Who Controls Agency Discretion?

At this point in the semester it should be clear to you that agencies have lots of discretion. There's nothing necessarily nefarious about this: after all, if agencies have discretion, it is because Congress and the President passed legislation to give them that discretion. The very name--"agency"--nods to "principal-agent theory," which involves delegation from a principal (the one ultimately in charge) to an agent (the one who is given responsibility for acting on behalf of the principal).

Still, a lot of people worry about this discretion. In any principal-agent relationship, the principal worries about "agency slack," which is just a fancy term for the agent going rogue. Principals wouldn't get much mileage from delegating to agencies if they had to confirm every detail of what agencies are doing, but too much of a hands-off approach to monitoring the agent allows the agent to take advantage of their discretion to pursue the agent's goals rather than the principal's. Ultimately, what the principal wants is to retain enough control to ensure that the agency is not systematically exploiting agency slack, but not so much that it obliterates the benefits of delegating discretion to the agent in the first place. In short, the principal wants "optimal political control of the bureaucracy." See Matthew C. Stephenson, Optimal Political Control of the Bureaucracy, 107 Mich. L. Rev. 53 (2008).

In some sense, everything in administrative law is part of this effort to optimally control the discretion of agencies. Procedures are a form of political control--they make certain outcomes more likely and some less likely, and they "stack the deck" in favor of constituencies that Congress means to benefit. See Mathew D. McCubbins, Roger G. Noll, & Barry R. Weingast, Structure and Process, Politics and Policy: Administrative Arrangements and the Political Control of Agencies, 75 Va. L. Rev. 431 (1989). Administrative procedures are a textbook example of ex ante political controls--ex ante because they operate even before agencies act.

This subsection focuses on another form of political control: ex post political control, meaning control after the agency has acted. It focuses on the ex post controls that Congress has used to monitor and reverse agency decisions, and also on the ex post political controls that the President, acting through the Office of Management and Budget's subagency, the Office of Information and Regulatory Affairs (OIRA), has deployed to review proposed agency regulations. Judicial review of agency action and inaction (covered in Section 4 of this casebook) for compliance with the law is also an ex post method of political control.

As you read the materials in this section, ask yourself whether you think that these controls seem to contribute to an "optimal" amount of control of administrative discretion, or whether they in fact seem to be miscalibrated: either the controls are not enough to rein in agency discretion or are gross overkill. Relatedly, notice that the perspectives of Congress and the President on whether agencies need to be corrected may not be aligned. Do the controls you are reviewing seem to respond to those concerns about multiple and competing principals? Does one principal seem more dominant than the other? Are agencies able to play principals off against each other to avoid any real oversight?

Ultimately, this subsection raises a perennial question in administrative law about where, exactly, agencies fall in the "organizational chart" of government--that is, who agencies ultimately need to answer to. The formal answer to the question is that agencies are part of the Executive Branch, even when they exercise powers that appear to be legislative and adjudicative as well. See City of Arlington v. FCC, 133 S. Ct. 1873 n.4 (2013). Yet, in practice, lots of putative principals across the three branches of government compete for control of the bureaucracy, with varying degrees of success. According to some, this makes the administrative state more like a "headless fourth branch" of government than anything else. See Jennifer L. Selin, The Headless Fourth Branch: Rethinking the Assumptions of Administrative Jurisprudence, 4 Persp. on Pub. Mgmt. & Governance 170 (2021). Are they right?