4 Wheat. 316, and Osborn v. Bank, 9 Wheat. 738, it is no longer an open question that Congress may establish banks for national purposes, only a small part of the capital of which is held by the Government, and a majority of the ownership in which is represented by shares of capital stock privately owned and held; the principal business of such banks being private banking conducted with the usual methods of such business. While the express power to create a bank or incorporate one is not found in the Constitution, the court, speaking by Chief Justice Marshall, in McCulloch v. Maryland, found authority so to do in the broad general powers conferred by the Constitution upon the Congress to levy and collect taxes, to borrow money, to regulate commerce, to pay the public debts, to declare and conduct war, to raise and support armies, and to provide and maintain a navy, etc. Congress it was held had authority to use such means as were deemed appropriate to exercise the great powers of the Government by virtue of Article I, § 8, cl. 18, of the Constitution granting to Congress the right to make all laws necessary and proper to make the grant effectual. In First National Bank v. Union Trust Co., 244 U.S. 416, 419, the Chief Justice, speaking for the court, after reviewing McCulloch v. Maryland, and Osborn v. Bank, and considering the power given to Congress to pass laws to make the specific powers granted effectual, said:

"In terms it was pointed out that this broad authority [209] was not stereotyped as of any particular time but endured, thus furnishing a perpetual and living sanction to the legislative authority within the limits of a just discretion enabling it to take into consideration the changing wants and demands of society and to adopt provisions appropriate to meet every situation which it was deemed required to be provided for."

That the formation of the bank was required in the judgment of the Congress for the fiscal operations of the Government, was a principal consideration upon which Chief Justice Marshall rested the authority to create the bank; and for that purpose being an appropriate measure in the judgment of the Congress, it was held not to be within the authority of the court to question the conclusion reached by the legislative branch of the Government.

Upon the authority of McCulloch v. Maryland, and Osborn v. Bank, the national banking system was established, and upon them this court has rested the constitutionality of the legislation establishing such banks. Farmers' & Mechanics' National Bank v. Dearing, 91 U.S. 29, 33, 34.

Congress has seen fit in § 6 of the act to make both classes of banks, when designated for that purpose by the Secretary of the Treasury, depositaries of public money, except receipts from customs, under regulations to be prescribed by the Secretary of the Treasury, and has authorized their employment as financial agents of the Government, and the banks are required to perform such reasonable duties, as depositaries of public moneys and financial agents as may be required of them. The Secretary of the Treasury shall require of the Federal Land Banks and the Joint Stock Land Banks, thus designated, satisfactory security, by the deposit of United States bonds or otherwise, for the safe-keeping and prompt payment of the public money deposited with them, and [210] for the faithful performance of their duties as the financial agents of the Government.

Section 6 also provides that no government funds deposited under the provisions of the section shall be invested in mortgage loans or Farm Loan Bonds.

It is said that the power to designate these banks as such depositaries has not been exercised by the Government, and that the Federal Land Banks have acted as federal agents only in the case of loans of money for seed purposes made in the summer of 1918, to which we have already referred. But the existence of the power under the Constitution is not determined by the extent of the exercise of the authority conferred under it. Congress declared it necessary to create these fiscal agencies, and to make them authorized depositaries of public money. Its power to do so is no longer open to question.

But, it is urged, the attempt to create these federal agencies, and to make these banks fiscal agents and public depositaries of the Government, is but a pretext. But nothing is better settled by the decisions of this court than that when Congress acts within the limits of its constitutional authority, it is not the province of the judicial branch of the Government to question its motives. Veazie Bank v. Fenno, 8 Wall. 533, 541; McCray v. United States, 195 U.S. 27; Flint v. Stone Tracy Co., 220 U.S. 107, 147, 153, 156, and cases cited.

That Congress has seen fit, in making these banks fiscal agencies and depositaries of public moneys, to grant to them banking powers of a limited character, in nowise detracts from the authority of Congress to use them for the governmental purposes named, if it sees fit to do so. A bank may be organized with or without the authority to issue currency. It may be authorized to receive deposits in only a limited way. Speaking generally, a bank is a moneyed institution to facilitate the borrowing, lending and caring for money. But whether [211] technically banks, or not, these organizations may serve the governmental purposes declared by Congress in their creation. Furthermore, these institutions are organized to serve as a market for United States bonds. Not less than 5% of the capital of the Federal Land Banks, for which stock is outstanding to Farm Loan Associations, is required to be invested in United States bonds. Both kinds of banks are empowered to buy and sell United States bonds.

In First National Bank v. Union Trust Co., supra, this court sustained the power of Congress to enable a national bank to transact business, which, by itself considered, might be beyond the power of Congress to authorize. In that case it was held to be within the authority of Congress to permit national banks to exercise, by permission of the Federal Reserve Board, when not in contravention of local law, the office of trustee, executor, administrator or registrar of stocks or bonds.

We, therefore, conclude that the creation of these banks, and the grant of authority to them to act for the Government as depositaries of public moneys and purchasers of Government bonds, brings them within the creative power of Congress although they may be intended, in connection with other privileges and duties, to facilitate the making of loans upon farm security at low rates of interest. This does not destroy the validity of these enactments any more than the general banking powers destroyed the authority of Congress to create the United States Bank, or the authority given to national banks to carry on additional activities, destroyed the authority of Congress to create those institutions.

In the brief filed upon reargument counsel for the appellant seem to admit the power of Congress to appropriate money for the direct purposes named, and in that brief they say: "Tax exemption is the real issue sought to be settled here." Deciding, as we do, that these institutions [212] have been created by Congress within the exercise of its legitimate authority, we think the power to make the securities here involved tax exempt necessarily follows. This principle was settled in McCulloch v. Maryland, and Osborn v. Bank, supra.

That the Federal Government can, if it sees fit to do so, exempt such securities from taxation, seems obvious upon the clearest principles. But, it is said to be an invasion of state authority to extend the tax exemption so as to restrain the power of the State. Of a similar contention made in McCulloch v. Maryland, Chief Justice Marshall uttered his often quoted statement: "That the power to tax involves the power to destroy; that the power to destroy may defeat and render useless the power to create; that there is a plain repugnance in conferring on one government a power to control the constitutional measures of another, which other, with respect to those very measures, is declared to be supreme over that which exerts the control, are propositions not to be denied." 4 Wheat. 431.

The same principle has been recognized in the National Bank Cases declaring the power of the States to tax the property and franchises of national banks only to the extent authorized by the laws of Congress. Owensboro National Bank v. Owensboro, 173 U.S. 664, involved the validity of a franchise tax in Kentucky on national banks. In that case this court declared (pp. 668, 669) that the States were wholly without power to levy any tax directly or indirectly upon national banks, their property, assets or franchises, except so far as the permissive legislation of Congress allowed such taxation; and the court declared that the right granted to tax the real estate of such banks, and the shares in the names of the shareholders, constituted the extent of the permission given by Congress, and any tax beyond these was declared to be void.

[213] In Farmers & Mechanics Savings Bank v. Minnesota, 232 U.S. 516, this court held that a State may not tax bonds issued by the municipality of a territory; that to tax such bonds as property in the hands of the holder is, in the last analysis, an imposition upon the right of a municipality to issue them.

The exercise of such taxing power by the States might be so used as to hamper and destroy the exercise of authority conferred by Congress, and this justifies the exemption. If the States can tax these bonds they may destroy the means provided for obtaining the necessary funds for the future operation of the banks. With the wisdom and policy of this legislation we have nothing to do. Ours is only the function of ascertaining whether Congress in the creation of the banks, and in exempting these securities from taxation, federal and state, has acted within the limits of its constitutional authority. For the reasons stated, we think the contention of the Government, and of the appellees, that these banks are constitutionally organized and the securities here involved legally exempted from taxation, must be sustained.

It follows that the decree of the District Court is

Affirmed.

MR. JUSTICE BRANDEIS took no part in the consideration or decision of this case.

MR. JUSTICE HOLMES, dissenting.

No doubt it is desirable that the question raised in this case should be set at rest, but that can be done by the Courts of the United States only within the limits of the jurisdiction conferred upon them by the Constitution and the laws of the United States. As this suit was brought by a citizen of Missouri against a Missouri corporation the [214] single ground upon which the jurisdiction of the District Court can be maintained is that the suit "arises under the Constitution or laws of the United States" within the meaning of § 24 of the Judicial Code. I am of opinion that this case does not arise in that way and therefore that the bill should have been dismissed.

It is evident that the cause of action arises not under any law of the United States but wholly under Missouri law. The defendant is a Missouri corporation and the right claimed is that of a stockholder to prevent the directors from doing an act, that is, making an investment, alleged to be contrary to their duty. But the scope of their duty depends upon the charter of their corporation and other laws of Missouri. If those laws had authorized the investment in terms the plaintiff would have had no case, and this seems to me to make manifest what I am unable to deem even debatable, that, as I have said, the cause of action arises wholly under Missouri law. If the Missouri law authorizes or forbids the investment according to the determination of this Court upon a point under the Constitution or acts of Congress, still that point is material only because the Missouri law saw fit to make it so. The whole foundation of the duty is Missouri law, which at its sole will incorporated the other law as it might incorporate a document. The other law or document depends for its relevance and effect not on its own force but upon the law that took it up, so I repeat once more the cause of action arises wholly from the law of the State.

But it seems to me that a suit cannot be said to arise under any other law than that which creates the cause of action. It may be enough that the law relied upon creates a part of the cause of action although not the whole, as held in Osborn v. Bank of the United States, 9 Wheat. 738, 819-823, which perhaps is all that is meant by the less guarded expressions in Cohens v. Virginia, 6 Wheat. 264, 379. I am content to assume this to be so, although the Osborn Case [215] has been criticized and regretted. But the law must create at least a part of the cause of action by its own force, for it is the suit, not a question in the suit, that must arise under the law of the United States. The mere adoption by a state law of a United States law as a criterion or test, when the law of the United States has no force proprio vigore, does not cause a case under the state law to be also a case under the law of the United States, and so it has been decided by this Court again and again. Miller v. Swann, 150 U.S. 132, 136, 137; Louisville & Nashville R.R. Co v. Western Union Telegraph Co., 237 U.S. 300, 303. See also Shoshone Mining Co. v. Rutter, 177 U.S. 505, 508, 509.

I find nothing contrary to my views in Brushaber v. Union Pacific R.R. Co., 240 U.S. 1, 10. It seems to me plain that the objection that I am considering was not before the mind of the Court or the subject of any of its observations, if open. I am confirmed in my view of that case by the fact that in the next volume of reports is a decision, reached not without discussion and with but a single dissent, that "a suit arises under the law that creates the cause of action." That was the ratio decidendi of American Well Works Co. v. Layne & Bowler Co., 241 U.S. 257, 260. I know of no decisions to the contrary and see no reason for overruling it now.

MR. JUSTICE McREYNOLDS concurs in this dissent. In view of our opinion that this Court has no jurisdiction we express no judgment on the merits.

[1] At the first hearing Mr. Solicitor General King and Mr. W.G. McAdoo, by leave of court, filed a brief on behalf of the United States as amici curiae.

CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT

[309] SOUTER, J., delivered the opinion for a unanimous Court. THOMAS, J., filed a concurring opinion, post, p. 320.

Eric H. Zagrans argued the cause for petitioner. On the briefs was Charles E. McFarland.

[310] Michael C. Walton argued the cause for respondent. With him on the brief were John M. Lichtenberg, Gregory G. Timmer, and Mary L. Tabin.

Irving L. Gornstein argued the cause for the United States as amicus curiae urging affirmance. With him on the brief were Acting Solicitor General Clement, Assistant Attorney General O'Connor, Deputy Solicitor General Hungar, and Gilbert S. Rothenberg.^[1]

JUSTICE SOUTER delivered the opinion of the Court.

The question is whether want of a federal cause of action to try claims of title to land obtained at a federal tax sale precludes removal to federal court of a state action with nondiverse parties raising a disputed issue of federal title law. We answer no, and hold that the national interest in providing a federal forum for federal tax litigation is sufficiently substantial to support the exercise of federal-question jurisdiction over the disputed issue on removal, which would not distort any division of labor between the state and federal courts, provided or assumed by Congress.

I

In 1994, the Internal Revenue Service seized Michigan real property belonging to petitioner Grable & Sons Metal Products, Inc., to satisfy Grable's federal tax delinquency. Title 26 U. S. C. § 6335 required the IRS to give notice of the seizure, and there is no dispute that Grable received actual notice by certified mail before the IRS sold the property to respondent Darue Engineering & Manufacturing. Although Grable also received notice of the sale itself, it did not exercise its statutory right to redeem the property within 180 days of the sale, § 6337(b)(1), and after that period [311] had passed, the Government gave Darue a quitclaim deed, § 6339.

Five years later, Grable brought a quiet title action in state court, claiming that Darue's record title was invalid because the IRS had failed to notify Grable of its seizure of the property in the exact manner required by § 6335(a), which provides that written notice must be "given by the Secretary to the owner of the property [or] left at his usual place of abode or business." Grable said that the statute required personal service, not service by certified mail.

Darue removed the case to Federal District Court as presenting a federal question, because the claim of title depended on the interpretation of the notice statute in the federal tax law. The District Court declined to remand the case at Grable's behest after finding that the "claim does pose a `significant question of federal law," Tr. 17 (Apr. 2, 2001), and ruling that Grable's lack of a federal right of action to enforce its claim against Darue did not bar the exercise of federal jurisdiction. On the merits, the court granted summary judgment to Darue, holding that although § 6335 by its terms required personal service, substantial compliance with the statute was enough. 207 F. Supp. 2d 694 (WD Mich. 2002).

The Court of Appeals for the Sixth Circuit affirmed. 377 F. 3d 592 (2004). On the jurisdictional question, the panel thought it sufficed that the title claim raised an issue of federal law that had to be resolved, and implicated a substantial federal interest (in construing federal tax law). The court went on to affirm the District Court's judgment on the merits. We granted certiorari on the jurisdictional question alone,^[2] 543 U. S. 1042 (2005), to resolve a split within the Courts of Appeals on whether Merrell Dow Pharmaceuticals Inc. v. Thompson, 478 U. S. 804 (1986), always requires [312] a federal cause of action as a condition for exercising federal-question jurisdiction.^[3] We now affirm.

II

Darue was entitled to remove the quiet title action if Grable could have brought it in federal district court originally, 28 U. S. C. § 1441(a), as a civil action "arising under the Constitution, laws, or treaties of the United States," § 1331. This provision for federal-question jurisdiction is invoked by and large by plaintiffs pleading a cause of action created by federal law (e. g., claims under 42 U. S. C. § 1983). There is, however, another longstanding, if less frequently encountered, variety of federal "arising under" jurisdiction, this Court having recognized for nearly 100 years that in certain cases federal-question jurisdiction will lie over state-law claims that implicate significant federal issues. E. g., Hopkins v. Walker, 244 U. S. 486, 490-491 (1917). The doctrine captures the commonsense notion that a federal court ought to be able to hear claims recognized under state law that nonetheless turn on substantial questions of federal law, and thus justify resort to the experience, solicitude, and hope of uniformity that a federal forum offers on federal issues, see ALI, Study of the Division of Jurisdiction Between State and Federal Courts 164-166 (1968).

The classic example is Smith v. Kansas City Title & Trust Co., 255 U. S. 180 (1921), a suit by a shareholder claiming that the defendant corporation could not lawfully buy certain bonds of the National Government because their issuance was unconstitutional. Although Missouri law provided the cause of action, the Court recognized federal-question jurisdiction because the principal issue in the case was the federal constitutionality of the bond issue. Smith thus held, in a [313] somewhat generous statement of the scope of the doctrine, that a state-law claim could give rise to federal-question jurisdiction so long as it "appears from the [complaint] that the right to relief depends upon the construction or application of [federal law]." Id., at 199.

The Smith statement has been subject to some trimming to fit earlier and later cases recognizing the vitality of the basic doctrine, but shying away from the expansive view that mere need to apply federal law in a state-law claim will suffice to open the "arising under" door. As early as 1912, this Court had confined federal-question jurisdiction over state-law claims to those that "really and substantially involv[e] a dispute or controversy respecting the validity, construction or effect of [federal] law." Shulthis v. McDougal, 225 U. S. 561, 569. This limitation was the ancestor of Justice Cardozo's later explanation that a request to exercise federal-question jurisdiction over a state action calls for a "common-sense accommodation of judgment to [the] kaleidoscopic situations" that present a federal issue, in "a selective process which picks the substantial causes out of the web and lays the other ones aside." Gully v. First Nat. Bank in Meridian, 299 U. S. 109, 117-118 (1936). It has in fact become a constant refrain in such cases that federal jurisdiction demands not only a contested federal issue, but a substantial one, indicating a serious federal interest in claiming the advantages thought to be inherent in a federal forum. E. g., Chicago v. International College of Surgeons, 522 U. S. 156, 164 (1997); Merrell Dow, supra, at 814, and n. 12; Franchise Tax Bd. of Cal. v. Construction Laborers Vacation Trust for Southern Cal., 463 U. S. 1, 28 (1983).

But even when the state action discloses a contested and substantial federal question, the exercise of federal jurisdiction is subject to a possible veto. For the federal issue will ultimately qualify for a federal forum only if federal jurisdiction is consistent with congressional judgment about the sound division of labor between state and federal courts governing [314] the application of § 1331. Thus, Franchise Tax Bd. explained that the appropriateness of a federal forum to hear an embedded issue could be evaluated only after considering the "welter of issues regarding the interrelation of federal and state authority and the proper management of the federal judicial system." Id., at 8. Because arising-under jurisdiction to hear a state-law claim always raises the possibility of upsetting the state-federal line drawn (or at least assumed) by Congress, the presence of a disputed federal issue and the ostensible importance of a federal forum are never necessarily dispositive; there must always be an assessment of any disruptive portent in exercising federal jurisdiction. See also Merrell Dow, supra, at 810.

These considerations have kept us from stating a "single, precise, all-embracing" test for jurisdiction over federal issues embedded in state-law claims between nondiverse parties. Christianson v. Colt Industries Operating Corp., 486 U. S. 800, 821 (1988) (STEVENS, J., concurring). We have not kept them out simply because they appeared in state raiment, as Justice Holmes would have done, see Smith, supra, at 214 (dissenting opinion), but neither have we treated "federal issue" as a password opening federal courts to any state action embracing a point of federal law. Instead, the question is, does a state-law claim necessarily raise a stated federal issue, actually disputed and substantial, which a federal forum may entertain without disturbing any congressionally approved balance of federal and state judicial responsibilities.

III

A

This case warrants federal jurisdiction. Grable's state complaint must specify "the facts establishing the superiority of [its] claim," Mich. Ct. Rule 3.411(B)(2)(c) (West 2005), and Grable has premised its superior title claim on a failure by the IRS to give it adequate notice, as defined by federal [315] law. Whether Grable was given notice within the meaning of the federal statute is thus an essential element of its quiet title claim, and the meaning of the federal statute is actually in dispute; it appears to be the only legal or factual issue contested in the case. The meaning of the federal tax provision is an important issue of federal law that sensibly belongs in a federal court. The Government has a strong interest in the "prompt and certain collection of delinquent taxes," United States v. Rodgers, 461 U. S. 677, 709 (1983), and the ability of the IRS to satisfy its claims from the property of delinquents requires clear terms of notice to allow buyers like Darue to satisfy themselves that the Service has touched the bases necessary for good title. The Government thus has a direct interest in the availability of a federal forum to vindicate its own administrative action, and buyers (as well as tax delinquents) may find it valuable to come before judges used to federal tax matters. Finally, because it will be the rare state title case that raises a contested matter of federal law, federal jurisdiction to resolve genuine disagreement over federal tax title provisions will portend only a microscopic effect on the federal-state division of labor. See n. 3, infra.

This conclusion puts us in venerable company, quiet title actions having been the subject of some of the earliest exercises of federal-question jurisdiction over state-law claims. In Hopkins, 244 U. S., at 490-491, the question was federal jurisdiction over a quiet title action based on the plaintiffs' allegation that federal mining law gave them the superior claim. Just as in this case, "the facts showing the plaintiffs' title and the existence and invalidity of the instrument or record sought to be eliminated as a cloud upon the title are essential parts of the plaintiffs' cause of action."^[4]Id., at [316] 490. As in this case again, "it is plain that a controversy respecting the construction and effect of the [federal] laws is involved and is sufficiently real and substantial." Id., at 489. This Court therefore upheld federal jurisdiction in Hopkins, as well as in the similar quiet title matters of Northern Pacific R. Co. v. Soderberg, 188 U. S. 526, 528 (1903), and Wilson Cypress Co. v. Del Pozo y Marcos, 236 U. S. 635, 643-644 (1915). Consistent with those cases, the recognition of federal jurisdiction is in order here.

B

Merrell Dow Pharmaceuticals Inc. v. Thompson, 478 U. S. 804 (1986), on which Grable rests its position, is not to the contrary. Merrell Dow considered a state tort claim resting in part on the allegation that the defendant drug company had violated a federal misbranding prohibition, and was thus presumptively negligent under Ohio law. Id., at 806. The Court assumed that federal law would have to be applied to resolve the claim, but after closely examining the strength of the federal interest at stake and the implications of opening the federal forum, held federal jurisdiction unavailable. Congress had not provided a private federal cause of action for violation of the federal branding requirement, and the Court found "it would . . . flout, or at least undermine, congressional intent to conclude that federal courts might nevertheless exercise federal-question jurisdiction and provide remedies for violations of that federal statute solely because the violation . . . is said to be a . . . `proximate cause' under state law." Id., at 812.

[317] Because federal law provides for no quiet title action that could be brought against Darue,^[5] Grable argues that there can be no federal jurisdiction here, stressing some broad language in Merrell Dow (including the passage just quoted) that on its face supports Grable's position, see Note, Mr. Smith Goes to Federal Court: Federal Question Jurisdiction over State Law Claims Post-Merrell Dow, 115 Harv. L. Rev. 2272, 2280-2282 (2002) (discussing split in Courts of Appeals over private right of action requirement after Merrell Dow). But an opinion is to be read as a whole, and Merrell Dow cannot be read whole as overturning decades of precedent, as it would have done by effectively adopting the Holmes dissent in Smith, see supra, at 314, and converting a federal cause of action from a sufficient condition for federal-question jurisdiction^[6] into a necessary one.

In the first place, Merrell Dow disclaimed the adoption of any bright-line rule, as when the Court reiterated that "in exploring the outer reaches of § 1331, determinations about federal jurisdiction require sensitive judgments about congressional intent, judicial power, and the federal system." 478 U. S., at 810. The opinion included a lengthy footnote explaining that questions of jurisdiction over state-law claims require "careful judgments," id., at 814, about the "nature of the federal interest at stake," id., at 814, n. 12 (emphasis deleted). And as a final indication that it did not mean to make a federal right of action mandatory, it expressly approved the exercise of jurisdiction sustained in Smith, despite the want of any federal cause of action available to Smith's shareholder plaintiff. 478 U. S., at 814, n. 12. [318] Merrell Dow then, did not toss out, but specifically retained, the contextual enquiry that had been Smith's hallmark for over 60 years. At the end of Merrell Dow, Justice Holmes was still dissenting.

Accordingly, Merrell Dow should be read in its entirety as treating the absence of a federal private right of action as evidence relevant to, but not dispositive of, the "sensitive judgments about congressional intent" that § 1331 requires. The absence of any federal cause of action affected Merrell Dow's result two ways. The Court saw the fact as worth some consideration in the assessment of substantiality. But its primary importance emerged when the Court treated the combination of no federal cause of action and no preemption of state remedies for misbranding as an important clue to Congress's conception of the scope of jurisdiction to be exercised under § 1331. The Court saw the missing cause of action not as a missing federal door key, always required, but as a missing welcome mat, required in the circumstances, when exercising federal jurisdiction over a state misbranding action would have attracted a horde of original filings and removal cases raising other state claims with embedded federal issues. For if the federal labeling standard without a federal cause of action could get a state claim into federal court, so could any other federal standard without a federal cause of action. And that would have meant a tremendous number of cases.

One only needed to consider the treatment of federal violations generally in garden variety state tort law. "The violation of federal statutes and regulations is commonly given negligence per se effect in state tort proceedings."^[7] Restatement [319] (Third) of Torts § 14, Reporters' Note, Comment a, p. 195 (Tent. Draft No. 1, Mar. 28, 2001) (hereinafter Restatement). See also W. Keeton, D. Dobbs, R. Keeton, & D. Owen, Prosser and Keeton on Law of Torts § 36, p. 221, n. 9 (5th ed. 1984) ("[T]he breach of a federal statute may support a negligence per se claim as a matter of state law" (collecting authority)). A general rule of exercising federal jurisdiction over state claims resting on federal mislabeling and other statutory violations would thus have heralded a potentially enormous shift of traditionally state cases into federal courts. Expressing concern over the "increased volume of federal litigation," and noting the importance of adhering to "legislative intent," Merrell Dow thought it improbable that the Congress, having made no provision for a federal cause of action, would have meant to welcome any state-law tort case implicating federal law "solely because the violation of the federal statute is said to [create] a rebuttable presumption [of negligence] . . . under state law." 478 U. S., at 811-812 (internal quotation marks omitted). In this situation, no welcome mat meant keep out. Merrell Dow's analysis thus fits within the framework of examining the importance of having a federal forum for the issue, and the consistency of such a forum with Congress's intended division of labor between state and federal courts.

As already indicated, however, a comparable analysis yields a different jurisdictional conclusion in this case. Although Congress also indicated ambivalence in this case by providing no private right of action to Grable, it is the rare state quiet title action that involves contested issues of federal law, see n. 3, supra. Consequently, jurisdiction over actions like Grable's would not materially affect, or threaten to affect, the normal currents of litigation. Given the absence of threatening structural consequences and the clear interest the Government, its buyers, and its delinquents have in the availability of a federal forum, there is no good reason to [320] shirk from federal jurisdiction over the dispositive and contested federal issue at the heart of the state-law title claim.^[8]

IV

The judgment of the Court of Appeals, upholding federal jurisdiction over Grable's quiet title action, is affirmed.

It is so ordered.

JUSTICE THOMAS, concurring.

The Court faithfully applies our precedents interpreting 28 U. S. C. § 1331 to authorize federal-court jurisdiction over some cases in which state law creates the cause of action but requires determination of an issue of federal law, e. g., Smith v. Kansas City Title & Trust Co., 255 U. S. 180 (1921); Merrell Dow Pharmaceuticals Inc. v. Thompson, 478 U. S. 804 (1986). In this case, no one has asked us to overrule those precedents and adopt the rule Justice Holmes set forth in American Well Works Co. v. Layne & Bowler Co., 241 U. S. 257 (1916), limiting § 1331 jurisdiction to cases in which federal law creates the cause of action pleaded on the face of the plaintiff's complaint. Id., at 260. In an appropriate case, and perhaps with the benefit of better evidence as to the original meaning of § 1331's text, I would be willing to consider that course.^[9]

[321] Jurisdictional rules should be clear. Whatever the virtues of the Smith standard, it is anything but clear. Ante, at 313 (the standard "calls for a `common-sense accommodation of judgment to [the] kaleidoscopic situations' that present a federal issue, in `a selective process which picks the substantial causes out of the web and lays the other ones aside'" (quoting Gully v. First Nat. Bank in Meridian, 299 U. S. 109, 117-118 (1936))); ante, at 314 ("[T]he question is, does a state-law claim necessarily raise a stated federal issue, actually disputed and substantial, which a federal forum may entertain without disturbing any congressionally approved balance of federal and state judicial responsibilities"); ante, at 317, 318 ("`[D]eterminations about federal jurisdiction require sensitive judgments about congressional intent, judicial power, and the federal system'"; "the absence of a federal private right of action [is] evidence relevant to, but not dispositive of, the `sensitive judgments about congressional intent' that § 1331 requires" (quoting Merrell Dow, supra, at 810)).

Whatever the vices of the American Well Works rule, it is clear. Moreover, it accounts for the "`vast majority'" of cases that come within § 1331 under our current case law, Merrell Dow, supra, at 808 (quoting Franchise Tax Bd. of Cal. v. Construction Laborers Vacation Trust for Southern Cal., 463 U. S. 1, 9 (1983)) — further indication that trying to sort out which cases fall within the smaller Smith category may not be worth the effort it entails. See R. Fallon, D. Meltzer, & D. Shapiro, Hart and Wechsler's The Federal [322] Courts and the Federal System 885-886 (5th ed. 2003). Accordingly, I would be willing in appropriate circumstances to reconsider our interpretation of § 1331.

[1] Mr. Zagrans filed a brief for Jerome R. Mikulski et ux. as amici curiae urging reversal.

[2] Accordingly, we have no occasion to pass upon the proper interpretation of the federal tax provision at issue here.

[3] Compare Seinfeld v. Austen, 39 F. 3d 761, 764 (CA7 1994) (finding that federal-question jurisdiction over a state-law claim requires a parallel federal private right of action), with Ormet Corp. v. Ohio Power Co., 98 F. 3d 799, 806 (CA4 1996) (finding that a federal private action is not required).

[4] The quiet title cases also show the limiting effect of the requirement that the federal issue in a state-law claim must actually be in dispute to justify federal-question jurisdiction. In Shulthis v. McDougal, 225 U. S. 561 (1912), this Court found that there was no federal-question jurisdiction to hear a plaintiff's quiet title claim in part because the federal statutes on which title depended were not subject to "any controversy respecting their validity, construction, or effect." Id., at 570. As the Court put it, the requirement of an actual dispute about federal law was "especially" important in "suit[s] involving rights to land acquired under a law of the United States," because otherwise "every suit to establish title to land in the central and western states would so arise [under federal law], as all titles in those States are traceable back to those laws." Id., at 569-570.

[5] Federal law does provide a quiet title cause of action against the Federal Government. 28 U. S. C. § 2410. That right of action is not relevant here, however, because the Federal Government no longer has any interest in the property, having transferred its interest to Darue through the quitclaim deed.

[6] For an extremely rare exception to the sufficiency of a federal right of action, see Shoshone Mining Co. v. Rutter, 177 U. S. 505, 507 (1900).

[7] Other jurisdictions treat a violation of a federal statute as evidence of negligence or, like Ohio itself in Merrell Dow Pharmaceuticals Inc. v. Thompson, 478 U. S. 804 (1986), as creating a rebuttable presumption of negligence. Restatement § 14, Reporters' Note, Comment c, at 196. Either approach could still implicate issues of federal law.

[8] At oral argument Grable's counsel espoused the position that after Merrell Dow, federal-question jurisdiction over state-law claims absent a federal right of action could be recognized only where a constitutional issue was at stake. There is, however, no reason in text or otherwise to draw such a rough line. As Merrell Dow itself suggested, constitutional questions may be the more likely ones to reach the level of substantiality that can justify federal jurisdiction. 478 U. S., at 814, n. 12. But a flat ban on statutory questions would mechanically exclude significant questions of federal law like the one this case presents.

[9] This Court has long construed the scope of the statutory grant of federal-question jurisdiction more narrowly than the scope of the constitutional grant of such jurisdiction. See Merrell Dow Pharmaceuticals Inc. v. Thompson, 478 U. S. 804, 807-808 (1986). I assume for present purposes that this distinction is proper — that is, that the language of 28 U. S. C. § 1331, "[t]he district courts shall have original jurisdiction of all civil actions arising under the Constitution, laws, or treaties of the United States" (emphasis added), is narrower than the language of Art. III, § 2, cl. 1, of the Constitution, "[t]he judicial Power shall extend to all Cases, in Law and Equity, arising under this Constitution, the Laws of the United States, and Treaties made, or which shall be made, under their Authority . . ." (emphases added).

[682] Anthony F. Shelley argued the cause for petitioner. With him on the briefs were Alan I. Horowitz, Laura G. Ferguson, Kathleen M. Sullivan, Roger G. Wilson, Paul F. Brown, and William A. Breskin.

Sri Srinivasan argued the cause for the United States as amicus curiae urging reversal. On the brief were Solicitor General Clement, Assistant Attorney General Keisler, Deputy Solicitor General Kneedler, James A. Feldman, Mark B. Stern, Alisa B. Klein, Mark A. Robbins, and James S. Green.

Thomas J. Stock argued the cause for respondent. With him on the brief were Harry Raptakis and Victor A. Carr.^[1]

JUSTICE GINSBURG delivered the opinion of the Court.

The Federal Employees Health Benefits Act of 1959 (FEHBA), 5 U. S. C. § 8901 et seq. (2000 ed. and Supp. III), establishes a comprehensive program of health insurance for federal employees. The Act authorizes the Office of Personnel Management (OPM) to contract with private carriers to offer federal employees an array of health-care plans. See § 8902(a) (2000 ed.). Largest of the plans for which OPM has contracted, annually since 1960, is the Blue Cross Blue Shield Service Benefit Plan (Plan), administered by local Blue Cross Blue Shield companies. This case concerns the proper forum for reimbursement claims when a Plan beneficiary, injured in an accident, whose medical bills have been paid by the Plan administrator, recovers damages (unaided by the carrier-administrator) in a state-court tort action against a third party alleged to have caused the accident.

[683] FEHBA contains a preemption clause, § 8902(m)(1), displacing state law on issues relating to "coverage or benefits" afforded by health-care plans. The Act contains no provision addressing the subrogation or reimbursement rights of carriers. Successive annual contracts between OPM and the Blue Cross Blue Shield Association (BCBSA) have obligated the carrier to make "a reasonable effort" to recoup amounts paid for medical care. App. 95, 125. The statement of benefits distributed by the carrier alerts enrollees that all recoveries they receive "must be used to reimburse the Plan for benefits paid." Id., at 132; see also id., at 146, 152.

The instant case originated when the administrator of a Plan beneficiary's estate pursued tort litigation in state court against parties alleged to have caused the beneficiary's injuries. The carrier had notice of the state-court action, but took no part in it. When the tort action terminated in a settlement, the carrier filed suit in federal court seeking reimbursement of the full amount it had paid for the beneficiary's medical care. The question presented is whether 28 U. S. C. § 1331 (authorizing jurisdiction over "civil actions arising under the . . . laws . . . of the United States") encompasses the carrier's action. We hold it does not.

FEHBA itself provides for federal-court jurisdiction only in actions against the United States. Congress could decide and provide that reimbursement claims of the kind here involved warrant the exercise of federal-court jurisdiction. But claims of this genre, seeking recovery from the proceeds of state-court litigation, are the sort ordinarily resolved in state courts. Federal courts should await a clear signal from Congress before treating such auxiliary claims as "arising under" the laws of the United States.

I

FEHBA assigns to OPM responsibility for negotiating and regulating health-benefits plans for federal employees. See [684] 5 U. S. C. § 8902(a). OPM contracts with carriers, FEHBA instructs, "shall contain a detailed statement of benefits offered and shall include such maximums, limitations, exclusions, and other definitions of benefits as [OPM] considers necessary or desirable." § 8902(d). Pursuant to FEHBA, OPM entered into a contract in 1960 with the BCBSA to establish a nationwide fee-for-service health plan, the terms of which are renegotiated annually. As FEHBA prescribes, the Federal Government pays about 75% of the premiums; the enrollee pays the rest. § 8906(b). Premiums thus shared are deposited in a special Treasury Fund, the Federal Employees Health Benefits Fund, § 8909(a). Carriers draw against the Fund to pay for covered health-care benefits. Ibid.; see also 48 CFR § 1632.170(b) (2005).

The contract between OPM and the BCBSA provides: "By enrolling or accepting services under this contract, [enrollees and their eligible dependents] are obligated to all terms, conditions, and provisions of this contract." App. 90. An appended brochure sets out the benefits the carrier shall provide, see id., at 89, and the carrier's subrogation and recovery rights, see id., at 100. Each enrollee, as FEHBA directs, receives a statement of benefits conveying information about the Plan's coverage and conditions. 5 U. S. C. § 8907(b). Concerning reimbursement and subrogation, matters FEHBA itself does not address, the BCBSA Plan's statement of benefits reads in part:

"If another person or entity . . . causes you to suffer an injury or illness, and if we pay benefits for that injury or illness, you must agree to the following:

"All recoveries you obtain (whether by lawsuit, settlement, or otherwise), no matter how described or designated, must be used to reimburse us in full for benefits we paid. Our share of any recovery extends only to the amount of benefits we have paid or will pay to you or, if applicable, to your heirs, administrators, successors, or assignees.

.....

[685] "If you do not seek damages for your illness or injury, you must permit us to initiate recovery on your behalf (including the right to bring suit in your name). This is called subrogation.

"If we pursue a recovery of the benefits we have paid, you must cooperate in doing what is reasonably necessary to assist us. You must not take any action that may prejudice our rights to recover." App. 165.^[2]

If the participant does not voluntarily reimburse the Plan, the contract requires the carrier to make a "reasonable effort to seek recovery of amounts . . . it is entitled to recover in cases . . . brought to its attention." Id., at 95, 125. Pursuant to the OPM—BCBSA master contract, reimbursements obtained by the carrier must be returned to the Treasury Fund. See id., at 92, 118-119.

FEHBA contains a preemption provision, which originally provided:

"The provisions of any contract under this chapter which relate to the nature or extent of coverage or benefits (including payments with respect to benefits) shall supersede and preempt any State or local law, or any regulation issued thereunder, which relates to health insurance or plans to the extent that such law or regulation is inconsistent with such contractual provisions." 5 U. S. C. § 8902(m)(1) (1994 ed.).

[686] To ensure uniform coverage and benefits under plans OPM negotiates for federal employees, see H. R. Rep. No. 95-282, p. 1 (1977), § 8902(m)(1) preempted "State laws or regulations which specify types of medical care, providers of care, extent of benefits, coverage of family members, age limits for family members, or other matters relating to health benefits or coverage," id., at 4-5 (noting that some States mandated coverage for services not included in federal plans, for example, chiropractic services). In 1998, Congress amended § 8902(m)(1) by deleting the words "to the extent that such law or regulation is inconsistent with such contractual provisions." Thus, under § 8902(m)(1) as it now reads, state law—whether consistent or inconsistent with federal plan provisions—is displaced on matters of "coverage or benefits."

FEHBA contains but one provision addressed to federalcourt jurisdiction. That provision vests in federal district courts "original jurisdiction, concurrent with the United States Court of Federal Claims, of a civil action or claim against the United States founded on this chapter." § 8912. The purpose of this provision—evident from its reference to the Court of Federal Claims—was to carve out an exception to the statutory rule that claims brought against the United States and exceeding $10,000 must originate in the Court of Federal Claims. See 28 U. S. C. § 1346(a)(2) (establishing district courts' jurisdiction, concurrent with the Court of Federal Claims, over claims against the United States that do not exceed $10,000); see also S. Rep. No. 1654, 83d Cong., 2d Sess., 4-5 (1954) (commenting, with respect to an identical provision in the Federal Employees' Group Life Insurance Act, 5 U. S. C. § 8715, that the provision "would extend the jurisdiction of United States district courts above the $10,000 limitation now in effect").

Under a 1995 OPM regulation, suits contesting final OPM action denying health benefits "must be brought against OPM and not against the carrier or carrier's subcontractors." 5 CFR § 890.107(c) (2005). While this regulation channels [687] disputes over coverage or benefits into federal court by designating a United States agency (OPM) sole defendant, no law opens federal courts to carriers seeking reimbursement from beneficiaries or recovery from tortfeasors. Cf. 29 U. S. C. § 1132(e)(1) (provision of the Employee Retirement Income Security Act (ERISA) vesting in federal district courts "exclusive jurisdiction of civil actions under this subchapter"). And nothing in FEHBA's text prescribes a federal rule of decision for a carrier's claim against its insured or an alleged tortfeasor to share in the proceeds of a state-court tort action.

II

Petitioner Empire HealthChoice Assurance, Inc., doing business as Empire Blue Cross Blue Shield (Empire), is the entity that administers the BCBSA Plan as it applies to federal employees in New York State. Respondent Denise Finn McVeigh (McVeigh) is the administrator of the estate of Joseph E. McVeigh (Decedent), a former enrollee in the Plan. The Decedent was injured in an accident in 1997. Plan payments for the medical care he received between 1997 and his death in 2001 amounted to $157,309. McVeigh, on behalf of herself, the Decedent, and a minor child, commenced tort litigation in state court against parties alleged to have caused Decedent's injuries. On learning that the parties to the state-court litigation had agreed to settle the tort claims, Empire sought to recover the $157,309 it had paid out for the Decedent's medical care.^[3] Of the $3,175,000 for which the settlement provided, McVeigh, in response to Empire's asserted reimbursement right, agreed to place $100,000 in escrow.

Empire then filed suit in the United States District Court for the Southern District of New York, alleging that McVeigh [688] was in breach of the reimbursement provision of the Plan. As relief, Empire demanded $157,309, with no offset for attorney's fees or other litigation costs McVeigh incurred in pursuing the state-court settlement. McVeigh moved to dismiss on various grounds, among them, lack of subject-matter jurisdiction. See 396 F. 3d 136, 139 (CA2 2005). Answering McVeigh's motion, Empire urged that the District Court had jurisdiction under 28 U. S. C. § 1331 because federal common law governed its reimbursement claim. In the alternative, Empire asserted that the Plan itself constituted federal law. See 396 F. 3d, at 140. The District Court rejected both arguments and granted McVeigh's motion to dismiss for want of subject-matter jurisdiction. Ibid.

A divided panel of the Court of Appeals for the Second Circuit affirmed, holding that "Empire's clai[m] arise[s] under state law." Id., at 150. FEHBA's text, the court observed, contains no authorization for carriers "to vindicate [in federal court] their rights [against enrollees] under FEHBAauthorized contracts"; therefore, the court concluded, "federal jurisdiction exists over this dispute only if federal common law governs Empire's claims." Id., at 140. Quoting Boyle v. United Technologies Corp., 487 U. S. 500, 507, 508 (1988), the appeals court stated that courts may create federal common law only when "the operation of state law would (1) `significant[ly] conflict' with (2) `uniquely federal interest[s].'" 396 F. 3d, at 140.

Empire maintained that its contract-derived claim against McVeigh implicated "`uniquely federal interest[s],'" because (1) reimbursement directly affects the United States Treasury and the cost of providing health benefits to federal employees; and (2) Congress had expressed its interest in maintaining uniformity among the States on matters relating to federal health-plan benefits. Id., at 141. The court acknowledged that the case involved distinctly federal interests, but found that Empire had not identified "specific ways in which the operation of state contract law, or indeed of [689] other laws of general application, would conflict materially with the federal policies underlying FEHBA in the circumstances presented." Id., at 150 (Sack, J., concurring); see id., at 142.

The Court of Appeals next considered and rejected Empire's argument that FEHBA's preemption provision, 5 U. S. C. § 8902(m)(1), independently conferred federal jurisdiction. 396 F. 3d, at 145-149. That provision, the court observed, is "a limited preemption clause that the instant dispute does not trigger." Id., at 145. Unlike § 8912, which "authoriz[es] federal jurisdiction over FEHBA-related . . . claims `against the United States,'" the court noted, § 8902(m)(1) "makes no reference to a federal right of action [in] or to federal jurisdiction [over]" the contract-derived reimbursement claim here at issue. 396 F. 3d, at 145, and n. 7.

Judge Raggi dissented. Id., at 151. In her view, FEHBA's preemption provision, § 8902(m)(1), as amended in 1998, both calls for the application of uniform federal common law to terms in a FEHBA plan and establishes federal jurisdiction over Empire's complaint.

We granted certiorari, 546 U. S. 1085 (2005), to resolve a conflict among lower federal courts concerning the proper forum for claims of the kind Empire asserts. Compare Blue Cross & Blue Shield of Ill. v. Cruz, 396 F. 3d 793, 799-800 (CA7 2005) (upholding federal jurisdiction), Caudill v. Blue Cross & Blue Shield of N. C., 999 F. 2d 74, 77 (CA4 1993) (same), and Medcenters Health Care v. Ochs, 854 F. Supp. 589, 593, and n. 3 (Minn. 1993) (same), aff'd, 26 F. 3d 865 (CA8 1994), with Goepel v. National Postal Mail Handlers Union, 36 F. 3d 306, 314-315 (CA3 1994) (rejecting federal jurisdiction), and 396 F. 3d, at 139 (decision below) (same).

III

Title 28 U. S. C. § 1331 vests in federal district courts "original jurisdiction" over "all civil actions arising under the Constitution, laws, or treaties of the United States." A [690] case "aris[es] under" federal law within the meaning of § 1331, this Court has said, if "a well-pleaded complaint establishes either that federal law creates the cause of action or that the plaintiff's right to relief necessarily depends on resolution of a substantial question of federal law." Franchise Tax Bd. of Cal. v. Construction Laborers Vacation Trust for Southern Cal., 463 U. S. 1, 27-28 (1983).

Empire and the United States, as amicus curiae, present two principal arguments in support of federal-question jurisdiction. Emphasizing our opinion in Jackson Transit Authority v. Transit Union, 457 U. S. 15, 22 (1982), and cases cited therein, they urge that Empire's complaint raises a federal claim because it seeks to vindicate a contractual right contemplated by a federal statute, a right that Congress intended to be federal in nature. See Brief for Petitioner 14-31; Brief for United States 12-23. FEHBA's preemption provision, Empire and the United States contend, demonstrates Congress' intent in this regard. The United States argues, alternatively, that there is federal jurisdiction here, as demonstrated by our recent decision in Grable & Sons Metal Products, Inc. v. Darue Engineering & Mfg., 545 U. S. 308 (2005), because "federal law is a necessary element of [Empire's] claim." Brief for United States 25; accord Brief for Petitioner 41, n. 5. We address these arguments in turn. But first, we respond to the dissent's view that Empire and the United States have engaged in unnecessary labor, for Clearfield Trust Co. v. United States, 318 U. S. 363 (1943), provides "a basis for federal jurisdiction" in this case. Post, at 702.

A

Clearfield is indeed a pathmarking precedent on the authority of federal courts to fashion uniform federal common law on issues of national concern. See Friendly, In Praise of Erie—and of the New Federal Common Law, 39 N. Y. U. L. Rev. 383, 409-410 (1964). But the dissent is mistaken in supposing that the Clearfield doctrine covers this case. [691] Clearfield was a suit by the United States to recover from a bank the amount paid on a Government check on which the payee's name had been forged. 318 U. S., at 365. Because the United States was the plaintiff, federal-court jurisdiction was solidly grounded. See ibid. ("This suit was instituted. . . by the United States . . ., the jurisdiction of the federal District Court being invoked pursuant to the provisions of § 24(1) of the Judicial Code, 28 U. S. C. § 41(1)," now contained in 28 U. S. C. §§ 1332, 1345, 1359). The case presented a vertical choice-of-law issue: Did state law under Erie R. Co. v. Tompkins, 304 U. S. 64 (1938), or a court-fashioned federal rule of decision (federal common law) determine the merits of the controversy? The Court held that "[t]he rights and duties of the United States on commercial paper which it issues are governed by federal rather than [state] law." 318 U. S., at 366.

In post-Clearfield decisions, and with the benefit of enlightened commentary, see, e. g., Friendly, supra, at 410, the Court has "made clear that uniform federal law need not be applied to all questions in federal government litigation, even in cases involving government contracts," R. Fallon, D. Meltzer, & D. Shapiro, Hart and Wechsler's The Federal Courts and the Federal System 700 (5th ed. 2003) (hereinafter Hart and Wechsler).^[4] "[T]he prudent course," we have recognized, is often "to adopt the readymade body of state [692] law as the federal rule of decision until Congress strikes a different accommodation." United States v. Kimbell Foods, Inc., 440 U. S. 715, 740 (1979).

Later, in Boyle, the Court telescoped the appropriate inquiry, focusing it on the straightforward question whether the relevant federal interest warrants displacement of state law. See 487 U. S., at 507, n. 3. Referring simply to "the displacement of state law," the Court recognized that prior cases had treated discretely (1) the competence of federal courts to formulate a federal rule of decision, and (2) the appropriateness of declaring a federal rule rather than borrowing, incorporating, or adopting state law in point. The Court preferred "the more modest terminology," questioning whether "the distinction between displacement of state law and displacement of federal law's incorporation of state law ever makes a practical difference." Ibid. Boyle made two further observations here significant. First, Boyle explained, the involvement of "an area of uniquely federal interest. . . establishes a necessary, not a sufficient, condition for the displacement of state law." Id., at 507. Second, in some cases, an "entire body of state law" may conflict with the federal interest and therefore require replacement. Id., at 508. But in others, the conflict is confined, and "only particular elements of state law are superseded." Ibid.

The dissent describes this case as pervasively federal, post, at 702, and "the provisions . . . here [as] just a few scattered islands in a sea of federal contractual provisions," post, at 709. But there is nothing "scattered" about the provisions on reimbursement and subrogation in the OPM-BCBSA master contract. See supra, at 684-685. Those provisions are linked together and depend upon a recovery from a third party under terms and conditions ordinarily governed by state law. See infra, at 698.^[5] The Court of [693] Appeals, whose decision we review, trained on the matter of reimbursement, not, as the dissent does, on FEHBA-authorized contracts at large. So focused, the appeals court determined that Empire has not demonstrated a "significant conflict . . . between an identifiable federal policy or interest and the operation of state law." 396 F. 3d, at 150 (Sack, J., concurring) (quoting Boyle, 487 U. S., at 507); see 396 F. 3d, at 140-141. Unless and until that showing is made, there is no cause to displace state law, much less to lodge this case in federal court.

B

We take up next Empire's Jackson Transit-derived argument, which is, essentially, a more tailored variation of the theme sounded in the dissent. It is undisputed that Congress has not expressly created a federal right of action enabling insurance carriers like Empire to sue health-care beneficiaries in federal court to enforce reimbursement rights under contracts contemplated by FEHBA. Empire and the United States nevertheless argue that, under our 1982 opinion in Jackson Transit, Empire's claim for reimbursement, arising under the contract between OPM and the BCBSA, "states a federal claim" because Congress intended all rights and duties stemming from that contract to be "federal in nature." Brief for United States as Amicus Curiae 12; see Brief for Petitioner 18-29. We are not persuaded by this argument.

The reliance placed by Empire and the United States on Jackson Transit is surprising, for that decision held there was no federal jurisdiction over the claim in suit. The federal statute there involved, § 13(c) of the Urban Mass Transportation Act of 1964 (UMTA), 78 Stat. 307 (then codified at 49 U. S. C. § 1609(c) (1976 ed.)), conditioned a governmental unit's receipt of federal funds to acquire a privately owned transit company on preservation of collective-bargaining rights enjoyed by the acquired company's employees. 457 U. S., at 17-18. The city of Jackson, Tennessee, with federal financial assistance, acquired a failing private bus company [694] and turned it into a public entity, the Jackson Transit Authority. Id., at 18. To satisfy the condition on federal aid, the transit authority entered into a "§ 13(c) agreement" with the union that represented the private company's employees, and the Secretary of Labor certified that agreement as "fair and equitable." Ibid. (internal quotation marks omitted).

For several years thereafter, the transit authority covered its unionized workers in a series of collective-bargaining agreements. Eventually, however, the Authority notified the union that it would no longer adhere to collective-bargaining undertakings. Id., at 19. The union commenced suit in federal court alleging breach of the § 13(c) agreement and of the latest collective-bargaining agreement. Ibid. This Court determined that the case did not arise under federal law, but was instead "governed by state law [to be] applied in state cour[t]." Id., at 29.

The Court acknowledged in Jackson Transit that "on several occasions [we had] determined that a plaintiff stated a federal claim when he sued to vindicate contractual rights set forth by federal statutes, [even though] the relevant statutes lacked express provisions creating federal causes of action." Id., at 22 (emphasis added) (citing Machinists v. Central Airlines, Inc., 372 U. S. 682 (1963) (union had a federal right of action to enforce an airline-adjustment-board award included in a collective-bargaining contract pursuant to a provision of the Railway Labor Act); Norfolk & Western R. Co. v. Nemitz, 404 U. S. 37 (1971) (railroad's employees stated federal claims when they sought to enforce assurances made by the railroad to secure Interstate Commerce Commission approval of a consolidation under a provision of the Interstate Commerce Act); Transamerica Mortgage Advisors, Inc. v. Lewis, 444 U. S. 11, 18-19 (1979) (permitting federal suit for rescission of a contract declared void by a provision of the Investment Advisers Act of 1940)). But prior decisions, we said, "d[id] not dictate the result in [the Jackson Transit] case," for in each case, "the critical factor" in determining "the scope of [695] rights and remedies under a federal statute . . . is the congressional intent behind the particular provision at issue." 457 U. S., at 22.

"In some ways," the Jackson Transit Court said, the UMTA "seem[ed] to make § 13(c) agreements and collective-bargaining contracts creatures of federal law." Id., at 23. In this regard, the Court noted, § 13(c)

"demand[ed] `fair and equitable arrangements' as prerequisites for federal aid; it require[d] the approval of the Secretary of Labor for those arrangements; it specifie[d] five different varieties of protective provisions that must be included among the § 13(c) arrangements; and it expressly incorporate[d] the protective arrangements into the grant contract between the recipient and the Federal Government." Ibid. (quoting 49 U. S. C. § 1609(c) (1976 ed.)).

But there were countervailing considerations. The Court observed that "labor relations between local governments and their employees are the subject of a longstanding statutory exemption from the National Labor Relations Act." 457 U. S., at 23. "Section 13(c)," the Court continued, "evince[d] no congressional intent to upset the decision in the [NLRA] to permit state law to govern the relationships between local governmental entities and the unions representing their employees." Id., at 23-24. Legislative history was corroborative. "A consistent theme," the Court found, "[ran] throughout the consideration of § 13(c): Congress intended that labor relations between transit workers and local governments would be controlled by state law." Id., at 24. We therefore held that the union had come to the wrong forum. Congress had indeed provided for § 13(c) agreements and collective-bargaining contracts stemming from them, but in the Court's judgment, the union's proper recourse for enforcement of those contracts was a suit in state court.

[696] Measured against the Court's discussion in Jackson Transit about when a claim arises under federal law, Empire's contract-derived claim for reimbursement is not a "creatur[e] of federal law." Id., at 23. True, distinctly federal interests are involved. Principally, reimbursements are credited to a federal fund, and the OPM-BCBSA master contract could be described as "federal in nature" because it is negotiated by a federal agency and concerns federal employees. See supra, at 683-684. But, as in Jackson Transit, countervailing considerations control. Among them, the reimbursement right in question, predicated on a FEHBA-authorized contract, is not a prescription of federal law. See supra, at 684. And, of prime importance, "Congress considered jurisdictional issues in enacting FEHBA[,] . . . confer[ring] federal jurisdiction where it found it necessary to do so." 396 F. 3d, at 145, n. 7.

FEHBA's jurisdictional provision, 5 U. S. C. § 8912, opens the federal district-court door to civil actions "against the United States." See supra, at 686. OPM's regulation, 5 CFR § 890.107(c) (2005), instructs enrollees who seek to challenge benefit denials to proceed in court against OPM "and not against the carrier or carrier's subcontractors." See ibid. Read together, these prescriptions "ensur[e] that suits brought by beneficiaries for denial of benefits will land in federal court." 396 F. 3d, at 145, n. 7. Had Congress found it necessary or proper to extend federal jurisdiction further, in particular, to encompass contract-derived reimbursement claims between carriers and insured workers, it would have been easy enough for Congress to say so. Cf. 29 U. S. C. § 1132(a)(3) (authorizing suit in federal court "by a participant, beneficiary, or fiduciary" of a pension or health plan governed by ERISA to gain redress for violations of "this subchapter or the terms of the plan"). We have no warrant to expand Congress' jurisdictional grant "by judicial decree." See Kokkonen v. Guardian Life Ins. Co. of America, 511 U. S. 375, 377 (1994).

[697] Jackson Transit, Empire points out, referred to decisions "demonstrat[ing] that . . . private parties in appropriate cases may sue in federal court to enforce contractual rights created by federal statutes." 457 U. S., at 22. See Brief for Petitioner 15. This case, however, involves no right created by federal statute. As just reiterated, while the OPM—BCBSA master contract provides for reimbursement, FEHBA's text itself contains no provision addressing the reimbursement or subrogation rights of carriers.

Nor do we read 5 U. S. C. § 8902(m)(1), FEHBA's preemption prescription, see supra, at 685-686, as a jurisdiction-conferring provision. That choice-of-law prescription is unusual in that it renders preemptive contract terms in health insurance plans, not provisions enacted by Congress. See 396 F. 3d, at 143-145; id., at 151 (Sack, J., concurring). A prescription of that unusual order warrants cautious interpretation.

Section 8902(m)(1) is a puzzling measure, open to more than one construction, and no prior decision seems to us precisely on point. Reading the reimbursement clause in the master OPM—BCBSA contract as a condition or limitation on "benefits" received by a federal employee, the clause could be ranked among "[contract] terms . . . relat[ing] to . . . coverage or benefits" and "payments with respect to benefits," thus falling within § 8902(m)(1)'s compass. See Brief for United States as Amicus Curiae 20; Reply Brief 8-9. On the other hand, a claim for reimbursement ordinarily arises long after "coverage" and "benefits" questions have been resolved, and corresponding "payments with respect to benefits" have been made to care providers or the insured. With that consideration in view, § 8902(m)(1)'s words may be read to refer to contract terms relating to the beneficiary's entitlement (or lack thereof) to Plan payment for certain healthcare services he or she has received, and not to terms relating to the carrier's postpayments right to reimbursement. See Brief for Julia Cruz as Amicus Curiae 10, 11.

[698] To decide this case, we need not choose between those plausible constructions. If contract-based reimbursement claims are not covered by FEHBA's preemption provision, then federal jurisdiction clearly does not exist. But even if FEHBA's preemption provision reaches contract-based reimbursement claims, that provision is not sufficiently broad to confer federal jurisdiction. If Congress intends a preemption instruction completely to displace ordinarily applicable state law, and to confer federal jurisdiction thereby, it may be expected to make that atypical intention clear. Cf. Columbus v. Ours Garage & Wrecker Service, Inc., 536 U. S. 424, 432-433 (2002) (citing Wisconsin Public Intervenor v. Mortier, 501 U. S. 597, 605 (1991)). Congress has not done so here.

Section 8902(m)(1)'s text does not purport to render inoperative any and all state laws that in some way bear on federal employee-benefit plans. Cf. 29 U. S. C. § 1144(a) (portions of ERISA "supersede any and all State laws insofar as they may now or hereafter relate to any employee benefit plan"). And, as just observed, see supra, at 697, given that § 8902(m)(1) declares no federal law preemptive, but instead, terms of an OPM—BCBSA negotiated contract, a modest reading of the provision is in order. Furthermore, a reimbursement right of the kind Empire here asserts stems from a personal-injury recovery, and the claim underlying that recovery is plainly governed by state law. We are not prepared to say, based on the presentations made in this case, that under § 8902(m)(1), an OPM—BCBSA contract term would displace every condition state law places on that recovery.

As earlier observed, the BCBSA Plan's statement of benefits links together the carrier's right to reimbursement from the insured and its right to subrogation. See supra, at 684-685. Empire's subrogation right allows the carrier, once it has paid an insured's medical expenses, to recover directly from a third party responsible for the insured's injury or [699] illness. See 16 G. Couch, Cyclopedia of Insurance Law § 61:1 (2d ed. 1982). Had Empire taken that course, no access to a federal forum could have been predicated on the OPM-BCBSA contract right. The tortfeasors' liability, whether to the insured or the insurer, would be governed not by an agreement to which the tortfeasors are strangers, but by state law, and § 8902(m)(1) would have no sway.

In sum, the presentations before us fail to establish that § 8902(m)(1) leaves no room for any state law potentially bearing on federal employee-benefit plans in general, or carrier-reimbursement claims in particular. Accordingly, we extract from § 8902(m)(1) no prescription for federal-court jurisdiction.

C

We turn finally to the argument that Empire's reimbursement claim, even if it does not qualify as a "cause of action created by federal law," nevertheless arises under federal law for § 1331 purposes, because federal law is "a necessary element of the [carrier's] claim for relief." Brief for United States as Amicus Curiae 25-26 (quoting Grable, 545 U. S., at 312, and Jones v. R. R. Donnelley & Sons Co., 541 U. S. 369, 376 (2004)). This case, we are satisfied, does not fit within the special and small category in which the United States would place it. We first describe Grable, a recent decision that the United States identifies as exemplary,^[6] and then explain why this case does not resemble that one.

Grable involved real property belonging to Grable & Sons Metal Products, Inc. (Grable), which the Internal Revenue Service (IRS) seized to satisfy a federal tax deficiency. 545 U. S., at 310. Grable received notice of the seizure by certified mail before the IRS sold the property to Darue Engineering & Manufacturing (Darue). Ibid. Five years later, [700] Grable sued Darue in state court to quiet title. Grable asserted that Darue's record title was invalid because the IRS had conveyed the seizure notice improperly. Id., at 311. The governing statute, 26 U. S. C. § 6335(a), provides that "notice in writing shall be given . . . to the owner of the property . . . or shall be left at his usual place of abode or business . . . ." Grable maintained that § 6335(a) required personal service, not service by certified mail. 545 U. S., at 311.

Darue removed the case to federal court. Alleging that Grable's claim of title depended on the interpretation of a federal statutory provision, i. e., § 6335(a) of the Internal Revenue Code, Darue invoked federal-question jurisdiction under 28 U. S. C. § 1331. We affirmed lower court determinations that the removal was proper. "The meaning of the federal tax provision," we said, "is an important issue of federal law that sensibly belongs in a federal court." 545 U. S., at 315. Whether Grable received notice adequate under § 6335(a), we observed, was "an essential element of [Grable's] quiet title claim"; indeed, "it appear[ed] to be the only. . . issue contested in the case." Ibid.

This case is poles apart from Grable. Cf. Brief for United States as Amicus Curiae 27. The dispute there centered on the action of a federal agency (IRS) and its compatibility with a federal statute, the question qualified as "substantial," and its resolution was both dispositive of the case and would be controlling in numerous other cases. See 545 U. S., at 313. Here, the reimbursement claim was triggered, not by the action of any federal department, agency, or service, but by the settlement of a personal-injury action launched in state court, see supra, at 687-688, and the bottom-line practical issue is the share of that settlement properly payable to Empire.

Grable presented a nearly "pure issue of law," one "that could be settled once and for all and thereafter would govern numerous tax sale cases." Hart and Wechsler 65 (2005 Supp.). In contrast, Empire's reimbursement claim, McVeigh's [701] Vecounsel represented without contradiction, is fact-bound and situation-specific. McVeigh contends that there were overcharges or duplicative charges by care providers, and seeks to determine whether particular services were properly attributed to the injuries caused by the 1997 accident and not rendered for a reason unrelated to the accident. See Tr. of Oral Arg. 44, 53.

The United States observes that a claim for reimbursement may also involve as an issue "[the] extent, if any, to which the reimbursement should take account of attorney's fees expended . . . to obtain the tort recovery." Brief as Amicus Curiae 29. Indeed it may. But it is hardly apparent why a proper "federal-state balance," see id., at 28, would place such a nonstatutory issue under the complete governance of federal law, to be declared in a federal forum. The state court in which the personal-injury suit was lodged is competent to apply federal law, to the extent it is relevant, and would seem best positioned to determine the lawyer's part in obtaining, and his or her fair share in, the tort recovery.

The United States no doubt "has an overwhelming interest in attracting able workers to the federal workforce," and "in the health and welfare of the federal workers upon whom it relies to carry out its functions." Id., at 10. But those interests, we are persuaded, do not warrant turning into a discrete and costly "federal case" an insurer's contract-derived claim to be reimbursed from the proceeds of a federal worker's state-court-initiated tort litigation.

In sum, Grable emphasized that it takes more than a federal element "to open the `arising under' door." 545 U. S., at 313. This case cannot be squeezed into the slim category Grable exemplifies.

* * *

For the reasons stated, the judgment of the Court of Appeals for the Second Circuit is

Affirmed.

[702] JUSTICE BREYER, with whom JUSTICE KENNEDY, JUSTICE SOUTER, and JUSTICE ALITO join, dissenting.

This case involves a dispute about the meaning of terms in a federal health insurance contract. The contract, between a federal agency and a private carrier, sets forth the details of a federal health insurance program created by federal statute and covering 8 million federal employees. In all this the Court cannot find a basis for federal jurisdiction. I believe I can. See Clearfield Trust Co. v. United States, 318 U. S. 363 (1943).

I

A

There is little about this case that is not federal. The comprehensive federal health insurance program at issue is created by a federal statute, the Federal Employees Health Benefits Act of 1959 (FEHBA), 5 U. S. C. § 8901 et seq. (2000 ed. and Supp. III). This program provides insurance for Federal Government employees and their families. That insurance program today covers approximately 8 million federal employees, retirees, and dependents, at a total cost to the Government of about $22 billion a year. Brief for United States as Amicus Curiae 2.

To implement the statute, the Office of Personnel Management (OPM), the relevant federal agency, enters into contracts with a handful of major insurance carriers. These agency/carrier contracts follow a standard agency form of about 38,000 words, and contain the details of the plan offered by the carrier. See § 8902(d) (2000 ed.) (requiring contract between carrier and agency to contain a detailed statement of the terms of the plan); see also Federal Employees Health Benefits Program Standard Contract (CR—2003) (2005), online at http://www.opm.gov/insure/carriers/samplecontract.doc (sample form agency/carrier contract) (as visited June 7, 2006, and available in Clerk of Court's case file). The contract lists, for example, the benefits provided to the employees who enroll. It provides a patient's bill of [703] rights. It makes clear that the Government, not the carrier, will receive the premiums and will pay the benefits. It specifies that the carrier will administer the program that the contract sets forth, for which the carrier will receive an adjustable fee. The contract also states, "By enrolling or accepting services under this contract, [enrollees] are obligated to all terms, conditions, and provisions of this contract." App. 90.

As the statute requires, § 8907(b), the agency/carrier contract also provides that the carrier will send each enrolled employee a brochure that explains the terms of the plan, as set forth in the contract. The brochure explains that it "describes the benefits of the . . . [p]lan under [the carrier's] contract . . . with [the federal agency], as authorized by the [federal statute]." Id., at 158. The terms of the brochure are incorporated into the agency/carrier contract. Id., at 89. The carrier distributes the brochure with a seal attached to the front stating, "Authorized for distribution by the United States Office of Personnel Management Retirement and Insurance Service." Id., at 155.

The program is largely funded by the Federal Government. More specifically, the Federal Government pays about 75% of the plan premiums; the enrollee pays the rest. § 8906(b). These premiums are deposited into a special fund in the United States Treasury. § 8909(a). The carrier typically withdraws money from the fund to pay for covered health care services, ibid.; however, the fund's money belongs, not to the carrier, but to the federal agency that administers the program. After benefits are paid, any surplus in the fund can be used at the agency's discretion to reduce premiums, to increase plan benefits, or to make a refund to the Government and enrollees. § 8909(b); 5 CFR § 890.503(c)(2) (2005). The carrier is not at risk. Rather, it earns a profit, not from any difference between plan premiums and the cost of benefits, but from a negotiated service charge that the federal agency pays directly.

[704] Federal regulations provide that the federal agency will resolve disputes about an enrolled employee's coverage. § 890.105(a)(1); see also 5 U. S. C. § 8902(j) (requiring carrier to provide health benefit if OPM concludes that enrollee is entitled to the benefit under the contract). The agency's resolution is judicially reviewable under the Administrative Procedure Act in federal court. 5 CFR § 890.107 (2005).

In sum, the statute is federal, the program it creates is federal, the program's beneficiaries are federal employees working throughout the country, the Federal Government pays all relevant costs, and the Federal Government receives all relevant payments. The private carrier's only role in this scheme is to administer the health benefits plan for the federal agency in exchange for a fixed service charge.

B

The plan at issue here, the Blue Cross Blue Shield Service Benefit Plan, is the largest in the statutory program. The plan's details are contained in Blue Cross Blue Shield's contract with the federal agency and in the brochure, which binds the enrolled employee to that contract. In this case, the carrier seeks to require the enrolled employee's estate to abide by provisions that permit the carrier to obtain (and require the enrolled employee to pay) reimbursement from an enrollee for benefits provided if the enrollee recovers money from a third party (as compensation for the relevant injury or illness). The parties dispute the proper application of some of those provisions.

First, the agency's contract with the carrier requires the carrier to "mak[e] a reasonable effort to seek recovery of amounts to which it is entitled to recover." App. 95. And the carrier must do so "under a single, nation-wide policy to ensure equitable and consistent treatment for all [enrollees] under this contract." Ibid. Any money recovered by the carrier goes into the statutory fund in the United States [705] Treasury, and may be spent for the benefit of the program at the discretion of the federal agency. See supra, at 703.

Second, the agency/carrier contract and the brochure set forth the enrollee's obligation to reimburse the carrier under certain circumstances. The contract states, "The Carrier may . . . recover directly from the [enrollee] all amounts received by the [enrollee] by suit, settlement, or otherwise from any third party or its insurer . . . for benefits which have also been paid under this contract." App. 95. The agency/carrier contract also says that the "[c]arrier's subrogation rights, procedures and policies, including recovery rights, shall be in accordance with the provisions of the agreed-upon brochure text." Id., at 100. The relevant provisions in the brochure (which also appear in the appendix to the agency/carrier contract) tell the enrollee:

"If another person or entity, through an act or omission, causes you to suffer an injury or illness, and if we pay benefits for that injury or illness, you must agree to the following:

"All recoveries you obtain (whether by lawsuit, settlement, or otherwise), no matter how described or designated, must be used to reimburse us in full for benefits we paid. . . .

"We will not reduce our share of any recovery unless we agree in writing to a reduction, . . . because you had to pay attorneys' fees." Id., at 165.

The enrollee must abide by these requirements because, as explained above, the brochure tells the beneficiary that, by enrolling in the program, he or she is agreeing to the terms of the brochure, which in turn "describes the benefits of the [plan] under [the agency/carrier] contract." Id., at 158.

II

A

I have explained the nature of the program and have set forth the terms of the agency/carrier contract in some detail [706] because, once understood, their federal nature brings this case well within the scope of the relevant federal jurisdictional statute, 28 U. S. C. § 1331, which provides jurisdiction for claims "arising under" federal law. For purposes of this statute, a claim arises under federal law if federal law creates the cause of action. Merrell Dow Pharmaceuticals Inc. v. Thompson, 478 U. S. 804, 808 (1986); see also American Well Works Co. v. Layne & Bowler Co., 241 U. S. 257, 260 (1916) (opinion of Holmes, J.) (A "suit arises under the law that creates the cause of action"). And this Court has explained that § 1331's "statutory grant of `jurisdiction will support claims founded upon federal common law as well as those of a statutory origin.'" National Farmers Union Ins. Cos. v. Crow Tribe, 471 U. S. 845, 850 (1985); see also Illinois v. Milwaukee, 406 U. S. 91 (1972); 19 C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure § 4514, p. 455 (2d ed. 1996) ("A case `arising under' federal common law presents a federal question and as such is within the original subject-matter jurisdiction of the federal courts"). In other words, "[f]ederal common law as articulated in rules that are fashioned by court decisions are `laws' as that term is used in § 1331." National Farmers, supra, at 850.

It seems clear to me that the petitioner's claim arises under federal common law. The dispute concerns the application of terms in a federal contract. This Court has consistently held that "obligations to and rights of the United States under its contracts are governed exclusively by federal law." Boyle v. United Technologies Corp., 487 U. S. 500, 504 (1988). This principle dates back at least as far as Clearfield Trust, 318 U. S., at 366, where the Court held that the "rights and duties of the United States on [federal] commercial paper," namely a federal employee's paycheck, "are governed by federal rather than local law." The Court reasoned that "[w]hen the United States disburses its funds or [707] pays its debts, it is exercising a constitutional function or power," a power "in no way dependent on the laws of Pennsylvania or of any other state." Ibid. Accordingly, "[i]n [the] absence of an applicable Act of Congress it is for the federal courts to fashion the governing rule of law." Id., at 367.

This Court has applied this principle, the principle embodied in Clearfield Trust, to Government contracts of all sorts. See, e. g., West Virginia v. United States, 479 U. S. 305, 308-309 (1987) (contract regarding federal disaster relief efforts); United States v. Kimbell Foods, Inc., 440 U. S. 715, 726 (1979) (contractual liens arising from federal loan programs); United States v. Little Lake Misere Land Co., 412 U. S. 580, 592 (1973) (agreements to acquire land under federal conservation program); United States v. Seckinger, 397 U. S. 203, 209 (1970) (Government construction contracts); United States v. County of Allegheny, 322 U. S. 174, 183 (1944) (Government procurement contracts).

In this case, the words that provide the right to recover are contained in the brochure, which in turn explains the provisions of the contract between the Government and the carrier, provisions that were written by a federal agency acting pursuant to a federal statute that creates a federal benefit program for federal employees. At bottom, then, the petitioner's claim is based on the interpretation of a federal contract, and as such should be governed by federal common law. And because the petitioner's claim is based on federal common law, the federal courts have jurisdiction over it pursuant to § 1331. The lower federal courts have similarly found § 1331 jurisdiction over suits between private parties based on Federal Government contracts. See, e. g., Downey v. State Farm Fire & Casualty Co., 266 F.3d 675, 680-681 (CA7 2001) (Easterbrook, J.) (National Flood Insurance Program contracts); Almond v. Capital Properties, Inc., 212 F.3d 20, 22-24 (CA1 2000) (Boudin, J.) (Federal Railroad Administration [708] contract); Price v. Pierce, 823 F. 2d 1114, 1119-1120 (CA7 1987) (Posner, J.) (Dept. of Housing and Urban Development contracts).

B

What might one say to the contrary? First, I may have made too absolute a statement in claiming that disputes arising under federal common law are (for jurisdictional purposes) cases "arising under" federal law. After all, in every Supreme Court case I have cited (except National Farmers and Milwaukee, and not including the Courts of Appeals cases), the United States was a party, and that fact provides an independent basis for jurisdiction. See 28 U. S. C. §§ 1345, 1346(a)(2), 1491(a)(1). In those cases the decision to apply federal common law was, therefore, a "choice-of-law issue" only, ante, at 691, and the Court consequently did not need to address the application of the Clearfield Trust doctrine to § 1331 "arising under" jurisdiction.

But I have found no case where a federal court concluded that federal common law governed a plaintiff's contract claim but nevertheless decided that the claim did not arise under federal law. I have found several lower court cases (cited supra, at 707 and this page) where courts asserted § 1331 jurisdiction solely on the basis of federal common law. And in Machinists v. Central Airlines, Inc., 372 U. S. 682, 693, n. 17 (1963), this Court cited the Clearfield Trust cases in finding § 1331 jurisdiction over the contract suit before it, noting that although those cases "did not involve federal jurisdiction as such," nevertheless "they are suggestive" on the issue of § 1331 jurisdiction over suits involving Federal Government contracts "since they hold federal law determinative of the merits of the claim."

It is enough here, however, to assume that federal common law means federal jurisdiction where Congress so intends. Cf. Clearfield Trust, supra, at 367 ("In absence of an applicable Act of Congress it is for the federal courts to fashion the governing rule of law according to their own standards" (emphasis added)). If so, there are strong reasons for the [709] federal courts, following Clearfield Trust, to assume jurisdiction and apply federal common law to resolve this case.

First, although the nominal plaintiff in this case is the carrier, the real party in interest is the United States. Any funds that the petitioner recovers here it must pay directly to the United States, by depositing those funds in the FEHBA United States Treasury account managed by the federal agency. The carrier simply administers the reimbursement proceeding for the United States, just as it administers the rest of the agency/carrier contract. Accordingly, this case, just like the Clearfield Trust cases, concerns the "rights of the United States under its contracts." Boyle, 487 U. S., at 504.

Second, the health insurance system FEHBA establishes is a federal program. The Federal Government pays for the benefits, receives the premiums, and resolves disputes over claims for medical services. Given this role, the Federal Government's need for uniform interpretation of the contract is great. Given the spread of Government employees throughout the Nation and the unfairness of treating similar employees differently, the employees' need for uniform interpretation is equally great. That interest in uniformity calls for application of federal common law to disputes about the meaning of the words in the agency/carrier contract and brochure. See Clearfield Trust, 318 U. S., at 367 (applying federal common law because the "desirability of a uniform [federal] rule is plain"); see also Bank of America Nat. Trust & Sav. Assn. v. Parnell, 352 U. S. 29, 33, 34 (1956) ("[L]itigation with respect to Government paper. . . between private parties" may nevertheless "be governed by federal [common] law" where there is "the presence of a federal interest"). And that interest in uniformity also suggests that the doors of the federal courts should be open to decide such disputes.

Third, as discussed above, the provisions at issue here are just a few scattered islands in a sea of federal contractual provisions, all of which federal courts will interpret and [710] apply (when reviewing the federal agency's resolution of disputes regarding benefits). Given this context, why would Congress have wanted the courts to treat those islands any differently? I can find no convincing answer.

Regardless, the majority and the Court of Appeals believe they have come up with one possible indication of a contrary congressional intent. They believe that the statute's jurisdictional provision argues against federal jurisdiction where the United States is not formally a party. That provision gives the federal district courts "original jurisdiction, concurrent with the United States Court of Federal Claims, of a civil action or claim against the United States founded on this chapter." 5 U. S. C. § 8912. According to the majority, if Congress had wanted cases like this one to be brought in the federal courts, it would have extended § 8912 to cover them. Ante, at 696.

That is not so. Congress' failure to write § 8912 to include suits between carriers and enrollees over plan provisions may reflect inadvertence. Or it may reflect a belief that § 1331 covered such cases regardless. Either way, § 8912 tells us nothing about Congress' intent in respect to § 1331 jurisdiction.

But why then did Congress write § 8912 at all? After all, the cases there covered—contract claims against the Federal Government "founded on" the federal health insurance program—would also be governed by federal common law and (if my view is correct) would have fallen within the scope of § 1331. What need would there have been (if my view is correct) to write a special section, § 8912, expanding federal jurisdiction to encompass these claims?

The answer, as the majority itself points out, ante, at 686, is that Congress did not write § 8912 to expand the jurisdiction of the federal courts. It wrote that section to transfer a category of suits (claims against the United States exceeding $10,000) from one federal court (the Court of Federal Claims) to others (the federal district courts).

[711] In sum, given Clearfield Trust, supra, and its progeny, there is every reason to believe that federal common law governs disputes concerning the agency/carrier contract. And that is so even though "it would have been easy enough for Congress to say" that federal common law should govern these claims. See ante, at 696. After all, no such express statement of congressional intent was present in Clearfield Trust itself, or in any of the cases relying on Clearfield Trust for the authority to apply federal common law to interpret Government contracts. See, e. g., cases cited supra, at 707; see also Clearfield Trust, supra, at 367 ("In absence of an applicable Act of Congress it is for the federal courts to fashion the governing rule of law according to their own standards"). Accordingly, I would apply federal common law to resolve the petitioner's contract claim. And, as explained above, when the "governing rule of law" on which a claim is based is federal common law, then the federal courts have jurisdiction over that claim under § 1331.

C

The Court adds that, in spite of the pervasively federal character of this dispute, state law should govern it because the petitioner has not demonstrated a "`significant conflict. . . between an identifiable federal policy or interest and the operation of state law.'" Ante, at 693. But as I have explained, see supra, at 708-709, the Federal Government has two such interests: (1) the uniform operation of a federal employee health insurance program, and (2) obtaining reimbursement under a uniform set of legal rules. These interests are undermined if the amount a federal employee has to reimburse the FEHBA United States Treasury fund in cases like this one varies from State to State in accordance with state contract law. We have in the past recognized that this sort of interest in uniformity is sufficient to warrant application of federal common law. See, e. g., Boyle, supra, at 508 ("[W]here the federal interest requires a uniform rule, [712] the entire body of state law applicable to the area conflicts and is replaced by federal rules"); Kimbell Foods, 440 U. S., at 728 ("Undoubtedly, federal programs that `by their nature are and must be uniform in character throughout the Nation' necessitate formulation of controlling federal rules"); Clearfield Trust, 318 U. S., at 367 (applying federal common law because "application of state law . . . would subject the rights and duties of the United States to exceptional uncertainty" and "would lead to great diversity in results by making identical transactions subject to the vagaries of the laws of the several states," and therefore "[t]he desirability of a uniform rule is plain").

But even if the Court is correct that "`[t]he prudent course'" is "`to adopt the readymade body of state law as the federal rule of decision until Congress strikes a different accommodation,'" ante, at 691-692 (quoting Kimbell Foods, supra, at 740), there would still be federal jurisdiction over this case. That is because, as Clearfield Trust, Kimbell Foods, and other cases make clear, the decision to apply state law "as the federal rule of decision" is itself a matter of federal common law. See, e. g., Kimbell Foods, supra, at 728, n. 21 ("`Whether state law is to be incorporated as a matter of federal common law . . . involves the . . . problem of the relationship of a particular issue to a going federal program'" (emphasis added)); Clearfield Trust, supra, at 367 ("In our choice of the applicable federal rule we have occasionally selected state law" (emphasis added)); see also R. Fallon, D. Meltzer, & D. Shapiro, Hart and Wechsler's The Federal Courts and the Federal System 700 (5th ed. 2003) ("[T]he current approach, as reflected in [Kimbell Foods, supra], suggests that . . . while under Clearfield federal common law governs, in general it will incorporate state law as the rule of decision"); 19 C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure § 4518, at 572-573 ("In recent years, the Supreme Court has put increasing emphasis on the notion that when determining what should be the content [713] of federal common law, the law of the forum state should be adopted absent some good reason to displace it" (emphasis added; citing Kimbell Foods, supra, and Clearfield Trust, supra)).

On this view, the Clearfield Trust inquiry involves two questions: (1) whether federal common law governs the plaintiff's claim; (2) if so, whether, as a matter of federal common law, the Court should adopt state law as the proper "`federal rule of decision,'" ante, at 692 (emphasis added). See, e. g., Kimbell Foods, supra, at 727 (deciding that "[f]ederal law therefore controls" the dispute but concluding that state law gives "content to this federal rule"); United States v. Little Lake Misere Land Co., 412 U. S., at 593-594 (The "first step of the Clearfield analysis" is to decide whether "`the courts of the United States may formulate a rule of decision,'" and the "next step in our analysis is to determine whether" the federal rule of decision should "`borro[w]' state law"); see also Friendly, In Praise of Erie—and of the New Federal Common Law, 39 N. Y. U. L. Rev. 383, 410 (1964) ("Clearfield decided not one issue but two. The first . . . is that the right of the United States to recover for conversion of a Government check is a federal right, so that the courts of the United States may formulate a rule of decision. The second . . . is whether, having this opportunity, the federal courts should adopt a uniform nation-wide rule or should follow state law" (footnote omitted)). Therefore, even if the Court is correct that state law applies to claims involving the interpretation of some provisions of this contract, the decision whether and when to apply state law should be made by the federal courts under federal common law. Accordingly, for jurisdictional purposes those claims must still arise under federal law, for federal common law determines the rule of decision.

Finally, the footnote in Boyle cited by the Court did not purport to overrule Clearfield Trust on this point. See Boyle, 487 U. S., at 507, n. 3 ("If the distinction between displacement [714] of state law and displacement of federal law's incorporation of state law ever makes a practical difference, it at least does not do so in the present case").

With respect, I dissent.

[1] Clinton A. Krislov and Michael R. Karnuth filed a brief for Julia Cruz, as representative of Jose S. Cruz, as amicus curiae urging affirmance.

[2]The statement of benefits further provides:

"You must tell us promptly if you have a claim against another party for a condition that we have paid or may pay benefits for, and you must tell us about any recoveries you obtain, whether in or out of court. We may seek a lien on the proceeds of your claim in order to reimburse ourselves to the full amount of benefits we have paid or will pay.

"We may request that you assign to us (1) your right to bring an action or (2) your right to the proceeds of a claim for your illness or injury. We may delay processing of your claims until you provide the assignment.

"Note: We will pay the costs of any covered services you receive that are in excess of any recoveries made." App. 165.

[3] At oral argument, counsel for respondent McVeigh represented that "most of the [reimbursement claims] are not of th[is] magnitude"; "[m]ost of the cases involve [amounts like] $5,500 and $6,500." Tr. of Oral Arg. 52.

[4] The United States, in accord with the dissent in this regard, see post, at 707, several times cites United States v. County of Allegheny, 322 U. S. 174 (1944), see, e. g., Brief as Amicus Curiae 10, 15, 26, maintaining that the construction of a federal contract "necessarily present[s] questions of `federal law not controlled by the law of any State,'" id., at 26 (quoting 322 U. S., at 183). Allegheny does not stretch as widely as the United States suggests. That case concerned whether certain property belonged to the United States and, if so, whether the incidence of a state tax was on the United States or on a Government contractor. See id., at 181-183, 186-189. Neither the United States nor any United States agency is a party to this case, and the auxiliary matter here involved scarcely resembles the controversy in Allegheny.

[5] The dissent nowhere suggests that uniform, court-declared federal law would govern the carrier's subrogation claim against the tortfeasor. Nor does the dissent explain why the two linked provisions—reimbursement and subrogation—should be decoupled.

[6] As the Court in Grable observed, 545 U. S., at 312, the classic example of federal-question jurisdiction predicated on the centrality of a federal issue is Smith v. Kansas City Title & Trust Co., 255 U. S. 180 (1921).

CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT.

[717] Willard P. Owens argued the cause for petitioner. With him on the brief were E. H. Rayson and R. R. Kramer.

Clarence Walker argued the cause for respondent. With him on the brief was William Ables, Jr.

MR. JUSTICE BRENNAN delivered the opinion of the Court.

Respondent Paul Gibbs was awarded compensatory and punitive damages in this action against petitioner United Mine Workers of America (UMW) for alleged violations of § 303 of the Labor Management Relations Act, 1947, 61 Stat. 158, as amended,^[1] and of the common law of [718] Tennessee. The case grew out of the rivalry between the United Mine Workers and the Southern Labor Union over representation of workers in the southern Appalachian coal fields. Tennessee Consolidated Coal Company, not a party here, laid off 100 miners of the UMW's Local 5881 when it closed one of its mines in southern Tennessee during the spring of 1960. Late that summer, Grundy Company, a wholly owned subsidiary of Consolidated, hired respondent as mine superintendent to attempt to open a new mine on Consolidated's property at nearby Gray's Creek through use of members of the Southern Labor Union. As part of the arrangement, Grundy also gave respondent a contract to haul the mine's coal to the nearest railroad loading point.

On August 15 and 16, 1960, armed members of Local 5881 forcibly prevented the opening of the mine, threatening respondent and beating an organizer for the rival union.^[2] The members of the local believed Consolidated [719] had promised them the jobs at the new mine; they insisted that if anyone would do the work, they would. At this time, no representative of the UMW, their international union, was present. George Gilbert, the UMW's field representative for the area including Local 5881, was away at Middlesboro, Kentucky, attending an Executive Board meeting when the members of the local discovered Grundy's plan;^[3] he did not return to the area until late in the day of August 16. There was uncontradicted testimony that he first learned of the violence while at the meeting, and returned with explicit instructions from his international union superiors to establish a limited picket line, to prevent any further violence, and to see to it that the strike did not spread to neighboring mines. There was no further violence at the mine site; a picket line was maintained there for nine months; and no further attempts were made to open the mine during that period.^[4]

[720] Respondent lost his job as superintendent, and never entered into performance of his haulage contract. He testified that he soon began to lose other trucking contracts and mine leases he held in nearby areas. Claiming these effects to be the result of a concerted union plan against him, he sought recovery not against Local 5881 or its members, but only against petitioner, the international union. The suit was brought in the United States District Court for the Eastern District of Tennessee, and jurisdiction was premised on allegations of secondary boycotts under § 303. The state law claim, for which jurisdiction was based upon the doctrine of pendent jurisdiction, asserted "an unlawful conspiracy and an unlawful boycott aimed at him and [Grundy] to maliciously, wantonly and willfully interfere with his contract of employment and with his contract of haulage."^[5]

The trial judge refused to submit to the jury the claims of pressure intended to cause mining firms other than Grundy to cease doing business with Gibbs; he found those claims unsupported by the evidence. The jury's verdict was that the UMW had violated both § 303 and state law. Gibbs was awarded $60,000 as damages under the employment contract and $14,500 under the haulage contract; he was also awarded $100,000 punitive damages. On motion, the trial court set aside the award of damages with respect to the haulage contract on the ground that damage was unproved. It also held that union pressure on Grundy to discharge respondent as supervisor would constitute only a primary dispute with Grundy, as respondent's employer, and hence was not cognizable as a claim under § 303. Interference with the [721] employment relationship was cognizable as a state claim, however, and a remitted award was sustained on the state law claim.^[6] 220 F. Supp. 871. The Court of Appeals for the Sixth Circuit affirmed. 343 F. 2d 609. We granted certiorari. 382 U. S. 809. We reverse.

I.

A threshold question is whether the District Court properly entertained jurisdiction of the claim based on Tennessee law. There was no need to decide a like question in Teamsters Union v. Morton, 377 U. S. 252, since the pertinent state claim there was based on peaceful secondary activities and we held that state law based on such activities had been pre-empted by § 303. But here respondent's claim is based in part on proofs of violence and intimidation. "[W]e have allowed the States to grant compensation for the consequences, as defined by the traditional law of torts, of conduct marked by violence and imminent threats to the public order. United Automobile Workers v. Russell, 356 U. S. 634; United Construction Workers v. Laburnum Corp., 347 U. S. 656. . . . State jurisdiction has prevailed in these situations because the compelling state interest, in the scheme of our federalism, in the maintenance of domestic peace is not overridden in the absence of clearly expressed congressional direction." San Diego Building Trades Council v. Garmon, 359 U. S. 236, 247.

[722] The fact that state remedies were not entirely pre-empted does not, however, answer the question whether the state claim was properly adjudicated in the District Court absent diversity jurisdiction. The Court held in Hurn v. Oursler, 289 U. S. 238, that state law claims are appropriate for federal court determination if they form a separate but parallel ground for relief also sought in a substantial claim based on federal law. The Court distinguished permissible from nonpermissible exercises of federal judicial power over state law claims by contrasting "a case where two distinct grounds in support of a single cause of action are alleged, one only of which presents a federal question, and a case where two separate and distinct causes of action are alleged, one only of which is federal in character. In the former, where the federal question averred is not plainly wanting in substance, the federal court, even though the federal ground be not established, may nevertheless retain and dispose of the case upon the non-federal ground; in the latter it may not do so upon the non-federal cause of action." 289 U. S., at 246. The question is into which category the present action fell.

Hurn was decided in 1933, before the unification of law and equity by the Federal Rules of Civil Procedure. At the time, the meaning of "cause of action" was a subject of serious dispute;^[7] the phrase might "mean one thing for one purpose and something different for another." [723] United States v. Memphis Cotton Oil Co., 288 U. S. 62, 67-68.^[8] The Court in Hurn identified what it meant by the term by citation of Baltimore S. S. Co. v. Phillips, 274 U. S. 316, a case in which "cause of action" had been used to identify the operative scope of the doctrine of res judicata. In that case the Court had noted that " `the whole tendency of our decisions is to require a plaintiff to try his whole cause of action and his whole case at one time.' " 274 U. S., at 320. It stated its holding in the following language, quoted in part in the Hurn opinion:

"Upon principle, it is perfectly plain that the respondent [a seaman suing for an injury sustained while working aboard ship] suffered but one actionable wrong and was entitled to but one recovery, whether his injury was due to one or the other of several distinct acts of alleged negligence or to a combination of some or all of them. In either view, there would be but a single wrongful invasion of a single primary right of the plaintiff, namely, the right of bodily safety, whether the acts constituting such invasion were one or many, simple or complex.

"A cause of action does not consist of facts, but of the unlawful violation of a right which the facts show. The number and variety of the facts alleged do not establish more than one cause of action so long as their result, whether they be considered severally or in combination, is the violation of but one right by a single legal wrong. The mere multiplication of grounds of negligence alleged as causing the same injury does not result in multiplying the causes of action. `The facts are merely the means, [724] and not the end. They do not constitute the cause of action, but they show its existence by making the wrong appear.' " Id., at 321.

Had the Court found a jurisdictional bar to reaching the state claim in Hurn, we assume that the doctrine of res judicata would not have been applicable in any subsequent state suit. But the citation of Baltimore S. S. Co. shows that the Court found that the weighty policies of judicial economy and fairness to parties reflected in res judicata doctrine were in themselves strong counsel for the adoption of a rule which would permit federal courts to dispose of the state as well as the federal claims.

With the adoption of the Federal Rules of Civil Procedure and the unified form of action, Fed. Rule Civ. Proc. 2, much of the controversy over "cause of action" abated. The phrase remained as the keystone of the Hurn test, however, and, as commentators have noted,^[9] has been the source of considerable confusion. Under the Rules, the impulse is toward entertaining the broadest possible scope of action consistent with fairness to the parties; joinder of claims, parties and remedies is strongly encouraged.^[10] Yet because the Hurn question involves issues of jurisdiction as well as convenience, there has been some tendency to limit its application to cases in which the state and federal claims are, as in Hurn, "little more than the equivalent of different epithets to characterize the same group of circumstances." 289 U. S., at 246.^[11]

[725] This limited approach is unnecessarily grudging Pendent jurisdiction, in the sense of judicial power, exists whenever there is a claim "arising under [the] Constitution, the Laws of the United States, and Treaties made, or which shall be made, under their Authority . . . ," U. S. Const., Art. III, § 2, and the relationship between that claim and the state claim permits the conclusion that the entire action before the court comprises but one constitutional "case."^[12] The federal claim must have substance sufficient to confer subject matter jurisdiction on the court. Levering & Garrigues Co. v. Morrin, 289 U. S. 103. The state and federal claims must derive from a common nucleus of operative fact. But if, considered without regard to their federal or state character, a plaintiff's claims are such that he would ordinarily be expected to try them all in one judicial proceeding, then, assuming substantiality of the federal issues, there is power in federal courts to hear the whole.^[13]

[726] That power need not be exercised in every case in which it is found to exist. It has consistently been recognized that pendent jurisdiction is a doctrine of discretion, not of plaintiff's right.^[14] Its justification lies in considerations of judicial economy, convenience and fairness to litigants; if these are not present a federal court should hesitate to exercise jurisdiction over state claims, even though bound to apply state law to them, Erie R. Co. v. Tompkins, 304 U. S. 64. Needless decisions of state law should be avoided both as a matter of comity and to promote justice between the parties, by procuring for them a surer-footed reading of applicable law.^[15] Certainly, if the federal claims are dismissed before trial, even though not insubstantial in a jurisdictional sense, the state claims should be dismissed as well.^[16] Similarly, if it appears that the state issues substantially predominate, whether in terms of proof, of the scope of the issues raised, or of the comprehensiveness of the remedy sought, the state claims may be dismissed without prejudice and [727] left for resolution to state tribunals. There may, on the other hand, be situations in which the state claim is so closely tied to questions of federal policy that the argument for exercise of pendent jurisdiction is particularly strong. In the present case, for example, the allowable scope of the state claim implicates the federal doctrine of pre-emption; while this interrelationship does not create statutory federal question jurisdiction, Louisville & N. R. Co. v. Mottley, 211 U. S. 149, its existence is relevant to the exercise of discretion. Finally, there may be reasons independent of jurisdictional considerations, such as the likelihood of jury confusion in treating divergent legal theories of relief, that would justify separating state and federal claims for trial, Fed. Rule Civ. Proc. 42 (b). If so, jurisdiction should ordinarily be refused.

The question of power will ordinarily be resolved on the pleadings. But the issue whether pendent jurisdiction has been properly assumed is one which remains open throughout the litigation. Pretrial procedures or even the trial itself may reveal a substantial hegemony of state law claims, or likelihood of jury confusion, which could not have been anticipated at the pleading stage. Although it will of course be appropriate to take account in this circumstance of the already completed course of the litigation, dismissal of the state claim might even then be merited. For example, it may appear that the plaintiff was well aware of the nature of his proofs and the relative importance of his claims; recognition of a federal court's wide latitude to decide ancillary questions of state law does not imply that it must tolerate a litigant's effort to impose upon it what is in effect only a state law case. Once it appears that a state claim constitutes the real body of a case, to which the federal claim is only an appendage, the state claim may fairly be dismissed.

[728] We are not prepared to say that in the present case the District Court exceeded its discretion in proceeding to judgment on the state claim. We may assume for purposes of decision that the District Court was correct in its holding that the claim of pressure on Grundy to terminate the employment contract was outside the purview of § 303. Even so, the § 303 claims based on secondary pressures on Grundy relative to the haulage contract and on other coal operators generally were substantial. Although § 303 limited recovery to compensatory damages based on secondary pressures, Teamsters Union v. Morton, supra, and state law allowed both compensatory and punitive damages, and allowed such damages as to both secondary and primary activity, the state and federal claims arose from the same nucleus of operative fact and reflected alternative remedies. Indeed, the verdict sheet sent in to the jury authorized only one award of damages, so that recovery could not be given separately on the federal and state claims.

It is true that the § 303 claims ultimately failed and that the only recovery allowed respondent was on the state claim. We cannot confidently say, however, that the federal issues were so remote or played such a minor role at the trial that in effect the state claim only was tried. Although the District Court dismissed as unproved the § 303 claims that petitioner's secondary activities included attempts to induce coal operators other than Grundy to cease doing business with respondent, the court submitted the § 303 claims relating to Grundy to the jury. The jury returned verdicts against petitioner on those § 303 claims, and it was only on petitioner's motion for a directed verdict and a judgment n. o. v. that the verdicts on those claims were set aside. The District Judge considered the claim as to the haulage [729] contract proved as to liability, and held it failed only for lack of proof of damages. Although there was some risk of confusing the jury in joining the state and federal claims—especially since, as will be developed, differing standards of proof of UMW involvement applied— the possibility of confusion could be lessened by employing a special verdict form, as the District Court did. Moreover, the question whether the permissible scope of the state claim was limited by the doctrine of pre-emption afforded a special reason for the exercise of pendent jurisdiction; the federal courts are particularly appropriate bodies for the application of pre-emption principles. We thus conclude that although it may be that the District Court might, in its sound discretion, have dismissed the state claim, the circumstances show no error in refusing to do so.

II.

This Court has consistently recognized the right of States to deal with violence and threats of violence appearing in labor disputes, sustaining a variety of remedial measures against the contention that state law was pre-empted by the passage of federal labor legislation. Allen-Bradley Local v. Wisconsin Board, 315 U. S. 740; United Construction Workers v. Laburnum Construction Corp., 347 U. S. 656; United Automobile Workers v. Wisconsin Board, 351 U. S. 266; Youngdahl v. Rainfair, Inc., 355 U. S. 131; United Automobile Workers v. Russell, 356 U. S. 634. Petitioner concedes the principle, but argues that the permissible scope of state remedies in this area is strictly confined to the direct consequences of such conduct, and does not include consequences resulting from associated peaceful picketing or other union activity. We agree.

Our opinions on this subject, frequently announced over weighty arguments in dissent that state remedies [730] were being given too broad scope, have approved only remedies carefully limited to the protection of the compelling state interest in the maintenance of domestic peace. Thus, in San Diego Building Trades Council v. Garmon, 359 U. S. 236, we read our prior decisions as only allowing "the States to grant compensation for the consequences, as defined by the traditional law of torts, of conduct marked by violence and imminent threats to the public order," id., at 247, and noted that in Laburnum

"damages were restricted to the `damages directly and proximately caused by wrongful conduct chargeable to the defendants . . .' as defined by the traditional law of torts. . . . Thus there is nothing in the measure of damages to indicate that state power was exerted to compensate for anything more than the direct consequences of the violent conduct." Id., 248, n. 6. at 249.

In Russell, we specifically observed that the jury had been charged that to award damages it must find a proximate relation between the violence and threats of force and violence complained of, on the one hand, and the loss of wages allegedly suffered, on the other. 356 U. S., at 638, n. 3. In the two Wisconsin Board cases it was noted that the State's administrative-injunctive relief was limited to prohibition against continuation of the unlawful picketing, not all picketing. 315 U. S., at 748; 351 U. S., at 269-270, n. 3. And in Youngdahl, the Court held that a state court injunction which would have prohibited all picketing must be modified to permit peaceful picketing of the premises. We said, "[t]hough the state court was within its discretionary power in enjoining future acts of violence, intimidation and threats of violence by the strikers and the union, yet it is equally clear that such court entered the pre-empted domain [731] of the National Labor Relations Board insofar as it enjoined peaceful picketing . . . ." 355 U. S., at 139.^[17]

It is true that in Milk Wagon Drivers Union v. Meadowmoor Dairies, 312 U. S. 287, the Court approved sweeping state injunctive relief barring any future picketing in a labor dispute, whether peaceful or not. That case, however, was decided only on a constitutional claim of freedom of speech. We did not consider the impact of federal labor policy on state regulatory power. Moreover, as we recognized in Youngdahl, supra, at 139, the case was decided in the context of a strike marked by extreme and repeated acts of violence—"a pattern of violence . . . which would inevitably reappear in the event picketing were later resumed." The Court in Meadowmoor had stated the question presented as "whether a state can choose to authorize its courts to enjoin acts of picketing in themselves peaceful when they are enmeshed with contemporaneously violent conduct which is concededly outlawed," 312 U. S., at 292, and had reasoned that

"acts which in isolation are peaceful may be part of a coercive thrust when entangled with acts of violence. The picketing in this case was set in a background of violence. In such a setting it could justifiably be concluded that the momentum of fear generated by past violence would survive even though future picketing might be wholly peaceful." Id., at 294.

Such special facts, if they appeared in an action for damages after picketing marred by violence had occurred, [732] might support the conclusion that all damages resulting from the picketing were proximately caused by its violent component or by the fear which that violence engendered.^[18] Where the consequences of peaceful and violent conduct are separable, however, it is clear that recovery may be had only for the latter.

In the present case, petitioner concedes that violence which would justify application of state tort law within these narrow bounds occurred during the first two days of the strike. It is a separate issue, however, whether the pleadings, the arguments of counsel to the jury, or the instructions to the jury adequately defined the compass within which damages could be awarded under state law. The tort claimed was, in essence, a "conspiracy" to interfere with Gibbs' contractual relations. The tort of "conspiracy" is poorly defined, and highly susceptible to judicial expansion; its relatively brief history is colored by use as a weapon against the developing labor movement.^[19] Indeed, a reading of the record in this case gives the impression that the notion of "conspiracy" was employed here to expand the application of state law substantially [733] beyond the limits to be observed in showing direct union involvement in violence.

Thus, respondent's complaint alleged "an unlawful conspiracy and an unlawful boycott . . . to maliciously, wantonly and willfully interfere with his contract of employment and with his contract of haulage." No limitation to interference by violence appears. Similarly, counsel in arguing to the jury asserted, not that the conspiracy in which the union had allegedly participated and from which its liability could be inferred was a conspiracy of violence but that it was a conspiracy to impose the UMW and the UMW's standard contract on the coal fields of Tennessee.^[20] Under the state law, it would not have been relevant that the union had not actually authorized, participated in or ratified the particular violence involved or even the general use of violence. It would only be necessary to show a conspiracy in which the union had a part, and to show also that those who engaged in the violence were members of the conspiracy and their acts were related to the conspiracy's purpose.^[21]

The instructions to the jury also appear not to have kept the conspiracy concept within any proper bounds. The charge instructed the jury separately on the § 303 and conspiracy claims, characterizing each as predicated on an assertion that there had been "unlawful" picketing action, and distinguishing one from the other on the basis that in the conspiracy claim "the lawfulness of the means rather than the lawfulness of the object or the purpose [734] of the picketing . . . is controlling." But in charging the conspiracy claim, the court stressed that the "unlawfulness" of the picketing, rather than violence as such, would be controlling. Thus, in characterizing respondent's claim of a conspiracy intentionally to interfere with his contractual relations with Grundy, the trial judge said respondent asserted the interference to be "wrongful in that it was accomplished by unlawful means, including violence and threats of violence." Turning to the question of the international union's responsibility, he said this depended on a showing that it "was a party to a conspiracy pursuant to which the interference was committed." He defined conspiracy as

"an agreement between two or more . . . to do an unlawful thing, or to do a lawful thing by unlawful means. . . . It is not essential to the existence of a conspiracy that the agreement between the conspirators be formally made between the parties at any one time, if, for example, two persons agreed to pursue an unlawful purpose or pursue a lawful purpose by unlawful means, then later a third person with knowledge of the existence of the conspiracy assents to it either impliedly or expressly and participates in it, then all three are conspirators in the same conspiracy. . . . [A]ll that is required is that each party to the conspiracy know of the existence of the conspiracy and that each agrees to assist in some manner in the furtherance of the unlawful purpose . . . or any unlawful means of accomplishing an unlawful purpose."

The trial judge then charged, in accordance with the Tennessee common law on conspiracy,^[22] that the union, if a member of a conspiracy, would be liable for all acts "done in concert . . . with the common purpose, and to effect [735] a common design," whether or not it had authorized, participated in, or ratified the particular acts. The jury was told it might award "only such damages as . . . he has sustained as a proximate and direct result of the action of the defendant," and that "[n]o award of damages can be made . . . on the basis of losses sustained . . . as a result of lawful activity upon the part of the defendant or its agents." Such instructions do not focus the jury's attention upon violence or threats of violence as the essential predicate of any recovery it might award.

III.

Even assuming the conspiracy concept could be and was kept within limits proper to the application of state tort law under the pre-emption doctrine, reversal is nevertheless required here for failure to meet the special proof requirements imposed by § 6 of the Norris-LaGuardia Act:^[23]

"No officer or member of any association or organization, and no association or organization participating or interested in a labor dispute, shall be held responsible or liable in any court of the United States for the unlawful acts of individual officers, members, or agents, except upon clear proof of actual participation in, or actual authorization of, such acts, or of ratification of such acts after actual knowledge thereof."

Petitioner vigorously contends that § 6 applied to the state claims in this case; that, on this record, it cannot be charged with having participated in or authorized the violence of August 15-16; and that its acts once it learned of the violence fell short of what would be necessary to show either ratification of the violence or any intent to build its picketing campaign upon the fears the violence engendered. We agree.

[736] We held in Brotherhood of Carpenters v. United States, 330 U. S. 395, 403, that

"whether § 6 should be called a rule of evidence or one that changes the substantive law of agency . . . its purpose and effect was to relieve organizations. . . and members of those organizations from liability for damages or imputation of guilt for lawless acts done in labor disputes by some individual officers or members of the organization, without clear proof that the organization or member charged with responsibility for the offense actually participated, gave prior authorization, or ratified such acts after actual knowledge of their perpetration."

Shortly thereafter, Congress passed the Labor Management Relations Act, which expressly provides that for the purposes of that statute, including § 303, the responsibility of a union for the acts of its members and officers is to be measured by reference to ordinary doctrines of agency, rather than the more stringent standards of § 6.^[24] Yet although the legislative history indicates that Congress was well aware of the Carpenters decision,^[25] it did not repeal § 6 outright, but left it applicable to cases not arising under the new Act. This selectivity is not surprising, for on state claims, though not on § 303 claims, punitive damages may be recovered. The driving force behind § 6^[26] and the opposition to § 303, even in its limited form,^[27] was the fear that unions might be destroyed [737] if they could be held liable for damage done by acts beyond their practical control. Plainly, § 6 applies to federal court adjudications of state tort claims arising out of labor disputes, whether or not they are associated with claims under § 303 to which the section does not apply.^[28]

Although the statute does not define "clear proof," its history and rationale suggest that Congress meant at least to signify a meaning like that commonly accorded such similar phrases as "clear, unequivocal, and convincing proof." Under this standard, the plaintiff in a civil case is not required to satisfy the criminal standard of reasonable doubt on the issue of participation, authorization or ratification; neither may he prevail by meeting the ordinary civil burden of persuasion. He is required to persuade by a substantial margin, to come forward with "more than a bare preponderance of the evidence to prevail." Schneiderman v. United States, 320 U. S. 118, 125. In our view, that burden was not met.^[29]

[738] At the outset, it is clear that the requisite showing was not made as to possible union authorization of or participation in the violence of August 15 and 16. Although it is undoubtedly true that the officers and members of Local 5881 were present in force at the mine site on those days, neither the Local nor they are parties to this suit. Mr. Gilbert, the UMW representative, had left the area for a business meeting before the series of events culminating in the violence, and immediately upon his return, the violence subsided. The Sixth Circuit conceded that "[t]he proofs were sketchy as to defendant's responsibility for the [first two days' violence]." This view accurately reflects the state of the record. Petitioner was not even aware of Grundy's plan to open the Gray's Creek mine until after the violence had occurred.

The remaining issue is whether there was clear proof that the union ratified the violence which had occurred. Preliminarily, we note that it would be inconsistent with the fabric of national labor policy to infer ratification from the mere fact that petitioner involved itself in the dispute after the violence had occurred, or from the fact that it carried on some normal union functions, such as provision of strike relief. A union would ordinarily [739] undertake these tasks during the course of a lawful strike. National labor policy requires that national unions be encouraged to exercise a restraining influence on explosive strike situations; and when they seek to do so, they should not for these activities be made to risk liability for such harm as may already have been done. The fact that ripples of the earlier violence may still be felt should not be permitted, and under § 6 is not permitted, to impose such liability. Because the dispute which sparked the violence will often continue, the union will feel a responsibility to take up the dispute as well as to curb its excesses. There can be no rigid requirement that a union affirmatively disavow such unlawful acts as may previously have occurred. Cf. ILGWU v. Labor Board, 237 F. 2d 545. What is required is proof, either that the union approved the violence which occurred, or that it participated actively or by knowing tolerance in further acts which were in themselves actionable under state law or intentionally drew upon the previous violence for their force.

The record here is persuasive that the petitioner did what it could to stop or curtail the violence. There was repeated and uncontradicted testimony that when news of the violence reached the meeting that Gilbert was attending, he was given firm instructions to return to the scene, to assume control of the strike, to suppress violence, to limit the size of the picket line, and to assure that no other area mines were affected.^[30] He [740] succeeded. Although the day after his return two Consolidated officers were harassed by a large and unruly mob in a nearby town, this incident was unrelated to respondent, and was not repeated. There was no further violence at the mine site, and the number of pickets was reduced to a very few. Other mines in the immediate area, including two worked on lease by Gibbs, continued to operate, although strenuous effort was required to accomplish this; one union official testified, "I thought I was going to get whipped two or three times [by members of the Local who opposed this policy]."^[31]

To be sure, there was testimony that Gilbert and, through him, the international union were not pleased with respondent's role in the abortive venture to open the Gray's Creek mines with members of the Southern Labor Union. A company officer testified that when the mines finally opened respondent was not hired, because "Had I hired Mr. Paul Gibbs none of these mines would be open today." Respondent testified that Gilbert had told him, shortly after assuming control of the strike, "I want you to keep your damn hands off of that Gray's Creek area over there, and tell that Southern Labor Union that we don't intend for you to work that mine." To another, Gilbert is alleged to have said, "Hell, we can't let that [741] go on . . . Paul was trying to bring this other union in there, and [Gilbert said] he ain't going to get by with it." A third witness reported remarks of a similar tenor. Respondent testified that fear for his own safety caused him not to visit his mine leases after the events of August 15 and 16. His foreman testified to minor acts of violence at the mine site, never connected to any person or persons.

The relevant question, however, is whether Gilbert or other UMW representatives were clearly shown to have endorsed violence or threats of violence as a means of settling the dispute. The Sixth Circuit's answer was that they had. Its view of the record gave it

"the impression that the threat of violence remained throughout the succeeding days and months. The night and day picketing that followed its spectacular beginning was but a guaranty and warning that like treatment would be accorded further attempts to open the Gray's Creek area. The aura of violence remained to enhance the effectiveness of the picketing. Certainly there is a threat of violence when the man who has just knocked me down my front steps continues to stand guard at my front door." 343 F. 2d, at 616.

An "impression" is too ephemeral a product to be the result of "clear proof." As we have said, the mere fact of continued picketing at the mine site is not properly relied upon to show ratification. But even accepting the passage as a holding that "clear proof" of UMW involvement is present, we do not so read the record.

If there was a remaining threat of violence here, it was a threat which arose from the context of the dispute, and not from the manner in which the international union was shown to have handled it. This dispute began when unemployed miners in the Appalachian hills discovered [742] that jobs they believed had been promised to them were being given to others behind their backs. In considering the vicarious liability of the international union, accommodation must be made for that fact. The record here clearly bears the construction that the international union exerted pressure to assure that respondent would lose his present jobs and obtain no more. But the record fails to rebut petitioner's contention that it had been unwilling to see its ends accomplished through violence, and indeed had sought to control the excesses which had occurred. Since the record establishes only peaceful activities in this regard on the part of petitioner, respondent was limited to his § 303 remedy. Teamsters Union v. Morton, supra. Although our result would undoubtedly be firmer if the petitioner had assured respondent that, having assumed control of the strike, it would prevent further violence, in the circumstances of this case the crucial fact of petitioner's participation in or ratification of the violence that occurred was not proved to the degree of certainty required by § 6.

Reversed

THE CHIEF JUSTICE took no part in the decision of this case.

MR. JUSTICE HARLAN, whom MR. JUSTICE CLARK joins, concurring.

I agree with and join in Part I of the Court's opinion relating to pendent jurisdiction. As to Part II, I refrain from joining the Court's speculations about the uses to which it may put the pre-emption doctrine in similar future cases. The holding in Part III that the Norris-LaGuardia Act requires reversal here seems to me correct, but my interpretation of the statute is different and somewhat narrower than that of the Court.

The statutory requirement for union liability in this case is "clear proof of actual participation in, or actual [743] authorization of . . . [the unlawful acts], or of ratification of such acts after actual knowledge thereof."^[32] The Court construes this provision as fixing a new test of the quantum of proof, somewhere between ordinary civil and criminal standards. I do not think the admittedly vague legislative history imports this reading, and I believe it introduces a revealing inconsistency since the new test could not be applied to criminal cases, concededly governed by the same statutory language, without standing the statute on its head by having it reduce present quantum-of-proof requirements in criminal cases, that is, proof "beyond a reasonable doubt." The best reading I can give the statute, absent more light than has been shed upon it in this case, is one directing it against a particular type of inferential proof of authority or ratification unacceptable to those who framed the law. For me, the gist of the statute is that in the usual instance a union's carrying on of its normal strike functions and its failure to take affirmative action to dispel misconduct are not in themselves proof of authorization or ratification of the wrongdoing.^[33]

[744] In the present case, apart from a few quite ambiguous episodes, there was nothing to bring the violence home to the union except, as the Sixth Circuit stressed (see p. 741, ante), that the union continued through its picketing the threat that the earlier violence would be renewed and did not repudiate the violence or promise to oppose its renewal. Whatever arguments could be made for imposing liability in such a situation, I think it approximates what the statute was designed to forbid. On this basis, I concur in the reversal.

[1] Section 303 of the Labor Management Relations Act, 1947 provides:

"(a) It shall be unlawful, for the purpose of this section only, in an industry or activity affecting commerce, for any labor organization to engage in any activity or conduct defined as an unfair labor practice in section 158 (b) (4) of this title.

"(b) Whoever shall be injured in his business or property by reason [of] any violation of subsection (a) of this section may sue therefor in any district court of the United States subject to the limitations and provisions of section 185 of this title without respect to the amount in controversy, or in any other court having jurisdiction of the parties, and shall recover the damages by him sustained and the cost of the suit." 29 U. S. C. § 187 (1964 ed.).

Section 158 (b) (4) of Title 29 U. S. C. (1964 ed.), § 8 (b) (4) of the National Labor Relations Act, as amended, 73 Stat. 542, provides, in relevant part, that:

"(b) It shall be an unfair labor practice for a labor organization or its agents—

.....

"(4) (i) to engage in, or to induce or encourage any individual employed by any person engaged in commerce or in an industry affecting commerce to engage in, a strike or a refusal in the course of his employment to use, manufacture, process, transport, or otherwise, handle or work on any goods, articles, materials, or commodities or to perform any services; or (ii) to threaten, coerce, or restrain any person engaged in commerce or in an industry affecting commerce, where in either case an object thereof is—

.....

"(B) forcing or requiring any person to cease using, selling, handling, transporting, or otherwise dealing in the products of any other producer, processor, or manufacturer, or to cease doing business with any other person, or forcing or requiring any other employer to recognize or bargain with a labor organization as the representative of his employees unless such labor organization has been certified as the representative of such employees under the provisions of section 159 of this title: Provided, That nothing contained in this clause (B) shall be construed to make unlawful, where not otherwise unlawful, any primary strike or primary picketing . . . ."

[2] These events were also the subject of two proceedings before the National Labor Relations Board. In one, the Board found that Consolidated had unlawfully assisted the Southern Labor Union in violation of § 8 (a) (2) of the National Labor Relations Act, as amended, 49 Stat. 452, 29 U. S. C. § 158 (a) (2) (1964 ed.), Tennessee Consolidated Coal Co., 131 N. L. R. B. 536, enforcement denied sub nom. Labor Board v. Tennessee Consolidated Coal Co., 307 F. 2d 374 (C. A. 6th Cir. 1962). In the other, it found that Local 5881 had engaged in coercive picketing in violation of § 8 (b) (1) (A), 61 Stat. 141, 29 U. S. C. § 158 (b) (1) (A) (1964 ed.), Local 5881, UMWA, 130 N. L. R. B. 1181. The International itself was not charged in this proceeding, and the Board's consideration focused entirely on the events of August 16.

[3] The only testimony suggesting that Gilbert might have been at the mine site on August 15-16 was Gibbs' statement that "Well, everything happened so fast there, I'm thinking that I seen Mr. Gilbert drive up there, but where he went, I don't know." Whether such testimony could ever be sufficient to establish presence we need not decide, since respondent effectively conceded in the Sixth Circuit and here that Gilbert was in Middlesboro when the violence occurred.

[4] Immediately after the Board's order in the proceedings against it, note 2, supra, Consolidated reopened the mine it had closed during the spring of 1960, and hired the men of Local 5881. Later, and while this litigation was awaiting trial, that mine was closed as the result of an accident. At this point, the fall of 1962, the Gray's Creek mine was opened using members of Local 5881.

[5] See Dukes v. Brotherhood of Painters, Local No. 437, 191 Tenn. 495, 235 S. W. 2d 7 (1950); Brumley v. Chattanooga Speedway & Motordrome Co., 138 Tenn. 534, 198 S. W. 775 (1917); Dale v. Temple Co., 186 Tenn. 69, 208 S. W. 2d 344 (1948).

[6] The questions had been submitted to the jury on a special verdict form. The suggested remittitur from $60,000 to $30,000 for damages on the employment contract and from $100,000 to $45,000 punitive damages was accepted by respondent. In view of our disposition, we do not reach petitioner's contentions that the verdict must be set aside in toto for prejudicial summation by respondent's counsel, or because the actual damages awarded substantially exceeded the proof, and the punitive damage award may have rested in part on the award of actual damages for interference with the haulage contract, which was vacated as unproved.

[7] See Clark on Code Pleading 75 et seq. (1928); Clark, The Code Cause of Action, 33 Yale L. J. 817 (1924); McCaskill, Actions and Causes of Actions, 34 Yale L. J. 614 (1925); McCaskill, One Form of Civil Action, But What Procedure, for the Federal Courts, 30 Ill. L. Rev. 415 (1935); Gavit, A "Pragmatic Definition" of the "Cause of Action"? 82 U. Pa. L. Rev. 129 (1933); Clark, The Cause of Action, id., at 354 (1934); Gavit, The Cause of Action—a Reply, id., at 695 (1934).

[8] See also American Fire & Cas. Co. v. Finn, 341 U. S. 6, 12; Musher Foundation, Inc. v. Alba Trading Co., 127 F. 2d 9, 12 (C. A. 2d Cir. 1942) (dissenting opinion of Clark, J.).

[9] Shulman & Jaegerman, Some Jurisdictional Limitations on Federal Procedure, 45 Yale L. J. 393, 397-410 (1936); Wechsler, Federal Jurisdiction and the Revision of the Judicial Code, 13 Law & Contemp. Prob. 216, 232 (1948); Barron & Holtzoff, Federal Practice and Procedure § 23 (1965 Supp.).

[10] See, e. g., Fed. Rules Civ. Proc. 2, 18-20, 42.

[11] E. g., Musher Foundation v. Alba Trading Co., supra; Note, The Evolution and Scope of the Doctrine of Pendent Jurisdiction in the Federal Courts, 62 Col. L. Rev. 1018, 1029-1030 (1962).

[12] The question whether joined state and federal claims constitute one "case" for jurisdictional purposes is to be distinguished from the often equally difficult inquiry whether any "case" at all is presented, Gully v. First National Bank, 299 U. S. 109, although the issue whether a claim for relief qualifies as a case "arising under . . . the Laws of the United States" and the issue whether federal and state claims constitute one "case" for pendent jurisdiction purposes may often appear together, see Dann v. Studebaker-Packard Corp., 288 F. 2d 201, 211-215 (C. A. 6th Cir. 1961); Borak v. J. I. Case Co., 317 F. 2d 838, 847-848 (C. A. 7th Cir. 1963), aff'd on other grounds, 377 U. S. 426.

[13] Cf. Armstrong Co. v. Nu-Enamel Corp., 305 U. S. 315, 325. Note, Problems of Parallel State and Federal Remedies, 71 Harv. L. Rev. 513, 514 (1958). While it is commonplace that the Federal Rules of Civil Procedure do not expand the jurisdiction of federal courts, they do embody "the whole tendency of our decisions. . . to require a plaintiff to try his . . . whole case at one time," Baltimore S. S. Co. v. Phillips, supra, and to that extent emphasize the basis of pendent jurisdiction.

[14] Massachusetts Universalist Convention v. Hildreth & Rogers Co., 183 F. 2d 497 (C. A. 1st Cir. 1950); Moynahan v. Pari-Mutuel Employees Guild, 317 F. 2d 209, 211-212 (C. A. 9th Cir. 1963); op. cit. supra, notes 9 and 11.

[15] Some have seen this consideration as the principal argument against exercise of pendent jurisdiction. Thus, before Erie, it was remarked that "the limitations [on pendent jurisdiction] are in the wise discretion of the courts to be fixed in individual cases by the exercise of that statesmanship which is required of any arbiter of the relations of states to nation in a federal system." Shulman & Jaegerman, supra, note 9, at 408. In his oft-cited concurrence in Strachman v. Palmer, 177 F. 2d 427, 431 (C. A. 1st Cir. 1949), Judge Magruder counseled that "[f]ederal courts should not be overeager to hold on to the determination of issues that might be more appropriately left to settlement in state court litigation," at 433. See also Wechsler, supra, note 9, at 232-233; Note, 74 Harv. L. Rev. 1660, 1661 (1961); Note, supra, note 11, at 1043-1044.

[16] Note, supra, note 11, at 1025-1026; Wham-O-Mfg. Co. v. Paradise Mfg. Co., 327 F. 2d 748, 752-754 (C. A. 9th Cir. 1964).

[17] In Teamsters Union v. Morton, supra, a similar analysis was applied to permit recovery under § 303 of damages suffered during a strike characterized by proscribed secondary activity only to the extent that the damages claimed were the proximate result of such activity; damages for associated primary strike activity could not be recovered.

[18] It would of course be relevant if the Board had already intervened and as here, note 2, supra, issued an order which permitted the continuance of peaceful picketing activity.

[19] On the flexibility of "conspiracy" as a tort, see Original Ballet Russe, Ltd. v. Ballet Theatre, Inc., 133 F. 2d 187, 189 (C. A. 2d Cir. 1943); Riley v. Dun & Bradstreet, Inc., 195 F. 2d 812 (C. A. 6th Cir. 1952); Charlesworth, Conspiracy as a Ground of Liability in Tort, 36 L. Q. Rev. 38 (1920); Burdick, Conspiracy as a Crime, and as a Tort, 7 Col. L. Rev. 229 (1907); Burdick, The Tort of Conspiracy, 8 Col. L. Rev. 117 (1908). The anti-labor uses of the doctrine are well illustrated in Sayre, Labor and the Courts, 39 Yale L. J. 682, 684-687 (1930). Similar dangers are presented by the tort of malicious interference with contract, id., at 691-695, a doctrine equally young which in its origins required a showing of interference by force, threats, or fraud, but does so no more, Sayre, Inducing Breach of Contract, 36 Harv. L. Rev. 663 (1923); Comment, 56 Nw. U. L. Rev. 391 (1961).

[20] Respondent's attorney argued in summation:

". . . and here is the conspiracy. Mr. Pass [an official of petitioner's] testified, we want that contract all over this nation. That contract or better. I don't guess at that, there is his testimony. There is no deviation from that contract, Mr. Turnblazer so says, unless it is approved in Washington. They impose a nationwide contract all over this nation, all over. I don't care whether it is in Canada or West Virginia or California or Tennessee."

[21] Note 5, supra.

[22] Ibid.

[23] 47 Stat. 71, 29 U. S. C. § 106 (1964 ed.).

[24] National Labor Relations Act, as amended, § 2 (13), 61 Stat. 139, 29 U. S. C. § 152 (13) (1964 ed.); Labor Management Relations Act, 1947, §§ 301 (e), 303 (b), 61 Stat. 157, 159, 29 U. S. C. §§ 185 (e), 187 (b) (1964 ed.).

[25] See, e. g., S. Rep. No. 105, 80th Cong., 1st Sess., p. 21.

[26] The fullest statement of the basis for § 6 appears in S. Rep. No. 163, 72d Cong., 1st Sess., pp. 19-21.

[27] The present § 303 was introduced on the floor of the Senate by Senator Taft, in response to a more severe proposal which would have permitted injunctive relief as well as damages against secondary activity. 93 Cong. Rec. 4769-4770, 4833-4847, 4858-4875 (1947). The tenor of the opposition may be seen in those pages, and also at 93 Cong. Rec. 4765-4766 (remarks of Senator Thomas); 93 Cong. Rec. 6451-6452 (remarks of Senator Morse); 93 Cong. Rec. 6520-6521 (remarks of Senator Pepper).

[28] The argument might be made that if there were "clear proof" that the local union was responsible, the responsibility of the international union vis-a-vis its local would be governed by a less demanding standard than that applicable for determining the responsibility of a labor organization or its officers on the basis of the acts of "individual officers, members, or agents" of the organization. Since the local was not a party here, we have no occasion to assess this issue. Liability of the international union is premised on the acts of Gilbert and the UMW's other agents, or not at all.

[29] In charging the jury, the trial judge first instructed the jury at length that the plaintiff's burden was to prove his case by a preponderance of the evidence, and that "if the plaintiff carries the burden of proof by a preponderance of the evidence, however slight that preponderance might be, he has done all that is required of him and is entitled to a verdict." In connection with substantive discussion of the state claim, he then remarked:

"Before the defendant may be held responsible for the acts of its agents in entering into a conspiracy during the course of a labor dispute, there must be clear proof that the particular conspiracy charged or the act generally of that nature had been expressly authorized or necessarily followed from a granted authority by the defendant, or that such conspiracy was subsequently ratified by the defendant after actual knowledge thereof."

The phrase "clear proof," referred to just this once, was never explained. The possibility is strong that the jury either did not understand the phrase or completely overlooked it in the context of the lengthy charge given. No challenge is directly made to the charge, however, and it does not appear whether an objection was entered. Accordingly, we do not rest judgment on this point.

[30] Other international union personnel were also later sent, perhaps in part because the union wanted to put its best foot forward in the NLRB proceedings, note 2, supra,which ensued. One such person testified,

". . . I explained to them that the labor board was there investigating and that certainly any mass picketing would only cause them a great deal of trouble, and instructed them that they should limit the number of their pickets and under no circumstances have any violence or any threats of violence to any person coming into or near that area."

[31] About six days after the violence, an earthmoving equipment salesman driving by the entrance to the mine site stopped to ask how he might get to another mine. Gilbert was present among the picketers, and gave him instructions. Gilbert told the salesman that he "couldn't get through" the road chosen, and should approach by another route; he said the salesman should tell any union men he met that he had spoken to Gilbert. A sinister cast can be put on this incident, but it shows clearly only that Gilbert was in control of the strike and that operations unrelated to Gray's Creek were not being interfered with. It is significant that the salesman did not claim to have been stopped by force or threatened in any way; it appears he did no more than seek directions, and received no more in return.

[32] Norris-LaGuardia Act, § 6, 47 Stat. 71, 29 U. S. C. § 106 (1964 ed.). The section is quoted in full at p. 735, ante.

[33] The principal legislative document, S. Rep. No. 163, 72 Cong., 1st Sess., pp. 19-21, is not very illuminating but it does at the end of its discussion of the section make reference to Frankfurter & Greene, The Labor Injunction 74-75 (1930). At these pages, to illustrate rulings on union responsibility that are deemed improper, that book states: " `Authorization' has been found as a fact where the unlawful acts `have been on such a large scale, and in point of time and place so connected with the admitted conduct of the strike, that it is impossible on the record here to view them in any other light than as done in furtherance of a common purpose and as part of a common plan'; where the union has failed to discipline the wrong-doer; where the union has granted strike benefits." (Footnotes omitted.) See also id., at 220-221, n. 42; United Brotherhood of Carpenters v. United States, 330 U. S. 395, 418-419 and n. 2 (Frankfurter, J., dissenting).

[206] [207] Daniel H. Schlueter, Atlanta, Ga. (Thomas M. Byrne, Valerie S. Sanders, Sutherland Asbill & Brennan LLP, Atlanta, Ga.; John H. Beisner, Neil K. Gilman, Rachel A. Shapiro, O'Melveny & Myers LLP, Washington, D.C., on the brief), for Defendant-Appellant.

Darnley D. Stewart, New York, N.Y. (Daniel L. Berger, Bernstein Litowitz Berger & Grossman LLP, New York, N.Y.; Michael E. Terry, Terry & Gore, Nashville, Tenn.; Gary Klein, Grant & Roddy, Boston, Mass.; Kevin Greco, Sandak Friedman Hennessey & Greco, LLP, Stamford, Conn.; Wyman O. Gilmore, Grove Hill, Ala.; Clint W. Watkins, Brentwood, Tenn.; Stuart Rossman, National Consumer Law Center, Boston, Mass., on the brief), for Plaintiffs-Appellees.

Before: NEWMAN, SOTOMAYOR, and WESLEY, Circuit Judges.

JON O. NEWMAN, Circuit Judge.

This appeal concerns the availability of subject matter jurisdiction for permissive counterclaims. It also demonstrates the normal utility of early decision of a motion for class certification. Defendant-Appellant Ford Motor Credit Company ("Ford Credit") appeals from the June 14, 2002, judgment of the United States District Court for the Southern District of New York (Lawrence M. McKenna, District Judge) dismissing for lack of jurisdiction its permissive counterclaims against three of the four Plaintiffs-Appellees and its conditional counterclaims against members of the putative class that the Plaintiffs-Appellees seek to certify. Jones v. Ford Motor Credit Co., No. 00-CV-8330, 2002 WL 1334812 (S.D.N.Y. June 17, 2002). We conclude that supplemental jurisdiction authorized by 28 U.S.C. § 1367 may be available for the permissive counterclaims, but that the District Court's discretion under subsection 1367(c) should not be exercised in this case until a ruling on the Plaintiffs' motion for class certification. We therefore vacate and remand.

Background

Plaintiffs-Appellees Joyce Jones, Martha L. Edwards, Lou Cooper, and Vincent E. Jackson ("Plaintiffs"), individually and as class representatives, sued Ford Credit alleging racial discrimination under the Equal Credit Opportunity Act ("ECOA"), 15 U.S.C. § 1691 et seq. (2003). They had purchased Ford vehicles under Ford Credit's financing plan. They alleged that the financing plan discriminated against African-Americans. Although the financing rate was primarily based on objective criteria, Ford Credit permitted its dealers to mark up the rate, using subjective criteria to assess non-risk charges. The Plaintiffs alleged that the mark-up policy penalized African-American customers with higher rates than those imposed on similarly situated Caucasian customers.

In its Answer, Ford Credit denied the charges of racial discrimination and also asserted state-law counterclaims against Jones, Edwards, and Cooper for the amounts of their unpaid car loans. Ford Credit alleged that Jones was in default on [208] her obligations under her contract for the purchase of a 1995 Ford Windstar, and that Edwards and Cooper were in default on payments for their joint purchase of a 1995 Mercury Cougar. Additionally, in the event that a class was certified, Ford Credit asserted conditional counterclaims against any member of that class who was in default on a car loan from Ford Credit. The Plaintiffs moved to dismiss Ford Credit's counterclaims for lack of subject matter jurisdiction, Fed.R.Civ.P. 12(b)(1), lack of personal jurisdiction, Fed.R.Civ.P. 12(b)(2), improper venue, Fed.R.Civ.P. 12(b)(3), and failure to state a claim upon which relief could be granted, Fed.R.Civ.P. 12(b)(6).

The District Court granted the Plaintiffs' motion and dismissed Ford Credit's counterclaims, summarizing its reasons for doing so as follows: "[D]efendant's counterclaims do not meet the standard for compulsory counterclaims[, and] ... pursuant to § 1367(c)(4), ... there are compelling reasons to decline to exercise jurisdiction over the counterclaims." Jones, 2002 WL 1334812, at *3.

In reaching these conclusions, Judge McKenna acknowledged some uncertainty. After determining that the counterclaims were permissive, he expressed doubt as to the jurisdictional consequence of that determination. On the one hand, he believed, as the Plaintiffs maintain, that permissive counterclaims must be dismissed if they lack an independent basis of federal jurisdiction. On the other hand, he acknowledged, citing Solow v. Jenkins, No. 98-CV-8726, 2000 WL 489667, at *2 (S.D.N.Y. Apr.25, 2000), that "there [was] some authority to suggest that ... the court should determine, based on the particular circumstances of the case, whether it ha[d] authority to exercise supplemental jurisdiction under § 1367(a)" over a counterclaim, regardless of whether it was compulsory or permissive. Jones, 2002 WL 1334812, at *2.

To resolve his uncertainty, Judge McKenna initially ruled that the counterclaims, being permissive, "must be dismissed for lack of an independent basis of federal jurisdiction." Id. He then ruled that, if he was wrong and if supplemental jurisdiction under section 1367 was available, he would still dismiss the counterclaims in the exercise of the discretion subsection 1367(c) gives district courts. Id. Without explicitly stating on which of the four subdivisions of subsection 1367(c) he relied, Judge McKenna gave the following reasons for declining to exercise supplemental jurisdiction:

[1] The claims and counterclaims arise out of the same occurrence only in the loosest terms.... There does not exist a logical relationship between the essential facts [to be proven] in the claim and those of the counterclaims.

[2] [A]llowing defendant's counterclaims to proceed in this forum might undermine the ECOA enforcement scheme by discouraging plaintiffs from bringing ECOA claims due to the fear of counterclaims.

[3] [T]he interests of judicial economy will not be served by joining the claim and counterclaims in one suit [because of] what would most likely be a tremendous number of separate collection actions, each based on facts specific to the individual plaintiffs involved.

Id. at *2-*3. Judge McKenna stated his belief that it would be "unfair and inexpedient" to require absent class members who resided outside of New York to litigate their debt collection actions in the Southern District of New York and that there was no good reason to litigate the debt collection actions in a federal court. Id. at *3.

[209] On March 27, 2003, the District Court entered judgment pursuant to Fed.R.Civ.P. 54(b) in favor of the Plaintiffs, dismissing Ford Credit's counterclaims without prejudice. Ford Credit appeals from this decision.

Discussion

I. Are Ford Credit's Counterclaims Permissive?

Fed.R.Civ.P. 13(a) defines a compulsory counterclaim as

any claim which at the time of serving the pleading the pleader has against any opposing party, if it arises out of the transaction or occurrence that is the subject matter of the opposing party's claim and does not require for its adjudication the presence of third parties of whom the court cannot obtain jurisdiction.

Such counterclaims are compulsory in the sense that if they are not raised, they are forfeited. See Critical-Vac Filtration Corp. v. Minuteman International, Inc., 233 F.3d 697, 699 (2d Cir.2000). Fed.R.Civ.P. 13(b) defines a permissive counterclaim as "any claim against an opposing party not arising out of the transaction or occurrence that is the subject matter of the opposing party's claim."

Whether a counterclaim is compulsory or permissive turns on whether the counterclaim "arises out of the transaction or occurrence that is the subject matter of the opposing party's claim," and this Circuit has long considered this standard met when there is a "logical relationship" between the counterclaim and the main claim. See United States v. Aquavella, 615 F.2d 12, 22 (2d Cir.1979).^[1] Although the "logical relationship" test does not require "an absolute identity of factual backgrounds," id. (internal citation omitted), the "`essential facts of the claims [must be] so logically connected that considerations of judicial economy and fairness dictate that all the issues be resolved in one lawsuit.'" Critical-Vac, 233 F.3d at 699 (emphasis omitted) (quoting Adam v. Jacobs, 950 F.2d 89, 92 (2d Cir.1991)); see also Harris v. Steinem, 571 F.2d 119, 123 (2d Cir.1978); United Artists Corp. v. Masterpiece Productions, Inc., 221 F.2d 213, 216 (2d Cir.1955).

We agree with the District Court that the debt collection counterclaims were permissive rather than compulsory. The Plaintiffs' ECOA claim centers on Ford Credit's mark-up policy, based on subjective factors, which allegedly resulted in higher finance charges on their purchase contracts than on those of similarly situated White customers. Ford Credit's debt collection counterclaims are related to those purchase contracts, but not to any particular clause or rate. Rather, the debt collection counterclaims concern the individual Plaintiffs' non-payment after the contract price was set. Thus, the relationship between the counterclaims and the ECOA claim is "logical" only in the sense that the sale, allegedly on discriminatory credit terms, was the "but for" cause of the non-payment. That is not the sort of [210] relationship contemplated by our case law on compulsory counterclaims. The essential facts for proving the counterclaims and the ECOA claim are not so closely related that resolving both sets of issues in one lawsuit would yield judicial efficiency. Indeed, Ford Credit does not even challenge the ruling that its counterclaims are permissive.

II. Is There Jurisdiction over the Permissive Counterclaims?

For several decades federal courts have asserted that permissive counterclaims require an independent basis of jurisdiction, i.e., that the counterclaim must be maintainable in a federal district court on some jurisdictional basis that would have sufficed had it been brought in a separate action. The origin of this proposition, the questioning of it before the statutory authorization of supplemental jurisdiction in section 1367, and the impact of that provision upon the proposition all merit careful consideration.

(A) Origin of the independent basis doctrine. The first suggestion of the requirement of an independent basis for permissive counterclaims is believed to have appeared in Marconi Wireless Telegraph Co. of America v. National Electric Signaling Co., 206 F. 295 (E.D.N.Y.1913), a case involving former Equity Rule 30. See Thomas F. Green, Jr., Federal Jurisdiction over Counterclaims, 48 Nw. U.L.Rev. 271, 283 (1953). That rule distinguished in its two parts between what we would now call compulsory counterclaims "arising out of the transaction which is the subject matter of the suit" and what we would now call permissive counterclaims "which might be the subject of an independent suit in equity." Equity Rule 30, quoted in Moore v. New York Cotton Exchange, 270 U.S. 593, 609, 46 S.Ct. 367, 70 L.Ed. 750 (1926). In Moore, the Supreme Court ruled that the counterclaim in that case bore a sufficient relation to the underlying transaction under the first part of the equity rule to be properly within federal jurisdiction and explicitly declined to "consider the point that, under the second branch [of the rule], federal jurisdiction independent of the original bill must appear." Id.

By 1944, our Court somewhat tentatively observed that "[i]t seems to be accepted that a permissive counterclaim ... is not ancillary and requires independent grounds of jurisdiction." Lesnik v. Public Industrials Corp., 144 F.2d 968, 976 n. 10 (2d Cir.1944) (citations omitted). We stated the proposition more forcefully two years later in Libbey-Owens-Ford Glass Co. v. Sylvania Industrial Corp., 154 F.2d 814, 816 (2d Cir.1946), although our statement, like most of the language in the early cases concerning the proposition, was dictum.

Our first holding that independent jurisdiction is required for a permissive counterclaim occurred in 1968. See O'Connell v. Erie Lackawanna R.R., 391 F.2d 156, 163 (2d Cir.1968). Six years later, the Supreme Court stated, "If a counterclaim is compulsory, the federal court will have ancillary jurisdiction over it even though ordinarily it would be a matter for a state court," Baker v. Gold Seal Liquors, Inc., 417 U.S. 467, 469 n. 1, 94 S.Ct. 2504, 41 L.Ed.2d 243 (1974) (emphasis added), apparently implying that ancillary jurisdiction is not available for a permissive counterclaim.

Notably absent from this evolution of the case law is a reasoned explanation of why independent jurisdiction should be needed for permissive counterclaims. One early decision hinted at a reason by suggesting that the then restrictive rules concerning joinder of claims were applicable whether the claims were sought to be added [211] by a plaintiff or a defendant. See Electric Boat Co. v. Lake Torpedo Boat Co., 215 F. 377, 381 (D.N.J.1914). Likely also influencing the emergence of the doctrine was concern that, because under Rule 13 all counterclaims that are not compulsory are permissive, requiring independent jurisdiction was necessary to prevent federal court adjudication of every conceivable non-compulsory counterclaim that a defendant might happen to have against a plaintiff, some of which might be totally inappropriate for federal jurisdiction. A wife's federal law suit against her husband for a declaration of rights as a joint author, for example, ought not to be vehicle for his counterclaim seeking a divorce. Perhaps lurking beneath the surface of the casual statements about an independent jurisdiction requirement was apprehension that some counterclaims lacking such a basis would extend the lawsuit beyond Article III's limiting scope of "cases and controversies."

(B) Questioning the doctrine prior to section 1367. The first challenge to the independent jurisdiction requirement appeared in Professor Green's article in 1953. See Thomas F. Green, Jr., Federal Jurisdiction over Counterclaims, 48 Nw. U.L.Rev. 271, 283 (1953). He mounted a powerful argument against the doctrine, demonstrating how it emerged from unreasoned dicta into unexplained holdings and why it was an unwarranted deviation from the general principle that "[t]wo court actions should not be encouraged where one will do." Id. at 271 (footnote omitted). He particularly noted the incursion on the doctrine, well established even in 1953 when he wrote, that permitted some set-offs to be interposed against a plaintiff's claim in the absence of independent jurisdiction. Professor Green questioned why a defendant who can present evidence of a set-off that reduces a plaintiff's judgment to zero should not be able to obtain a counterclaim judgment to which his evidence would entitle him in a separate action. Id. at 287-88.

In 1970, Judge Friendly, the acknowledged jurisdictional scholar of our Court, changed his mind about the independent jurisdiction doctrine and "reject[ed] the conventional learning, which [he] followed too readily in O'Connell." United States v. Heyward-Robinson Co., 430 F.2d 1077, 1088 (2d Cir.1970) (Friendly, J., concurring).^[2] Judge Friendly noted Professor Green's persuasiveness: "The reasons why the conventional view is wrong are set out in detail in [Professor Green's article], and nothing would be gained by repetition." Id.

In 1984, the Third Circuit, in a thoughtful opinion by Judge Becker, rejected the view that independent jurisdiction is required for all permissive counterclaims. See Ambromovage v. United Mine Workers, 726 F.2d 972 (3d Cir.1984). Considering whether jurisdiction was available for a set-off, Judge Becker declined to uphold jurisdiction based on the previously recognized set-off exception to the independent jurisdiction requirement,^[3] and instead argued [212] broadly for ancillary jurisdiction, with some exceptions,^[4] over set-offs and permissive counterclaims that satisfy the test of sharing a "common nucleus of operative fact," United Mine Workers v. Gibbs, 383 U.S. 715, 725, 86 S.Ct. 1130, 16 L.Ed.2d 218 (1966), with the plaintiff's underlying claim. Ambromovage, 726 F.2d at 988-89. "We conclude that the determination that a counterclaim is permissive within the meaning of Rule 13 is not dispositive of the constitutional question whether there is federal jurisdiction over the counterclaim." Id. at 990.

(C) The impact of section 1367. The judge-made doctrine of ancillary jurisdiction, which had been invoked to provide a jurisdictional basis for compulsory counterclaims, was given statutory undergirding when Congress added section 1367 to Title 28 in 1990. See Judicial Improvements Act of 1990, Pub.L. No. 101-650, Title III, § 310(c), 104 Stat. 5114 (1990). The newly labeled "supplemental" jurisdiction explicitly extended federal courts' authority to "all other claims" in a civil action "so related to claims in the action within [the district court's] original jurisdiction that they form part of the same case or controversy under Article III of the United States Constitution." 28 U.S.C. § 1367(a) (2000).

The explicit extension to the limit of Article III of a federal court's jurisdiction over "all other claims" sought to be litigated with an underlying claim within federal jurisdiction recast the jurisdictional basis of permissive counterclaims into constitutional terms.^[5] After section 1367, it is no [213] longer sufficient for courts to assert, without any reason other than dicta or even holdings from the era of judge-created ancillary jurisdiction, that permissive counterclaims require independent jurisdiction. Rising to the challenge, after enactment of section 1367, in a case strikingly similar to our pending case, the Seventh Circuit vacated the dismissal of a permissive counterclaim and remanded for exercise of the discretion contemplated by section 1367. Channell v. Citicorp National Services, Inc., 89 F.3d 379 (7th Cir.1996). Channell involved a creditor's counterclaims to collect debts in a class action alleging violations of the Consumer Leasing Act, 15 U.S.C. §§ 1667-1667e (2000). As Judge Easterbrook stated, "Now that Congress has codified the supplemental jurisdiction in § 1367(a), courts should use the language of the statute to define the extent of their powers." Id. at 385. He viewed section 1367's reach to the constitutional limits of Article III as requiring only "[a] loose factual connection between the claims," id. (internal quotation marks omitted), a standard that appears to be broader than the Gibbs test of "a common nucleus of operative facts," appropriate for permitting joinder of a plaintiff's non-federal claim. In Channell, he readily found the requisite "loose connection" to exist between the Consumer Leasing Act claim and the debt collection counterclaim. Id. at 385-86.

We share the view that section 1367 has displaced, rather than codified, whatever validity inhered in the earlier view that a permissive counterclaim requires independent jurisdiction (in the sense of federal question or diversity jurisdiction). The issue in this case therefore becomes whether supplemental jurisdiction is available for Ford Credit's counterclaims.

III. Application of Section 1367's Standards for Supplemental Jurisdiction

Whether or not the Gibbs "common nucleus" standard provides the outer limit of an Article III "case,"^[6] and is therefore a requirement for entertaining a permissive counterclaim that otherwise lacks a jurisdictional basis, the facts of Ford Credit's counterclaims and those of the Plaintiffs' ECOA claims satisfy that standard, even though the relationship is not such as would make the counterclaims compulsory. See Channell, 89 F.3d at 385-86.^[7] The [214] counterclaims and the underlying claim bear a sufficient factual relationship (if one is necessary) to constitute the same "case" within the meaning of Article III and hence of section 1367. Both the ECOA claim and the debt collection claims originate from the Plaintiffs' decisions to purchase Ford cars.

Satisfying the constitutional "case" standard of subsection 1367(a), however, does not end the inquiry a district court is obliged to make with respect to permissive counterclaims. A trial court must consider whether any of the four grounds set out in subsection 1367(c) are present to an extent that would warrant the exercise of discretion to decline assertion of supplemental jurisdiction.^[8] Subsection 1367(c) provides:

The district courts may decline to exercise supplemental jurisdiction over a claim under subsection (a) if —

(1) the claim raises a novel or complex issue of State law,

(2) the claim substantially predominates over the claim or claims over which the district court has original jurisdiction,

(3) the district court has dismissed all claims over which it has original jurisdiction, or

(4) in exceptional circumstances, there are other compelling reasons for declining jurisdiction.

We have indicated that, where at least one of the subsection 1367(c) factors is applicable, a district court should not decline to exercise supplemental jurisdiction unless it also determines that doing so would not promote the values articulated in Gibbs, 383 U.S. at 726, 86 S.Ct. 1130: economy, convenience, fairness, and comity. See Itar-Tass Russian News Agency v. Russian Kurier, Inc., 140 F.3d 442, 445-47 (2d Cir.1998) (rejecting approach of 1st, 3rd, 7th, and D.C. Circuits in favor of approach adhered to by 8th, 9th, and 11th Circuits).

Clearly the exception set forth in subsection 1367(c)(1) does not apply since Ford Credit's counterclaims do not raise a novel or complex issue of state law, but merely a standard contract question. Nor does subsection 1367(c)(3) apply since the District Court has not dismissed all claims over which it has original jurisdiction. That leaves subsections 1367(c)(2), permitting declination of supplemental jurisdiction where "the [counter]claim substantially predominates over the claim or claims over which the district court has original jurisdiction," and 1367(c)(4), permitting declination "in exceptional circumstances, [where] there are other compelling reasons for declining jurisdiction." The District Court apparently based its decision on subsection 1367(c)(4), since it cited only that subsection in its opinion, but some of the concerns it discussed implicate the substantial predomination analysis of subsection 1367(c)(2) as well.

[215] In Channell, Judge Easterbrook canvassed the competing considerations bearing on whether the subsection (c)(2) and (c)(4) factors might permit declination of supplemental jurisdiction over collection counterclaims interposed against a claim under a consumer protection statute. See Channell, 89 F.3d at 386-87. Acknowledging a district court's discretion, the Seventh Circuit ultimately remanded because "[a]rguments under § 1367(c) are addressed to the district court's discretion." Id. at 387. In Channell, however, the class had been certified, class members had been notified, and some had opted out. Id. at 384. In our case, a ruling on the class motion has not yet been made.

Whether Ford Credit's counterclaims "predominate[ ]" over the Plaintiffs' claims and whether there are "exceptional circumstances" for declining jurisdiction cannot properly be determined until a decision has been made on the Plaintiffs' motion for class certification. Both the applicability of subsections 1367(c)(2) and (4), and the exercise of a district court's discretion in the event either or both are ruled applicable will be significantly influenced by the existence of a large class as sought by the Plaintiffs. The District Court's conclusions that it would be "unfair and inexpedient" to require out-of-state class members to litigate Ford's state law debt claims in New York, and that allowing the counterclaims might dissuade potential plaintiffs from joining the class, were therefore premature.^[9]

Class certification is to be decided "at an early practicable time" after the commencement of a suit. Fed.R.Civ.P. 23(c)(1) (amended Dec. 1, 2003). See 5 James Wm. Moore et al., Moore's Federal Practice-Civil § 23.61 (3d ed.2003); cf. Cottone v. Blum, 571 F.Supp. 437, 440-41 (W.D.N.Y.1983) (dismissing class action for, among other things, failure to move for class certification within 60 days of filing of complaint, as required by local rule). That course is especially important in this case. Accordingly, we remand this case with directions to rule on the class certification motion, and then, in light of that ruling, to proceed to determine whether to exercise or decline supplemental jurisdiction.

On remand, the District Court should exercise its discretion pursuant to subsection 1367(c) in light of our decision in Itar-Tass, particularly the caution there expressed concerning use of subsection 1367(c)(4), 140 F.3d at 448. In order to decline jurisdiction on this basis, the District Court should identify truly compelling circumstances that militate against exercising jurisdiction. Id. Moreover, if the Court certifies the class action, its substantial predomination analysis under subsection 1367(c)(2) should take into account the [216] methods by which the class action might be managed in order to prevent the state law counterclaims from predominating. By bifurcating the litigation, certifying a limited class (perhaps only in-state plaintiffs), or utilizing other management tools, the District Court might be able to structure the litigation in such a way as to prevent the state law claims from predominating over the federal basis of the action, while maintaining the advantages inherent in providing a forum in which all of the litigants' claims can be litigated.

Conclusion

The judgment dismissing Ford Credit's counterclaims is vacated, and the case is remanded for further proceedings consistent with this opinion. No costs.

[1] The phrase "logical relationship," in the context of counterclaims, was first used by the Supreme Court in Moore v. New York Cotton Exchange,270 U.S. 593, 610, 46 S.Ct. 367, 70 L.Ed. 750 (1926). Referring to a counterclaim "arising out of the transaction which is the subject matter of the suit," as stated in former Equity Rule 30, the Court explained:

"Transaction" is a word of flexible meaning. It may comprehend a series of many occurrences, depending not so much upon the immediateness of their connection as upon their logical relationship.

Id.

[2] Judge Friendly refrained from calling for in banc consideration after his change of position because, in agreement with his ultimate vote, his two colleagues had affirmed the judgment upholding federal jurisdiction on the ground, which he did not share, that the counterclaim was compulsory. See Heyward-Robinson, 430 F.2d at 1089.

[3] The set-off exception provides that "[w]here the permissive counterclaim is in the nature of a set-off interposed merely to defeat or reduce the opposing party's claim and does not seek affirmative relief, no independent jurisdictional grounds are required." Heyward-Robinson, 430 F.2d at 1081 n. 1. As the court noted in Ambromovage, the "defensive set-off exception does not fit squarely within the analytic framework set forth" in United Mine Workers v. Gibbs, 383 U.S. 715, 86 S.Ct. 1130, 16 L.Ed.2d 218 (1966), because "under the exception, all defensive set-offs, whether or not they satisfy the Gibbs test, are within the ancillary jurisdiction of the court." Ambromovage, 726 F.2d at 990-92 & n. 56.

[4] Anticipating considerations that would later be codified in 28 U.S.C. § 1367, Judge Becker suggested that a district court should decline to exercise ancillary jurisdiction over a defendant's claims where the exercise of jurisdiction would violate some federal policy limiting jurisdiction and be an inappropriate exercise of a district court's discretion, taking into account such factors as "fairness to the litigants, judicial economy, and the interests of federalism." Ambromovage, 726 F.2d at 991.

[5] There is some doubt as to whether section 1367's expansion of supplemental jurisdiction to its constitutional limits renders the provision's scope broader than was contemplated in Gibbs. The text of subsection 1367(a) unambiguously extends jurisdiction to the limits of Article III, and the provision's legislative history indicates that Congress viewed the Gibbs "common nucleus" test as delineating those constitutional limits. See H.R.Rep. No. 101-734, reprinted in 1990 U.S.C.C.A.N. 6360, 6374-75 (stating that "subsection (a) codifies the scope of supplemental jurisdiction first articulated by the Supreme Court in United Mine Workers v. Gibbs, 383 U.S. 715, 86 S.Ct. 1130, 16 L.Ed.2d 218 (1966)"). Several commentators have suggested, however, that the extent of constitutional jurisdiction is broader than the Gibbstest, as Article III may not require a particular factual relationship between joined claims or counterclaims and federal claims that provide the basis for jurisdiction.

In 1983, Professor Matasar, anticipating section 1367's authorization of supplemental jurisdiction for all claims within an Article III "case or controversy," advocated permitting joinder of all claims, whether those of the plaintiff or the defendant, to the full extent of "the system of rules lawfully adopted to govern procedure in the federal courts." See Richard A. Matasar, Rediscovering "One Constitutional Case": Procedural Rules and the Rejection of the Gibbs Test for Supplemental Jurisdiction, 71 Cal. L.Rev. 1399, 1478-79. (1983). He drew support from Chief Justice Marshall's statement that "`[the judicial] power is capable of acting only when the subject is submitted to [the judicial department], by a party who asserts his rights in the form prescribed by law.'" Id. at 1479 (quoting Osborn v. Bank of the United States, 22 U.S. (9 Wheat.) 738, 819, 6 L.Ed. 204 (1824) (Little, Brown 1864)). Professor Matasar would permit jurisdiction over any counterclaim authorized by federal rules, without requiring any factual relationship to the underlying claim.

More recently, Professor Fletcher (now Judge Fletcher) also advocated a broad view of the claims that could be joined in a "case or controversy" under section 1367. See William A. Fletcher, "Common Nucleus of Operative Fact" and Defensive Set-Off Beyond the Gibbs Test, 74 Ind. L.J. 171 (1998). Like Professor Matasar, Professor Fletcher urged that the constitutional test of a "case" did not require a factual connection between joined claims, but would be satisfied under either the joinder standards applicable when the Constitution was adopted or modern joinder rules. Id. at 178.

Congress's understanding of the extent of Article III is of course not binding as constitutional interpretation, and section 1367's legislative history cannot be read as an independent limit on subsection 1367(a)'s clear extension of jurisdiction to the limits of Article III. Thus, the correct reading of subsection 1367(a)'s reference to "the same case or controversy under Article III" remains unsettled.

[6] If the Gibbs standard marks the outer limit of an Article III "case," congressional authorization to join counterclaims with a more tenuous connection to the underlying claim would be unconstitutional unless Congress has some authority to expand the constitutional scope of "case."

[7] We note that the "common nucleus" test of Gibbs, expanding the prior test of Hurn v. Oursler, 289 U.S. 238, 245-47, 53 S.Ct. 586, 77 L.Ed. 1148 (1933), was developed to provide some limit upon the state law claims that a plaintiff could join with its federal law claims. That rationale does not necessarily apply to a defendant's counterclaims. A plaintiff might be tempted to file an insubstantial federal law claim as an excuse to tie to it one or more state law claims that do not belong in a federal court. There is no corresponding risk that a defendant will decline to file in state court an available state law claim, hoping to be lucky enough to be sued by his adversary on a federal claim so that he can assert a state law counterclaim.

[8] Subsection 1367(b), precluding exercise of supplemental jurisdiction under some circumstances where jurisdiction over the underlying claim is based solely on diversity of citizenship, does not apply here; the Plaintiffs' underlying claim is based on federal question jurisdiction.

[9] The District Court's assumption that plaintiffs' class would be certified, and its analysis of the state law counterclaims in light of that assumption, unduly weighted the subsection 1367(c) analysis in favor of the plaintiffs. Gibbs emphasizes that the question of "whether [supplemental] jurisdiction has been properly assumed is one which remains open throughout the litigation," and the analysis should be undertaken when the district court is best positioned to determine how the exercise of jurisdiction will affect the case as a whole. Gibbs, 383 U.S. at 727, 86 S.Ct. 1130. Thus, when pendent state law claims are asserted in the context of a putative class action, district courts should normally not dismiss the claims based solely on the problems that could arise if the class is eventually certified. See, e.g., Clark v. McDonald's Corp., 213 F.R.D. 198, 232 (D.N.J.2003) (noting that "[a]lthough the potential remains for this litigation to reach a crossroads where the better course may be to decline supplemental jurisdiction over claims arising under the laws of the Subclass States, or some of them," there was no reason to dismiss the claims prior to certification).

[107] [108] ARGUED: David Clifton Schroeder, Murphy, McGettigan, Richards & West, P.C., Alexandria, VA, for Appellant. David Rosenblum, Rosenblum & Rosenblum, P.C., Alexandria, VA, for Appellees.

Before WILKINSON and WILKINS, Circuit Judges, and PHILLIPS, Senior Circuit Judge.

Reversed and remanded by published opinion. Judge WILKINSON wrote the opinion, in which Judge WILKINS and Senior Judge PHILLIPS joined.

OPINION

WILKINSON, Circuit Judge:

This appeal presents the question whether the district court properly dismissed plaintiff's case for lack of jurisdiction because of an insufficient amount in controversy under 28 U.S.C. § 1332(a). The trial court felt compelled to dismiss all of plaintiffs claims when one count was eliminated and the remaining liquidated damages totalled less than fifty thousand dollars. Because we think the district court had discretion to retain the remaining counts, we reverse and remand the case for an appropriate exercise of that discretion.

I.

In 1993, Kathleen Shanaghan brought this diversity action in Virginia against John Cahill and his company Cahill & Associates, Inc., seeking to recover on three separate debts. Her complaint alleged that in 1987 and 1988 she made three loans to Cahill and his company, in the amounts of $40,000, $23,696, and $14,700. She further alleged that the defendants had refused payment despite repeated demands, and were in default on all three debts.

The defendants filed an answer, and discovery proceeded through March of 1994. Plaintiff produced two promissory notes memorializing the loans of $23,696 and $14,700. She was unable, however, to provide a writing for the alleged loan of $40,000, though she has always maintained that such a writing exists. The defendants filed for partial [109] summary judgment in March of 1994, arguing that the claim on the $40,000 loan was barred by the Virginia Statute of Frauds, Va.Code § 11-2(4), and the Statute of Limitations, Va.Code § 8.01-246. The district court agreed and granted summary judgment on the $40,000 claim. Then, noting that the amount in controversy had fallen below fifty thousand dollars, the court dismissed the remaining claims for lack of subject matter jurisdiction. Plaintiff has appealed, challenging only the dismissal of her two smaller claims.

II.

A.

Federal district courts possess jurisdiction over cases in diversity when "the matter in controversy exceeds the sum or value of $50,000." 28 U.S.C. § 1332(a) (1988). A plaintiff may aggregate smaller claims in order to reach this threshold, as was done in this case. Griffin v. Red Run Lodge, Inc., 610 F.2d 1198, 1204 (4th Cir. 1979). After Shanaghan's claim for $40,000 was dismissed, however, the district court concluded it was bound by § 1332 to dismiss her remaining aggregated claims of $38,669. The court apparently believed it had no discretion in this matter, but rather was faced with a mandatory obligation to dismiss the case in its entirety, despite the possibility of a statute of limitations bar to refiling in state court, and regardless of the potential merit of the remaining claims.

This assumption was in error. The basis for district court discretion in this context lies in the model of supplemental jurisdiction set forth in 28 U.S.C. § 1367. The doctrine of supplemental jurisdiction indicates that federal courts generally have discretion to retain or dismiss state law claims when the federal basis for an action drops away. See 28 U.S.C. § 1367 (1993). Federal supplemental jurisdiction was created by the Judicial Improvement Act of 1990, Pub.L. No. 101-650, Title III, § 310(a), which codified the doctrine of pendent jurisdiction developed by the Supreme Court in the case of United Mine Workers of America v. Gibbs, 383 U.S. 715, 86 S.Ct. 1130, 16 L.Ed.2d 218 (1966), and its progeny.^[1] Section 1367(a) provides that

in any civil action of which the district courts have original jurisdiction, the district courts shall have supplemental jurisdiction over all other claims that are so related to claims in the action within such original jurisdiction that they form part of the same case or controversy under Article III of the United States Constitution.

28 U.S.C. § 1367(a). Supplemental jurisdiction thus allows parties to append state law claims over which federal courts would otherwise lack jurisdiction, so long as they form part of the same case or controversy as the federal claims. Gibbs, 383 U.S. at 725, 86 S.Ct. at 1138.

Moreover, the statute is not limited to cases where the original basis for federal jurisdiction was a federal question. It clearly provides for the operation of supplemental jurisdiction in diversity cases. First, § 1367(a) is broadly phrased to provide for supplemental jurisdiction over claims appended to "any civil action" over which the court has "original jurisdiction." 28 U.S.C. § 1367(a). Second, § 1367(b) imposes specific limits on the use of supplemental jurisdiction in diversity cases in order to prevent the addition of parties that would destroy complete diversity as required by § 1332, but otherwise plainly contemplates the use of supplemental jurisdiction in that context. 28 U.S.C. § 1367(b). See also 28 U.S.C.A. § 1367, David Seigel, Practice Commentary 832 (1993) (noting that "[b]y no means does [§ 1367(b)] exclude [supplemental jurisdiction] from diversity cases in general."). The only possible interpretation of this language is that state law claims between diverse parties that do not, however, satisfy the jurisdictional amount requirements appended to diversity actions are cognizable under supplemental jurisdiction.

[110] The statute then goes on to provide that courts "may decline" to exercise supplemental jurisdiction in certain circumstances. 28 U.S.C. § 1367(c). In particular, a court has discretion to dismiss or keep a case when it "has dismissed all claims over which it has original jurisdiction." 28 U.S.C. § 1367(c)(3). Recent case law has emphasized that trial courts enjoy wide latitude in determining whether or not to retain jurisdiction over state claims when all federal claims have been extinguished. See, e.g., Noble v. White, 996 F.2d 797, 799 (5th Cir.1993). Among the factors that inform this discretionary determination are convenience and fairness to the parties, the existence of any underlying issues of federal policy, comity, and considerations of judicial economy. Carnegie-Mellon University v. Cohill, 484 U.S. 343, 350 n. 7, 108 S.Ct. 614, 619 n.7, 98 L.Ed.2d 720 (1988); Growth Horizons, Inc. v. Delaware County, 983 F.2d 1277, 1284 (3d Cir.1993). The doctrine of supplemental jurisdiction "thus is a doctrine of flexibility, designed to allow courts to deal with cases involving pendent claims in the manner that most sensibly accommodates a range of concerns and values." Cohill, 484 U.S. at 350, 108 S.Ct. at 619.

B.

There are several reasons why the supplemental jurisdiction model of discretion should apply when the amount in controversy falls below the fifty thousand dollar threshold, just as it does when a federal question or a diverse claim falls out of a case. First, the same basic pattern of circumstances exists in both contexts: the jurisdictional basis of the action fades away and the court is left with what would otherwise be a state law case. There is no way to distinguish a reduction of the amount in controversy from the disappearance of a federal claim as contemplated under § 1367(c)(3). Indeed, the factors applicable in the typical pendent jurisdiction case are equally applicable here — comity, the existence of a state limitations bar, and considerations of judicial economy. Whenever the basis for federal jurisdiction evaporates, Congress has provided for discretion. There are no situations wherein a federal court must retain jurisdiction over a state law claim, which would not by itself support jurisdiction. It makes little sense, then, to think of jurisdictional amounts as a separate category of cases when they so clearly fit into the congressional scheme of supplemental jurisdiction. It does make sense, however, to have one rule of district court discretion, not separate rules for separate jurisdictional bases when simple logic dictates that they be treated the same.

This point is illustrated by the fact that the grant of discretion to retain or dismiss residual state law claims in § 1367(c)(3) would apply in cases closely analogous to the one before us. For example, § 1367(c)(3) provides on its own terms for supplemental jurisdiction over an inadequate amount in controversy if the claim that is dismissed is in excess of $50,000. See 28 U.S.C. § 1367(c) ("courts may ... exercise supplemental jurisdiction ... if ... the district court has dismissed all claims over which it has original jurisdiction"). In such a case the remaining smaller claims would have been "pendent" to the larger dismissed claim, and thus the court could choose to exercise jurisdiction over them. See Worthams v. Atlanta Life Ins. Co., 533 F.2d 994, 997-98 (6th Cir.1976) (finding jurisdiction still existed over claim of $7,900 when other $100,000 claim was dismissed). Here, the only difference is that no single claim by itself carried plaintiff into federal court, and so she was forced to aggregate her various smaller claims to reach the jurisdictional threshold. Similarly, supplemental (ancillary) jurisdiction provides discretionary review of counterclaims of less than $50,000 even when the primary federal claim is dismissed. 28 U.S.C.A. § 1367, David Seigel, Practice Commentary 830 (1993). Hence, a reduction of the amount in controversy below the $50,000 limit simply does not affect the existence of discretion to either retain or dismiss the remaining claims.

Second, when § 1332 (diversity jurisdiction) and § 1367 (supplemental jurisdiction) are read together, the natural conclusion is that the discretionary model may be applied in the context of jurisdictional amounts in liquidated damages cases. Section 1332(a) provides positive, mandatory jurisdiction in diversity cases when the amount is over [111] $50,000. 28 U.S.C. § 1332(a). Section 1367(a) is a parallel to § 1332(a), providing a similar positive grant of supplementary jurisdiction. Section 1367(c), by contrast, is permissive, allowing a court to decline that mandatory jurisdiction when read in conjunction with § 1367(a). Thus, if § 1332(a) is read with § 1367(a), it fits together with the permissive language of § 1367(c), allowing courts to "decline" jurisdiction when the amount falls below $50,000.

Moreover, § 1367(b), which limits the use of supplemental jurisdiction in diversity cases, speaks only to preventing attempts by plaintiffs to circumvent complete diversity. It bars jurisdiction over certain claims against or by parties added through Fed. R.Civ.P. 14, 19, 20, or 24. 28 U.S.C. § 1367(b). Nothing in § 1367(b) precludes the use of a supplemental jurisdiction analysis in diversity cases when the issue is the amount in controversy rather than complete diversity. If Congress had wanted to limit the use of supplemental jurisdiction to prevent discretionary decisions in this context, § 1367(b) would have been the obvious place in which to do it.

Case law on jurisdictional amounts that predates § 1367 is in some tension with the statute. The case of St. Paul Mercury Indemnity Co. v. Red Cab Co., 303 U.S. 283, 58 S.Ct. 586, 82 L.Ed. 845 (1938), required retention of jurisdiction over cases falling below the jurisdictional amount threshold so long as the amount alleged in the complaint was made in good faith. St. Paul, 303 U.S. at 288, 58 S.Ct. at 590. See also Lynch v. Porter, 446 F.2d 225, 228 (8th Cir.1971), cert. denied, 404 U.S. 1047, 92 S.Ct. 711, 30 L.Ed.2d 739 (1972). St. Paul further held that a claim of good faith is negated only if it was a legal certainty that, at the time of the complaint, the plaintiff could not recover the requisite amount. St. Paul, 303 U.S. at 289, 58 S.Ct. at 590. However, the St. Paul test was developed long before the modern concepts of supplemental jurisdiction expressed in Gibbs. Under the Gibbs approach, federal courts have discretion to dismiss or retain residual state law claims. Congress, by adopting § 1367, expressed an emphatic preference for the Gibbs approach. A straightforward application of the St. Paul rule would nevertheless mandate retention of cases falling below the jurisdictional amount despite, for example, the existence of novel or complex issues of state law. Divesting federal courts of discretion in such circumstances simply cannot be squared with the statutory language. See 28 U.S.C. § 1367(c)(1). Thus, the strict St. Paul rule is inconsistent with the statutory framework of § 1367 and so must be modified to fit the contemporary congressional view of federal jurisdiction.

Indeed, any strict rule that mandates either retention or dismissal of residual claims in all cases operates in derogation of important competing interests. On the one hand, a standard tied exclusively to the St. Paul rule of good faith pleading fails to respect the congressional purpose in raising the jurisdictional amount to the fifty thousand dollar threshold. Congress recently raised the amount in an effort to prevent state law claims for modest sums from landing in federal court. See 136 Cong. Rec. S16295 (Section-by-Section Analysis of Judicial Improvements and Access to Justice Act) (noting increase intended to reduce federal caseload). See also Snyder v. Harris, 394 U.S. 332, 339-40, 89 S.Ct. 1053, 1058-59, 22 L.Ed.2d 319 (1969) (purpose of steady increases in jurisdictional amount is to stem rising caseload of federal courts). A strict rule, however, that required retention of all claims that pass the lenient threshold of "good faith" pleading violates this basic congressional purpose. It would require, for example, federal judges to resolve cases that Congress believed all along should be pressed in state court, because they involved both limited sums and purely state law claims. Thus, a rigid rule of retention would sacrifice the very purpose of a jurisdictional amount requirement, not to mention Congress' recent action in raising that amount five-fold. See U.S.C. § 1332(a).

On the other hand, a rigid rule requiring dismissal once the jurisdictional amount falls below the statutory standard would work a serious injustice. It might result in valid claims going unheard or in significant wastes of judicial resources. These were the same [112] concerns that motivated the creation of pendent jurisdiction in the first place. See Gibbs, 383 U.S. at 726, 86 S.Ct. at 1139. Moreover, a rule of automatic dismissal would encourage defendants in attempts to bootstrap a win on a single liquidated claim into an overall victory. If district courts lacked discretion, they would be unable to combat such manipulations, and as a result meritorious claims might fall by the wayside. Thus, the potential injustice to litigants of a Rule of Dismissal is as undesirable as the flood of trivial claims that might result from a Rule of Retention. Justice is better served by a jurisdictional rule that includes some measure of discretion for the district court.

C.

The application of § 1367's discretionary analysis in the context of a jurisdictional amount case is quite straightforward. First, the court should look to the face of the complaint itself to determine whether it is a legal certainty that plaintiff's claims do not reach the required amount. See St. Paul, 303 U.S. at 288, 58 S.Ct. at 590. Unless the claim for an amount over the jurisdictional prerequisite is made in bad faith, or unless it is plain from the complaint that an amount less than the jurisdictional amount is all that is at issue, the district court has jurisdiction over the case. This is akin to a "well-pleaded complaint" rule. Of course, if plaintiff has alleged only a small amount of damages or it is otherwise obvious that the jurisdictional amount under § 1332(a) cannot be satisfied, the court must dismiss the case outright for lack of jurisdiction. Wiggins v. North American Equitable Life Assurance Co., 644 F.2d 1014, 1016-18 (4th Cir.1981). In this case, by contrast, it appears from the face of the complaint that Shanaghan had alleged claims totaling well over the $50,000 limit.

Second, if some event subsequent to the complaint reduces the amount in controversy, such as the dismissal of one count based on the defendant's answer, the court must then decide in its discretion whether to retain jurisdiction over the remainder of the case. Here, courts should be guided by the same kind of factors that inform decisions in the supplemental jurisdiction context when the federal basis of a case disappears. In general, courts should weigh convenience and fairness to both parties, as well as the interests of judicial economy. See In Re Conklin, 946 F.2d 306, 322 (4th Cir.1991) (citing Gibbs, 383 U.S. at 725, 86 S.Ct. at 1138). In particular, it is important to consider whether the amount claimed in the complaint was made in good faith, or whether plaintiff was consciously relying on flimsy grounds to get into federal court. See Rosado v. Wyman, 397 U.S. 397, 404-05, 90 S.Ct. 1207, 1213-14, 25 L.Ed.2d 442 (1970) (distinguishing initial "insubstantiality" from subsequent "mootness"). Another critical equitable factor is the existence of any state limitations bars to refiling in state court.^[2]See Ridenour v. Andrews Federal Credit Union, 897 F.2d 715, 722 (4th Cir.1990). Where, as here, a plaintiff might suffer serious prejudice from the dismissal of her action, courts should be more willing to retain the case despite the smaller amount in controversy. Next, courts should account for the amount of time and energy that has already been expended, and decide whether it might be more efficient to simply retain jurisdiction. See Washington v. Union Carbide Corp., 870 F.2d 957, 962 (4th Cir.1989). Finally, other considerations may arise on a case-by-case basis, such as the existence of some significant issue of state law best resolved in state court, that could inform the trial court's determination whether to keep or dismiss the case entirely. See McCullough v. Branch Banking & Trust Co., Inc., 844 F.Supp. 258, 260-62 (E.D.N.C.1993) (describing discretionary analysis under § 1367(c)(3)), aff'd, 35 F.3d 127 (4th Cir. 1994).

In sum, we leave it to the sound judgment of the district court to decide whether to exercise jurisdiction over residual liquidated [113] claims under $50,000, so long as it was not a legal certainty from the outset that the plaintiff had no business being in federal court.

III.

For the foregoing reasons, the judgment of the district court is reversed, and this matter is remanded with instructions to determine whether the court, in its discretion, should maintain jurisdiction over plaintiff's liquidated damages claims of $23,969 and $14,700.

REVERSED AND REMANDED.

[1] Of course, supplemental jurisdiction also incorporates the doctrine of ancillary jurisdiction. See 28 U.S.C.A. § 1367.

[2] It is unclear whether the savings provision in § 1367(d), which tolls state statutes of limitations for state claims brought under § 1367(a), strictly applies to smaller aggregated claims which reach federal court through § 1332. District courts therefore should make an effort to determine whether the presence of a state savings or tolling statute would permit plaintiffs to proceed in state court free from any limitations bar.

[238] [239] Charles S. Hey-Maestre, with whom Adalina deJesus-Morales was on brief for Manuel A. Iglesias.

Roberto O. Maldonado-Nieves, was on brief for Mutual Life Insurance Company of New York.

Before TORRUELLA, Chief Judge, CAMPBELL, Senior Circuit Judge, and STAHL, Circuit Judge.

LEVIN H. CAMPBELL, Senior Circuit Judge.

Manuel A. Iglesias appeals from the district court's grant of summary judgment on his discrimination and contract claims against his former employer, Mutual Life Insurance Company of New York ("MONY"). MONY appeals from the court's dismissal of its counterclaim for restitution of money that Iglesias obtained by submitting admittedly overstated expense reports. We affirm the judgment for MONY against Iglesias. We vacate the order dismissing MONY's counterclaim and remand the counterclaim with directions to dismiss the counterclaim without prejudice for want of jurisdiction.

I. Iglesias's Claims.

We review a grant of summary judgment de novo, taking the facts in the light most favorable to the non-moving party. See One Nat'l Bank v. Antonellis, 80 F.3d 606, 608 (1st Cir.1996). After considering the record, briefs, and oral arguments, we affirm the dismissal of Iglesias's discrimination and contract claims substantially for the reasons stated in the district court's memoranda and orders. We add only the following short discussion.

Regarding Iglesias's discrimination claims,^[1] Iglesias received clear and unequivocal notice that he was terminated as MONY's San Juan Agency Manager on February 9, 1989, at the latest. On that date, MONY sent him a letter informing him that he would no longer be authorized to act as Agency Manager. In the context of the parties' prior discussions regarding Iglesias's retirement, the February 1989 letter put Iglesias on notice that he had been fired.^[2] Iglesias did not file administrative charges with the Anti-Discrimination Division of the Puerto Rico Department of Labor until August 28, 1990, and his initial complaint in federal court was not filed until April 17, 1991. Iglesias's [240] action for discrimination was, therefore, barred by the applicable statutes of limitations. See 42 U.S.C. § 2000e-5(e) (establishing a 300-day limitation period for actions under Title VII that were first presented to an administrative agency); 29 U.S.C. § 626(d)(2) (setting a 300-day limitation period for actions under the ADEA that were first brought before a state authority); Olmo v. Young & Rubicam, 110 D.P.R. 740, 745, 1981 WL 176523 (P.R.1981) (applying a one year limitation period to claims brought under Law 100). The Puerto Rico Supreme Court's decision in Vélez Rodríguez v. Pueblo Int'l, Inc., 94 JTS 37 (P.R. March 18, 1994) does not apply to Iglesias's claims for the reasons stated in the district court's memorandum.

Regarding Iglesias's contract claims, MONY's employment contract with Iglesias did not limit MONY's ability to withdraw products from the Puerto Rico market. Iglesias argues that the provision in the contract authorizing him to "solicit applications for insurance in MONY of the types of insurance which MONY is issuing" created a vested right that prevented MONY from ever altering its policy offerings in Puerto Rico. We recognize that "when the facts support plausible but conflicting inferences on a pivotal issue in the case, the judge may not choose between those inferences at the summary judgment stage." Coyne v. Taber Partners I, 53 F.3d 454, 460 (1st Cir.1995). We need not indulge a nonmoving party's inferences, however, if they do not "flow rationally from the underlying facts." Rubinovitz v. Rogato, 60 F.3d 906, 911 (1st Cir. 1995).

No reasonable jury could agree with Iglesias's reading of his contract. The plain language of the provision purports only to authorize Iglesias to sell MONY's product line, it does not guarantee stability of that line. MONY is a national company that deals with many agents in diverse geographic areas. To interpret this contract provision so as to grant each of those agents the power to control which products MONY offers within that particular agent's jurisdiction would give Iglesias and other agents the power to veto policy decisions made by MONY's upper management at the company's national headquarters. This would flip the employment relationship between MONY and its agents on its head. Further, each of Iglesias's contracts contained clauses in which MONY reserved the right to change the contract. The district court's grant of summary judgment was entirely appropriate.

II. MONY's Counterclaim.

MONY reimbursed Iglesias for expenses that he incurred as a result of his professional activities. To supplement his income, Iglesias admits that he "padded" his expense reimbursement requests from 1981 to 1988. MONY first learned of Iglesias's practice of overstating his expenses in 1987. At that time, MONY took no legal action against Iglesias. Instead, after notifying him that his conduct conflicted with company policy, MONY requested that he submit accurate reports in the future. Subsequently, Iglesias brought the present action against MONY for discrimination and breach of contract. On November 7, 1991, in the course of a deposition, Iglesias admitted that he had continued submitting exaggerated expense reports even after MONY's 1987 warning.

MONY tried to use the information obtained at the 1991 deposition in three ways. First, MONY questioned Iglesias about his expense reports during trial in an effort to impeach his credibility. Second, MONY attempted to raise the falsified expense requests as an affirmative defense to Iglesias's discrimination claims, arguing that they provided MONY with a valid non-discriminatory reason to terminate Iglesias. The court did not allow MONY to pursue this strategy because MONY learned of these improprieties only after it had terminated Iglesias.

Third, and more importantly for our purposes, MONY sought leave to amend its answer to add a counterclaim for restitution. On March 17, 1992, the magistrate judge granted MONY's request. For four years, the parties conducted discovery on the counterclaim. On August 6, 1996, Iglesias moved for dismissal of MONY's counterclaim arguing, inter alia, that it was barred by laches. On August 14, 1996, the district court dismissed [241] the counterclaim as untimely. MONY now appeals from that dismissal.

Although neither of the parties has raised the issue, we have an obligation to inquire into our subject matter jurisdiction over MONY's counterclaim. See Clark v. Paul Gray, Inc., 306 U.S. 583, 588, 59 S.Ct. 744, 83 L.Ed. 1001 (1939) (holding that although district court's jurisdiction had not been challenged, Court had duty to raise question of whether jurisdictional amount was involved). See also White v. Gittens, 121 F.3d 803, 806 (1st Cir.1997). Federal courts are courts of limited jurisdiction; a court may not address the merits until "after ... the court has assumed jurisdiction over the controversy." Bell v. Hood, 327 U.S. 678, 681, 66 S.Ct. 773, 90 L.Ed. 939 (1946); see also Steel Co. v. Citizens for a Better Environment, ___ U.S. ___, ___, 118 S.Ct. 1003, 1012, 140 L.Ed.2d 210 (1998) (stating that deciding the merits before confirming jurisdiction "carries the courts beyond the bounds of authorized judicial action and thus offends fundamental principles of separation of powers"). Since MONY's counterclaim sounds in state law, a federal court may not hear it unless it falls within the ambit of supplemental jurisdiction or is supported by an independent jurisdictional basis. See Toste Farm Corp. v. Hadbury, Inc., 70 F.3d 640, 646 (1st Cir.1995) (noting that a federal court has jurisdiction over a counterclaim only if it resides within the court's supplemental jurisdiction or is supported by an independent jurisdictional basis).

The parties appear to have assumed that MONY's counterclaim was within the court's supplemental jurisdiction. When considering supplemental jurisdiction, the nature of the counterclaim is crucial. Federal Rule of Civil Procedure 13 describes two types of counterclaims: compulsory and permissive. A compulsory counterclaim is one that "arises out of the transaction or occurrence that is the subject matter of the opposing party's claim." Fed.R.Civ.P. 13(a). All counterclaims that are not compulsory are permissive. See Fed.R.Civ.P. 13(b). Only compulsory counterclaims can rely upon supplemental jurisdiction; permissive counterclaims require their own jurisdictional basis.

The supplemental jurisdiction statute provides: "[I]n any civil action of which the district courts have original jurisdiction, the district courts shall have supplemental jurisdiction over all other claims that are so related to claims in the action within such original jurisdiction that they form part of the same case or controversy under Article III of the United States Constitution." 28 U.S.C. § 1367(a). Section 1367 incorporates the common law doctrines of pendent and ancillary jurisdiction. See Penobscot Indian Nation v. Key Bank, 112 F.3d 538, 563 (1st Cir.1997), cert. denied, ___ U.S. ___, 118 S.Ct. 297, 139 L.Ed.2d 229 (1997). Federal courts have ancillary jurisdiction over compulsory counterclaims. See McCaffrey v. Rex Motor Transp., Inc., 672 F.2d 246, 248 (1st Cir.1982); 6 Charles Alan Wright, Arthur R, Miller & Mary Kay Kane, Federal Practice & Procedure § 1414, at 99 (2d ed.1990). Permissive counterclaims, however, do not fall within ancillary jurisdiction and therefore may not be heard in federal court unless supported by an independent basis of jurisdiction. See McCaffrey, 672 F.2d at 248; Wright, Miller & Kane, § 1422, at 169-70.

There are at least four tests to determine whether a counterclaim is compulsory or permissive:

1) Are the issues of fact and law raised by the claim and counterclaim largely the same?

2) Would res judicata bar a subsequent suit on defendant's claim absent the compulsory counterclaim rule?

3) Will substantially the same evidence support or refute plaintiff's claim as well as defendant's counterclaim?

4) Is there any logical relation between the claim and the counterclaim?

McCaffrey, 672 F.2d at 248 (quoting 6 Wright, Miller & Kane, supra, § 1410, at 42 [currently 52-57]) (internal quotation marks omitted). Of these tests, only the fourth — the "logical relation" test — could possibly encompass MONY's counterclaim. In McCaffrey, we adopted the Fifth Circuit's treatment of the logical relation test. Under this [242] approach, a claim qualifies as compulsory only if:

it arises out of the same aggregate of operative facts as the original claim in two senses: (1) that the same aggregate of operative facts serves as the basis of both claims; or (2) that the aggregate core of facts upon which the original claim rests activates additional legal rights in a party defendant that would otherwise remain dormant.

McCaffrey, 672 F.2d at 249 (quoting Revere Copper & Brass, Inc. v. Aetna Cas. & Sur. Co., 426 F.2d 709, 715 (5th Cir.1970)).

Applying this standard, MONY's counterclaim is permissive. First, the aggregate of operative facts giving rise to Iglesias's claims is entirely different from the facts supporting MONY's counterclaim. Iglesias alleges that MONY removed its disability insurance product from the Puerto Rico market in 1981 for discriminatory reasons. That change in policy, he claims, led to his eventual termination, which therefore was also discriminatory. The evidence relevant to these claims includes the terms of Iglesias's employment contract and documentary, testimonial or other evidence concerning MONY's actions in Puerto Rico. By contrast, MONY's counterclaim is for restitution of monies that Iglesias wrongfully obtained from the company over the period from 1981 to 1988. The facts supporting the restitution claim include Iglesias's records and expense vouchers for travel and meals, his admissions during deposition, and documentary, testimonial or other evidence regarding MONY's reimbursement policies. Notwithstanding both claims arose during the same time period, they therefore rest on different sets of supporting facts.

Second, MONY's ability to recover restitution does not depend on the success or failure of Iglesias's claims. Iglesias has admitted that he overbilled his expenses to enhance his compensation. Although MONY discovered Iglesias's false reimbursement requests for 1988 in a deposition that was part of the present action, Iglesias's claims did not "activate" otherwise dormant claims. MONY could have sought reimbursement for its payments based on Iglesias's false expense reports prior to 1988 at any point. Even MONY's claim relating to 1988 expenses is not tied to the merits of the discrimination and contract claims. In other words, MONY's counterclaim bears no logical relation to the claims in the main action.

We conclude that MONY's counterclaim is permissive rather than compulsory, and is not supported by federal supplemental jurisdiction. We look, therefore, to whether it rests on an independent federal jurisdictional basis.

The only possible jurisdictional basis would be diversity of citizenship. While Iglesias and MONY are citizens of different states, MONY did not allege that its claim exceeded any specified sum, nor is its amount ascertainable from the record. See 28 U.S.C. § 1332(a) (1993) (limiting diversity jurisdiction to claims that exceed a certain amount-in-controversy). At the time that MONY asserted its counterclaim, the amount-in-controversy required was in excess of $50,000, exclusive of interest and costs.^[3] The burden of alleging jurisdiction is on the party asserting the claim. See Viqueira v. First Bank, 140 F.3d 12, 16 (1st Cir.1998) (stating that "the proponent ... must carry the burden of demonstrating the existence of federal jurisdiction").

Although "the sum claimed by the [proponent] controls if the claim is apparently made in good faith," Saint Paul Mercury Indem. Co. v. Red Cab Co., 303 U.S. 283, 288, 58 S.Ct. 586, 82 L.Ed. 845 (1938) (footnote omitted), MONY, as noted, has failed to allege that its counterclaim exceeds the required amount-in-controversy nor is there evidence in the record from which its value can reliably be determined as exceeding that amount. MONY stated merely that it was:

in the process of reviewing each and everyone [sic] of the expense reports submitted by Mr. Iglesias to determine what was the scope of this practice by Mr. Iglesias and the approximate amount of funds wrongfully taken by Mr. Iglesias from Mutual [243] through the scheme of false expense reports acknowledged by the plaintiff.

Later, MONY submitted a more detailed accounting of specific instances in which Iglesias alleged excessive or false mileage, but did not state the value of the wrongful claims. MONY noted that Iglesias had charged $48,500.87 in meal expenses between 1981 and 1988, but it did not indicate what portion of these claims was false or overstated.

In these circumstances — lacking MONY's allegation that the counterclaim meets the jurisdictional amount, and lacking facts of record indicative that this is so — we must conclude that the district court was without jurisdiction to proceed. We, therefore, remand the counterclaim with directions that it be dismissed for want of jurisdiction. See Fed.R.Civ.P. 12(h)(3)("Whenever it appears by suggestion of the parties or otherwise that the court lacks jurisdiction of the subject matter, the court shall dismiss the action."). See also United Food & Com. Workers Union v. CenterMark Properties Meriden Square, Inc., 30 F.3d 298, 306 (2d Cir.1994) (addressing for first time on appeal whether district court lacked subject matter jurisdiction and remanding for determination of whether amount in controversy requirement had been met).

We recognize that Iglesias never pointed out below the absence of a proper amount-in-controversy allegation, nor was the subject raised by anyone. Had it been raised, MONY might have been able to remedy the deficiency by amendment if it believed in good faith that its counterclaim exceeded the jurisdictional minimum. In these circumstances, the dismissal we now order shall be without prejudice to MONY's right, if it can do so in good faith, to amend the counterclaim within such time as the district court allows so as to allege a supportably proper jurisdictional amount.^[4] If MONY can do so, the district court may lift the jurisdictional bar. Cf. United Food, 30 F.3d at 306. If that occurs, the court will be faced with whether to reinstate its previous dismissal with prejudice for untimeliness. We comment briefly.

Because jurisdiction over the counterclaim has not been established, we are without authority to rule whether the district court's dismissal for untimeliness was proper. Cf. Steel Co., ___ U.S. at ___, 118 S.Ct. at 1012. The issue was, however, presented in this appeal and we believe it would be helpful if we were to state our initial impressions, recognizing their present non-binding character. MONY brought its counterclaim only four months after Iglesias's November 7, 1991 deposition revealing for the first time that he had continued to falsify his expenses even after similar improprieties were noted in 1987. In all the circumstances, it is not readily apparent why MONY should be thought tardy in bringing the counterclaim when it did.

Iglesias argues that the counterclaim is barred by laches. The equitable doctrine of laches allows a court to dismiss a claim "where a party's delay in bringing suit was (1) unreasonable, and (2) resulted in prejudice to the opposing party." K-Mart Corp. v. Oriental Plaza, Inc., 875 F.2d 907, 911 (1st Cir.1989). Even if we were to find unreasonable delay within the period between 1987, when MONY first learned that Iglesias had been padding expenses, and 1992, when, after having warned Iglesias earlier, MONY asserted its counterclaim based on the subsequent defalcations, we fail to see how any of that delay prejudiced Iglesias. Indeed, such prejudice as there was appears to have worked the other way. The magistrate judge had granted permission to MONY to assert its counterclaim, and the parties thereafter conducted four years of discovery. MONY sifted through receipts for meals and business trips in an effort to reveal false claims. The court's dismissal after four years of discovery had been allowed seems to work a hardship on MONY. Given these equities, we would have serious questions as to the strength of Iglesias's laches argument.

The judgment for MONY against Iglesias is affirmed. The order dismissing the counterclaim [244] for untimeliness is vacated and the counterclaim remanded with directions to dismiss the counterclaim without prejudice for want of jurisdiction. Costs awarded to MONY on the appeal from the district court's grant of summary judgment; each party to bear its own costs on the appeal from the dismissal of the counterclaim.

[1] Iglesias sued for discrimination under Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e (1994); the Age Discrimination in Employment Act, 29 U.S.C. §§ 621 et seq. (1985); and Law 100 of June 30, 1959, P.R. Law Ann. tit. 29, § 146 (1985).

[2] According to Iglesias, a MONY manager had told him in 1988: "Either you retire or I'll retire you." At trial, Iglesias said that he did not respond to the February 9, 1989 letter from MONY because he "felt humiliated by what [MONY] was offering me because [MONY] was firing me." Thus, Iglesias was aware that he was being terminated in early 1989.

[3] After MONY asserted its counterclaim, § 1332(a) was amended to require an amount-incontroversy greater than $75,000. See Pub.L. No. 104-317, § 205(a)(1)(1996).

[4] A cursory examination leaves it unclear that the applicable amount can be asserted in good faith, but we leave that issue to MONY and the district court in the first instance.