Review of the Exercise of Discretion

Prepared for the Scope of Judicial Review portion of the ABA Administrative

Law Section’s Project on the Administrative Procedure Act

Lisa Schultz Bressman, Vanderbilt Law School

Third Draft: July 2001

Restatement of the Law

A court may set aside an agency action as an abuse of discretion under the “arbitrary and capricious” test. The word “arbitrary” suggests the deferential “minimal rationality” standard of review applicable to legislative policy decisions under the Due Process Clause. The precise standard of review – or more specifically, the precise level of deference – applicable to administrative policy decisions is not entirely clear, however. The Supreme Court has sent conflicting signals. It has described the arbitrary and capricious inquiry as “searching and careful.” Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416 (1971). At the same time, it has characterized “the ultimate standard of review as a narrow one” and admonished courts not to “substitute [their] judgment for that of the agency.” Id; see also Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto Ins. Co., 463 U.S. 29, 43 (1983). Lower courts routinely invoke both of these sentiments, either ignoring the tension between them or eliminating the tension by emphasizing one or the other to fit the level of scrutiny ultimately applied.

Commentators also have debated the appropriate level of scrutiny. See Patricia M. Wald, Judicial Review in Midpassage: The Uneasy Partnership Between Courts and Agencies Plays On, 32 Tulsa L.J. 221, 257 (1996) (“After fifty years . . . [w]e have yet to agree on how this review should operate in practice. We are still struggling with where to draw the line between obsequious deference and intensive scrutiny.”). Some have defended relatively stringent review as “a device, admittedly highly imperfect, for reducing the risk that agency decisions will result from ‘political’ considerations that are sometimes illegitimate and at any rate ought not to be concealed.” Cass R. Sunstein, Deregulation and the Hard-Look Doctrine, 1983 Sup. Ct. Rev. 177, 211. Others have argued for minimal review on the theory that agency decisions appropriately reflect political judgments that courts should respect rather than impede. See Antonin Scalia, Chairman’s Message, 34 Admin. L. Rev. v. (Summer 1982).

More recently, the debate over the level of deference due administrative policy judgments has shifted to the question whether intrusive judicial review causes “ossification” of the rulemaking process. Many commentators have claimed that intensive judicial review forces agencies to devote excessive resources to insulating their policy decisions from reversal (or to rehabilitating those policy decisions on remand) and often discourages agencies from issuing such decisions at all. See, e.g., Thomas O. McGarity, The Courts and the Ossification of Rulemaking: A Response to Professor Seidenfeld, 75 Tex. L. Rev. 525 (1997); Richard J. Pierce, Jr. Judicial Review of Agency Actions in a Period of Diminishing Agency Resources, 49 Admin. L. Rev. 61 (1997); Stephen G. Breyer, Breaking the Vicious Circle – Toward Effective Risk Regulation 48 (1993); Thomas O. McGarity, Some Thoughts on “Deossifying” the Rulemaking Process, 41 Duke L.J. 1385 (1992); Jerry L. Mashaw & David L. Harfst, The Struggle for Auto Safety, 19, 199, 224-55 (1990); Richard J. Pierce, Jr., Two Problems in Administrative Law: Political Polarity on the District of Columbia Circuit and Judicial Deterrence of Agency Rulemaking, 37 Duke L. J. 300 (1988). Other commentators have disputed this claim as an empirical matter. See, e.g., William S. Jordan III, Ossification Revisited: Does Arbitrary and Capricious Review Significantly Interfere with Agency Ability to Achieve Regulatory Goals Through Informal Rulemaking?, 94 Nw. U. L. Rev. 393 (2000). Still others have acknowledged the risk of agency paralysis but have maintained that the risk is worth running to improve the quality of rulemaking. See, e.g., Thomas O. Sargentich, The Critique of Active Judicial Review of Administrative Agencies: A Reevaluation, 49 Admin, L. Rev. 599 (1997); Mark Seidenfeld, DeMystifying Deossification: Rethinking Recent Proposals to Modify Judicial Review of Notice and Comment Rulemaking, 75 Tex. L. Rev. 483 (1997); Patricia M. Wald, Judicial Review in the Time of Cholera, 49 Admin. L. Rev. 659 (1997); Jim Rossi, The Hard Look Doctrine and Federal Regulatory Efforts to Restructure the Electric Utility Industry, 1994 Wis. L. Rev. 763 (1994).

This debate aside, one aspect of arbitrary and capricious review has become relatively certain. In practice, the ground for setting aside an agency action under the arbitrary and capricious test varies according to the nature and magnitude of that action. Thus, a court typically will apply the criteria set forth in this section rigorously during judicial review of high-stakes rulemaking proceedings (a practice commonly termed “hard look” review), but much more leniently when reviewing an adjudicative matter that an agency would be expected to dispose of quickly. Compare State Farm, 463 U.S. 29, 43 (applying the requirement that an agency explain the basis for its decision stringently to NHTSA’s rescission of its passive restraint rule) withCamp v. Pitts, 411 U.S. 138, 142-43 (1973)(applying the requirement less stringently to the Comptroller of the Currency’s action on an application for a certificate authorizing organization of a new national bank) and Dunlop v. Bachowski, 421 U.S. 560, 572-75 (1975) (applying the requirement less stringently to the Secretary of Labor’s determination not to sue to set aside an allegedly invalid labor union election).

Before discussing the criteria of the arbitrary and capricious test, one other general issue is important to mention. That issue concerns the relationship between the arbitrary and capricious test and Chevron Step Two. Chevron Step Two directs courts to defer to “reasonable” agency interpretations of ambiguous statutory provisions. Such interpretations constitute policy decisions of the sort to which the arbitrary and capricious test also applies. See APA § 706. A question arises whether the Chevron Step Two “reasonableness” test incorporates, overlaps, or remains distinct from the arbitrary and capricious test. This question is addressed in [Liz’s section]. For purposes of this section, note that many cases applying the arbitrary and capricious test do so to assess the “reasonableness” of an agency interpretation either as part of the Chevron Step Two inquiry or as a separate requirement. In these cases, the components of the arbitrary and capricious test are those set forth in this section.

It is now possible to turn to the components of arbitrary and capricious review. A court may set aside an agency action under the arbitrary and capricious test on any of several grounds. As one lower court has aptly noted, “the ‘arbitrary or capricious’ concept, needless to say, is not easy to encapsulate in a single list of rubrics because it embraces a myriad of possible faults and depends heavily upon the circumstances of the case.” Puerto Rico Sun Oil Co. v. EPA, 8 F.3d 73, 77 (1st Cir. 1993). There are “rules of thumb,” however. Id. The place to start is with the Supreme Court’s list of examples in the State Farm case. In State Farm, the Court said that “the agency must examine the relevant data and articulate a satisfactory explanation for its action including a ‘rational connection between the facts found and the choice made.’” State Farm, 463 U.S. at 43 (quoting Burlington Truck Lines v. United States, 371 U.S. 156, 168 (1962)). In reviewing that explanation, the Court continued, “we must ‘consider whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment.’” Id. (quoting Bowman Transp. Inc. v. Arkansas-Best Freight System, 419 U.S. 281, 285 (1975) and citing Overton Park, 401 U.S. at 416)). Furthermore, the Court stated, “an agency rule would be arbitrary and capricious if the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.” Id. It also would be arbitrary and capricious if the agency failed to consider certain policy alternatives. Id. at 51. The following subsections organize and amplify this language, with the exception of the “relevant factors” language, which is better understood as part of the court’s legal analysis – that is, part of the analysis whether the agency has considered all relevant legal factors, such as all of the relevant statutory requirements. [Cross reference Beth’s setion?]

B. The action (including the agency’s choice of remedy in a particular case) rests upon a policy judgment that is seriously unacceptable, [although a court will not substitute its judgment of sound policy for that of the agency].

This subsection corresponds to the “clear error of judgment” test of Overton Park, 401 U.S. at 416. The Overton Park phraseology should not be confused with similar language in Federal Rule of Civil Procedure 52, under which district court findings of fact must stand unless they are “clearly erroneous.” Nothing in Overton Park suggests that the similarity is more than mere coincidence. Furthermore, courts and commentators long have recognized that Rule 52 criteria are out of place in this context. SeeNAACP v. FCC, 682 F.2d 993, 998 n.4 (D.C. Cir. 1982); J. Skelly Wright, The Courts and the Rulemaking Process: The Limits of Judicial Review, 59 Cornell L. Rev. 375, 392 n. 84 (1974); see also Stephen G. Breyer, Richard B. Stewart, et al., Administrative Law and Regulatory Policy 366 (4th ed. 1998) (“[O]n the conventional view, reviewing courts should be more deferential to agency policy decisions than to findings of fact by lower courts.”). To avoid confusion, the present report captures the Overton Park test by referring to “a policy judgment that is seriously unacceptable.” This formulation also is meant to capture the State Farm reference to an agency rule that is “so implausible that it could not be ascribed to a difference in view or the product of agency expertise.” State Farm, 463 U.S. at 43.

Courts rarely overturn an agency action simply on the ground that the substance of that decision is “unacceptable” or unwise. Perhaps courts refrain from doing so because the Supreme Court has warned them not to “substitute [their] judgment for that of the agency.” Overton Park, 401 U.S. at 416. In any event, courts clearly prefer to rely on the less overtly value-laden rationales described in later subsections. Although courts once examined whether an agency has made an unpalatable discretionary choice as a core part of “arbitrary and capricious” review, see, e.g.,ICC v. Illinois Cent. R.R., 215 U.S. 452, 470 (1910) (reviewing agency policy choice for “unreasonableness” as distinct from constitutional or statutory violations), they no longer perform this function vigorously.

Nonetheless, courts occasionally do invoke the irrationality of an agency’s policy decision as a reason for invalidating it. See, e.g., Athens Community Hosp. , Inc. v. Shalala, 21 F.3d 1176 (D.C. Cir. 1994); Dugan v. Ramsay, 727 F.2d 192 (1st Cir. 1984); Conway v. Watt, 717 F.2d 512 (10th Cir. 1983); Community Nutrition Inst. v. Bergland, 493 F. Supp. 488 (D.D.C. 1989). It is possible to re-characterize some of these cases under other subsections. It also is possible to break them out to the extent they find fatal substantive flaws in the rules rather than potentially curable process defects, such as reliance on flawed reasoning, see § V(D).

There is another set of cases that illustrate the § V(B) inquiry. These cases deal with judicial review of remedies – for example, when a court vacates or modifies the relief ordered by an agency because it deems the remedy to be too severe, FTC v. Royal Milling Co., 288 U.S. 212, 217 (1933); Chrysler Corp. v. FTC, 561 F.2d 357 (D.C. Cir. 1977); Spiegel, Inc. v. FTC, 540 F.2d 287 (7th Cir. 1976), or unfairly retroactive, Buccaneer Point Estates, Inc. v. United States, 729 F.2d 1297 (11th Cir. 1984); McDonald v. Watt, 653 F.2d 1035 (5th Cir. 1981); NLRB v. E & B Brewing Co., 276 F.2d 594 (6th Cir. 1960), cert. denied, 366 U.S. 908 (1961). In such cases, it is hard to see the courts as doing anything other than declaring that the agency has made an improvident discretionary decision. Of course, this judicial power should be restricted to correction of “arbitrary” errors and does not extend to every case where a court disagrees with an agency’s choice of remedy. See Butz v. Glover Livestock Comm’n Co., 411 U.S. 182 (1973).

It is impossible to generalize when a court is likely to find “an unacceptable policy judgment.” The § V(B) inquiry obviously is subjective and context dependent. (Indeed, this is true about the inquiries in the remaining subsections as well.) The point to make for present purposes is that this type of review theoretically is available to courts, even though they rarely use it in practice.

C. The asserted or necessary factual premises of the action do not withstand scrutiny under the relevant standard of review (see § IV).

This provision essentially is a cross reference to § IV, which states in detail the standards of review for factual issues. It is included here so that this section will contain a comprehensive list of the types of errors that can lead to reversal.

The phrase “asserted or necessary” comes from the first version of this report, which in turn states that it comes from certain regulatory reform bills. The “necessary” factual premises of an agency action are derived from substantive law. When a court decides what legal test the agency is required to apply, it also determines what factual findings are necessary to support the action. For example, in Industrial Union Dep’t, AFL-CIO v. American Petroleum Inst. (the Benzene Case), 448 U.S. 607 (1980), the Supreme Court held that the Occupational Safety and Health Act required OSHA to demonstrate that exposure to a toxic substance presented a “significant risk” of harm before regulating that substance. After the Benzene Case, a factual finding of “significant risk” is necessary to the validity of any OSHA toxic substance regulation.

The “asserted” factual premises are determined by the agency itself. When an agency actually relies on a factual finding in defending its action, it subjects that finding to scrutiny under the standards of § IV – regardless of whether the finding is “necessary,” as defined above. The requirement that “asserted” factual findings pass muster under § IV follows from the principle that an agency action can be sustained only on the grounds upon which the agency relied. See § VI. (The factual assertion need not be documented, of course, if the action can be sustained on an alternative and independent basis upon which the agency relied. See id.)

D. The action is unsupported by any explanation or rests upon reasoning that is seriously flawed.

This subsection addresses weaknesses in the agency’s chain of reasoning. It recognizes that even when an agency’s factual premises themselves are sustainable (as determined under § V(C)), the agency must show a “rational connection between the facts found and the choice made,” and may not “offer[] an explanation for its decision that runs counter to the evidence” before it. State Farm, 463 U.S. at 43 (quotations omitted). At the same time, a court must “uphold a decision of less than ideal clarity if the agency’s path may be reasonably discerned.” Bowman Transp., Inc. v. Arkansas-Best Freight Sys., Inc., 419 U.S. 281, 286 (1974). As one lower court explained, an agency “‘must inform the court and the petitioner of the grounds of decision and the essential facts upon which the administrative decisions was based.’” U.S. v. Dierckman, 201 F.3d 915, 926 (7th Cir. 2000) (citing Bagdonas v. Dep't of the Treasury, 93 F.3d 422, 426 (7th Cir. 1996) quoting Kitchens v. Dept. of the Treasury, 535 F.2d 1197, 1200 (9th Cir. 1976)).

At bottom, this subsection covers instances in which an agency provides an explanation for its policy choice that is inadequate to suggest reasoned decisionmaking. “Inadequate explanation” is the most commonly offered reason for invalidating an agency policy decision. It covers a variety of errors and takes a variety of forms. The easiest to spot occurs when an agency fails to provide any explanation whatsoever for its policy choice. Although courts occasionally attribute such omission to oversight, they find more often that it indicates failure to engage in reasoned decisionmaking. See, e.g.,American Petroleum Inst. v. EPA, 216 F.3d 50, 57-58 (D.C. Cir. 2000); United Transportation v. ICC, 52 F.3d 1074 (D.C. Cir. 1995); Allied-Signal, Inc. v. U.S. Nuclear Regulatory Comm’n, 988 F.2d 146, 150 (D.C. Cir. 1993); National Tank Truck Carriers, Inc. v. EPA, 907 F.2d 177, 184-85 (D.C. Cir. 1990); UMWA v. Dole, 870 F.2d 662, 673 (D.C. Cir. 1989). At a minimum, the absence of any explanation leaves courts without the ability to judge whether the agency has engaged in reasoned decisionmaking.

Even when an agency does provide an explanation for its policy choice, it nonetheless may convey less than fully rational decisionmaking if it relies on faulty reasoning. Sometimes it relies on unconvincing or implausible reasoning. See, e.g., Fresno Mobile Radio, Inc. v. FCC, 165 F.3d 965, 969-70 (D.C. Cir. 1999), Time Warner Entertainment Co. v. FCC, 56 F.3d 151, 209-10 (D.C. Cir. 1995); Building & Construction Trades Dept., AFL-CIO v. Brock, 838 F.2d 1258, 1277 (D.C. Cir. 1988); Alltel Corp. v. FCC, 838 F.2d 551, 558-59 (D.C. Cir. 1988). Sometimes it relies on illogical reasoning. See, e.g., Exxon Co., U.S.A. v. FERC, 182 F.3d 30, 42 (D.C. Cir. 1999). Other times it simply offers insufficient reasoning. See, e.g., Air Transport Ass’n of America v. Dept. of Transportation, 119 F.3d 38, 44-45 (D.C. Cir. 1997); Petroleum Communications, Inc. v. FCC, 22 F.3d 1164, 1172-73 (D.C. Cir. 1994); American Mining Congress v. EPA, 907 F.2d 1179, 1190 (D.C. Cir. 1990); Hazardous Waste Treatment Council v. EPA, 886 F.2d 355, 365-66 (D.C. Cir. 1989). It also may rely on inconsistent reasoning. See, e.g., Engine Manufacturers Ass’n v. EPA, 20 F.3d 1177, 1182 (D.C. Cir. 1994); Illinois Bell Telephone v. FCC, 911 F.2d 776, 784 (D.C. Cir. 1990). Or it may rely on skewed or biased reasoning. See, e.g., Sierra Club v. Thomas, 105 F.3d 248, 251-53 (5th Cir. 1997). Many of these categories overlap. Moreover, they are intended to be illustrative rather than exhaustive.

Two issues arise that are of specific concern to scientific or technical policy decisions: (1) the extent to which agencies are entitled to rely on predictive judgments, and (2) the extent to which agencies are entitled to rely on broad models or tests. Courts routinely state that agencies are entitled to significant latitude with respect to scientific or technical decisions. See, e.g.,Baltimore Gas & Elec. Co. v. NRDC, 462 U.S. 87, 103 (1983); American Iron and Steel Inst. v. EPA, 115 F.3d 979, 1006 (D.C. Cir. 1997); American Legion v. Derwinski, 54 F.3d 789, 795 (D.C. Cir. 1995). Thus, agencies are entitled to rely on predictive judgments when they possess an imperfect basis for more concrete judgments. See, e.g., People of the State of California v. FCC, 75 F.3d 1350, 1359 (9th Cir. 1996); Cellnet Communications. Inc. v. FCC, 149 F.3d 429, 441 (6th Cir. 1998); American Postal Workers Union v. U.S. Postal Serv., 891 F.2d 304, 413 (D.C. Cir. 1989), rev’d on other grounds, 498 U.S. 517 (1991). And they are entitled to rely on broad models or tests when faced with a wide range of potential applications. See, e.g., American Iron and Steel Inst. v. EPA, 115 F.3d at 1004; Eagle-Picher Indus., Inc. v. EPA, 759 F.2d 905, 921 (D.C. Cir. 1985). At the same time, courts find that agencies are not entitled to rely on mere speculation or unsubstantiated predictions. See, e.g., Cincinnati Bell Telephone Co. v. FCC, 69 F.3d 752, 760 (6th Cir. 1995); American Mining Congress v. EPA, 907 F.2d 1179, 1188 (D.C. Cir. 1990). Nor are they entitled to rely on models or tests that lack a rational relationship to their specific applications. See, e.g., Leather Industries of America, Inc. v. EPA, 40 F.3d 392, 402-03, 404-05 (D.C. Cir. 1994); Anne Arundel Country, Maryland v. EPA, 963 F.2d 412, 313-14 (D.C. Cir. 1992); Chemical Manufacturers Ass’n v. EPA, 28 F.3d 1259, 1265-66 (D.C. Cir. 1994); Edison Elec. Institute v. EPA, 2 F.3d 438, 446-47 (D.C. Cir. 1993). These divergent strains leave agencies without much guidance on when a court will reject the use of a predictive judgment or a general model. [Cross reference Jim’s section? Shorten this discussion?]