In 1946 Congress enacted a "comprehensive statement of the right, mechanics, and scope of judicial review."2 Section 10 of the Administrative Procedure Act was designed to prescribe "when there may be judicial review and how far the court may go in examining into a given case," and the courts were supposed to "prevent avoidance of the requirements of the [Act] by any manner or form of indirection.3"
That was the ambition of the 79th Congress, but the reality has quite been different. Decades after the enactment of the APA, Professor Kenneth Culp Davis accurately observed: "Perhaps about nine-tenths of American administrative law is judge-made law . . . . Most of it is common law in every sense, that is, it is law made by judges in absence of relevant constitutional or statutory provision."4 Common law has been particularly prevalent in judicial review, an area that Professor Louis Jaffe described in 1965 (again, quite accurately) as encompassing "a whole congeries of judicial theories and practices," which constitutes "the common law of review, and which is a significant part of the `administrative law' of the jurisdiction."5
A perfect example of the conflict between the APA and administrative common law is provided by Darby v. Cisneros, 509 U.S. 137 (1993). Darby involved a simple issue: whether a party aggrieved by administrative action must, prior to seeking judicial review, exhaust all administrative appeals. The APA provides that intra-agency appeals do not have to be pursued unless the agency requires the appeal by rule and suspends the effect of the administrative action pending the appeal. 5 U.S.C. Section 704. The judge-made doctrine of exhaustion gave a different answer: It required a party to exhaust all administrative remedies, including intra-agency appeals, that are "(a) available to him on his initiative (b) more or less immediately and (c) will substantially protect his claim of right."6 Darby abandoned the common law in favor of the APA rule.
Yet until Darby was decided in 1993—nearly a half century after 1946—the APA rule had been overshadowed by the judge-made doctrine. Indeed, the APA rule was not applied at all until the 1970's and, while a few circuit court decisions then considered the APA rule, most held that it did not supersede the common-law doctrine. Indeed, administrative common law held so much sway that the first circuit to follow the APA rule reversed course within seven years, holding that the APA could not limit exhaustion requirements "of judicial origin" because numerous cases "clearly demonstrate the existence of an active judicial control over certain exhaustion requirements."7 In 1983, Professor Davis lamented that the APA provision "is relevant in hundreds of cases and is customarily overlooked."8 Though Darby finally vindicated the APA's rule, even the Court expressed its "surpris[e]" both at the length of time needed for the issue to be resolved and at the mere "handful of opinions in the Courts of Appeals" that had considered the effect of the APA on the judge-made law of exhaustion. 509 U.S. at 145.
The clash between the APA and administrative common law, especially the administrative common law governing judicial review, is not unique to Darby. And, as in Darby, administrative common law is now finally beginning to be replaced by doctrine more consistent with the original meaning of the APA's judicial review provisions. The change is slow and halting, for courts are reluctant to abandon their own creations in favor of the statutory law created by Congress. Major judge-made doctrines such as the ubiquitous Chevron doctrine continue to thrive even though the courts have not even attempted to reconcile their doctrine with the statute. Yet undeniably the federal courts are treating the statutory law with more care. Why this change is occurring more than a half century after Congress enacted the APA is a fascinating tale—one that provides insights into our constitutional system and into the vast changes that the last five decades have wrought in the federal courts' conceptions of their own legitimate powers in that system.
Traditional Limitations on Judicial Lawmaking in the Federal Courts
The distinction between law based on written text (whether statutory or constitutional) and law fashioned by judges or "common law" has been important to the federal courts since the early Republic. As early as 1807, the Supreme Court in Ex parte Bollman held that there were basic differences between federal courts and "[c]ourts which originate in the common law." 8 U.S. (4 Cranch) 75, 93 (1807). Because federal courts "are created by written law," they may not decide "any question" unless "the power to determine [the question] . . . [is] given by written law." Id. at 93-94. By 1834, the Court found it "clear" that "there can be no common law of the United States" since "[t]here is no principle which pervades the union and has the authority of law, that is not embodied in the constitution or laws of the union."9
These traditional restrictions on federal common law reinforce two fundamental features of our Constitution: federalism and separation of powers. Federal common law threatens federalism because it circumvents the enumerated legislative powers of the federal government. If the federal courts could fashion law in any case over which they had jurisdiction, then there appears to be no good reason why the courts would be limited in that lawmaking enterprise to the list of enumerated powers in Article I, Section 8, which, after all, applies only to Congress, not to the courts. While this problem would be mitigated if the federal courts choose to confine their lawmaking to the enumerated powers, that "solution" would still leave a tremendous separation of powers problem.
Article I, Section 1 of the Constitution vests "All legislative Powers" in the Congress. This clause itself provides a sound textual basis for doubting that the federal courts have power to fashion law without legislative authorization. Structural features of the Constitution provide further reason. In designing the Article I legislative process, the framers included a host of checks to help insure that federal laws would be enacted by democratically accountable officials, and that the legislative process could move forward only with a consensus among political actors with different constituencies. For example, one does not have to be a fan of the ultimate result in Garcia v. San Antonio Metro. Transit Authority, 469 U.S. 528 (1985), to acknowledge that the Court correctly recognized that the federal legislative process contains many protections of State interests. Similarly, the bicameralism and presentment features of Article I help to protect the separation of powers by insuring that federal laws—including those that might shift power between the branches—are not enacted without either agreement between the Executive and Congress, or a supermajority in both Houses of Congress. These structural features provide good evidence that the framers did not intend the Article I legislative process to be circumvented by unaccountable courts manufacturing law sua sponte.
All of this is not to deny that broadly worded statutes will often generate a case law resembling common law. The Sherman Act is the most frequently cited example. Such statutes may be interpreted as authorizations for common law—i.e., as delegations to the judiciary to create law, just as standard administrative statutes delegate similar power to the Executive Branch. But one might expect that courts engaging in judicial review—courts demanding that executive branch agencies identify the basis in law for their actions—would be careful to identify the textual basis for their power to fashion judge-made law. The history of administrative common law shows that has not always been so.
The Administrative Common Law, the APA, and the New Federal Common Law
If the judicial review provisions in the APA were intended to provide a comprehensive statement of the right, mechanics and scope of judicial review, what explains the numerous judge-made doctrines such as exhaustion, ripeness and Chevron that dominate the area? And, perhaps more importantly, how can these doctrines be reconciled with basic restrictions on federal common law if the doctrines are (as K.C. Davis claimed) "law made by judges in absence of relevant constitutional or statutory provision"? The first question is easier, for a justification is more difficult than an explanation.
Judicial review in the early administrative era grew up in the federal equity jurisdiction, with parties wronged by administrative action seeking injunctive relief against arbitrary and unlawful actions by the responsible officials. Equity had been the domain of federal judge-made law since the founding of the Republic, and many doctrines of administrative common law grew up by analogy to other judge-made equity law. For example, the exhaustion doctrine began with equity courts drawing an analogy to the rule that parties having an adequate remedy at law were not entitled to equitable relief. The dominance of judge-made law in equity was, however, just another example of a statutorily-authorized common law. The statutes conferring equity jurisdiction had been interpreted, correctly in all likelihood, as vesting the federal courts with a power to administer a judge-made law of equity. But, while early administrative law was dominated by judge-made law, it remained closely tied to, and restrained by, traditional principles of equity. Thus, as late as 1944, two years prior to the enactment of the APA, Justice Frankfurter could plausibly claim that there was "no such thing as a common law of judicial review in the federal courts." Stark v. Wickard, 321 U.S. 288, 312 (1944) (Frankfurter, J., dissenting). But there was such a thing as the federal common law of chancery, and it provided an avenue of relief from unlawful administrative action.
Things should have changed in 1946. Statutory law should have assumed the dominant position in cases covered by the APA, which means just about all cases reviewing federal administrative action. Yet for decades, judge-made law continued to thrive. Perhaps it is not surprising that judges would cling to their roles as law-givers in the doctrinal area that was one of waning equity's most significant contributions to the twentieth century—judicial review over the newly created administrative agencies. But other forces were also working against the ascendancy of statutory law.
First, there was the influential Attorney General's Manual on the APA, a highly political document designed to minimize the impact of the new statute on Executive agencies. By characterizing the judicial review provisions of the APA as a mere "restatement" of "general principles" that "must be carefully coordinated with," and "interpreted in light of," existing case law, the AG's Manual went a long way in suggesting that the APA should be treated as less binding than other statutes.10 That suggestion was wrong. Attorney General Thomas Clark had no support in the text of the statute or in its legislative history for his broad gloss on the statute. In fact, most of the passages from the legislative history cited by Clark to bolster his "restatement" thesis were letters sent to Congress by Clark himself.11
Some members of the bar noticed that the Executive Branch was placing quite a spin on the new Act. For example, a 1947 editorial by the American Bar Association warned that Executive agencies were "striv[ing] to minimize the import and effect" of the APA's judicial review provisions by claiming that the Act "merely embodied" prior practice and procedure. Editorial, The Agencies and the Administrative Procedure Act, 33 A.B.A. J. 16 (1947). This "official interpretative pattern" showed a "[d]ivergence" between "the legislative intent" and "the views of those whose procedures and practices were to be regulated and curbed by the new Act." Id. Yet despite that warning, the Attorney General's shrewd restatement theory was soon reflected in the case law. See Willapoint Oysters, Inc. v. Ewing, 174 F.2d 676, 690 & n.22 (9th Cir. 1949) (citing the AG's Manual for the proposition that the judicial review provisions of the APA are a restatement). Even today, though the Supreme Court in 1990 rejected the view that "the judicial review provisions of the APA [are] no more than a restatement of pre-existing law," Lujan v. Defenders of the Wildlife, 497 U.S. 871, 883 (1990) (Scalia, J.), courts continue to be misled by the AG's Manual, see, e.g., Old Town Trolley Tours of Washington, Inc. v. Washington Metro Area Transit Comm'n, 129 F.3d 201, 204 (D.C. Cir. 1997) (citing the Manual for the restatement thesis).
But the AG's spin on the APA might have been less successful were it not for a pervasive and overarching intellectual movement in federal law. For the APA had misfortune of being enacted at the dawn of the era of "The New Federal Common Law" (in the words of Judge Friendly)—an era when leaders of the federal bar and bench were calling for an aggressive formulation of federal common law. For a time stretching from the late 1940's into the 1970's, this view reigned. To understand just how different that era is from our own time, one need only read Judge Friendly's article In Praise of Erieand of the New Federal Common Law, 39 N.Y.U. L. Rev. 383 (1964). Judge Friendly, it must be remembered, was one of the most careful, restrained and conservative judges of the era. Yet he praised the federal courts of his time for employing a variety of "techniques" to fashion federal common law, including implying private rights of action from statutory silence, interpreting grants of jurisdiction as mandates for federal judge-made law, and even creating law solely from the existence of federal interests—a technique he called the "spontaneous generation" of federal common law. Id. at 421. Moreover, Friendly claimed:
"One of the beauties of the . . . [New Federal Common Law] doctrine for our day and age is that it permits overworked federal legislators, who must vote with one eye on the clock and the other on the next election, so easily to transfer a part of their load to federal judges, who have time for reflection and freedom from fear as to tenure and are ready, even eager, to resume their historic law-making function—with Congress always able to set matters right if they go too far off the desired beam." Id at 419.
Democratic accountability was itself seen as a problem: The work of legislators was to be disfavored because they had "one eye on the next election," and the legislation of "eager" judges was to be favored because they have "freedom from fear as to tenure." With these assumptions, there was nothing so extraordinary with finding an authorization for judge-made law in "only the smallest bit of legislating," "a bit of legislative history," or even nothing at all. Id. at 414, 413 & 421.
The New Federal Common Law gave the courts not only a reason for departing from the APA, but also a doctrinal home once they left, for the administrative common law was part of a much larger edifice of federal common law. Precisely because so much federal common law existed, Louis Jaffe was able to organize his Judicial Control of Administrative Action (a work accurately described by Judge Posner as the "summa theologica of this era of administrative law scholarship.")12 around the basic theme that, in the words of Dean Daniel Rodriguez of the University of San Diego, "courts ought to use their federal common law powers to scrutinize administrative action."13 Jaffe's theme did not raise eyebrows in the era of New Federal Common Law, because the power of federal courts to fashion judge-made law was then seen as not only untroubling, but commendable. Administrative common law would flourish in that environment, and the APA would languish.
The Decline and Fall of the Administrative Common Law
Since at least Justice Powell's influential dissent in Cannon v. University of Chicago, 441 U.S. 677, 730 (1978), the New Federal Common Law has been receding into history. Federal judges are less "eager" to be federal law-givers and are devoting renewed attention to the traditional limits on their powers to act as common-law judges. This change has had profound effects across the law. Private rights of action are now seldom implied into statutory silence (contrast J.I. Case v. Borak, 377 U.S. 426 (1964)); whole fields of state law are no longer converted into federal law based only on a jurisdiction statute (contrast Textile Workers v. Lincoln Mills, 353 U.S. 448 (1957)). The shift is now being felt in administrative law.
As previously mentioned, Darby v. Cisneros is one of the best examples of a transition between judge-made law and statutory law. Another recent example is Dickinson v. Zurko, 119 S.Ct. 1816 (1999). In Zurko, the Supreme Court considered the Federal Circuit's long practice of reviewing decisions of the Patent and Trademark Office under a "clear error" standard , which, the Federal Circuit acknowledged, was a "common law" standard. As in Darby, the Supreme Court confronted an entrenched common law rule and rejected it in favor of the APA. Moreover, even the differences between Darby and Zurko show a continuing erosion in administrative common law. For in Zurko, the circuit court at least tried to reconcile its standard with the APA: The court relied on a savings clause in the APA that preserves "additional requirements . . . recognized by law," 5 U.S.C. Section 559, and held that this savings clause could preserve a more demanding standard of review in judge-made law prior to the APA. Though more accommodating to common law than the Supreme Court would allow, the Federal Circuit's reasoning nonetheless showed far more attention to the APA than, for example, that shown by the lower court in Darby, which ignored the APA entirely even though the party trying to avoid the common law exhaustion doctrine made the statute as his lead argument in briefing. The Zurko litigation demonstrates that even defenses of administrative common law are now litigated within the terms of the APA.
Other examples of administrative common law can be seen in the ripeness doctrine and in the Vermont Yankee decision, both of which I have detailed in a longer law review article, Administrative Common Law in Judicial Review, 77 Tex. L. Rev. 113 (1998). But the example with the most importance to modern administrative law is the Chevron doctrine.
No true fan of judicial restraint should be enamored of the opinion in Chevron USA v. NRDC, 467 U.S. 837 (1984), for it provides one of the best examples of a pure judicial lawmaking. The Chevron Court did not trouble itself to consider the APA or any other statutory authority; it justified its ruling with case law and its own assessment of the policy reasons (agency expertise and democratic accountability) for preferring agency interpretation over judicial interpretation. To leave no doubt as to its method, the Court even relied on Roscoe Pound's The Spirit of the Common Law. See id., at 844 n.(14).
Chevron was an APA case, so any attempt to justify its rule should begin with the APA. The first sentence of Section 706 of the APA requires a reviewing court to "decide all relevant questions of law" and to "interpret constitutional and statutory provisions." In the years immediately after the APA was enacted, courts routinely interpreted this language as an "explicit" command that "questions of law are for the courts rather than agencies to decide."14 Surprisingly enough, even after Chevron, federal appellate courts that focus on the APA rather than on administrative common law continue to read Section 706 as requiring de novo review on issues of law.15
The legislative history of the APA also leaves no doubt that, as Representative Francis Walter (the Chairman of the House Subcommittee on Administrative Law and author of the House Committee Report on the APA) explained to the House just before it passed the bill, Section 706 "requires courts to determine independently all relevant questions of law, including the interpretation of constitutional or statutory provisions."16 Even Justice Scalia, who otherwise defends Chevron, has noted that the 79th Congress was laboring under "the quite mistaken assumption that questions of law would always be decided de novo by the courts."17 But if Congress enacted that assumption into law with Section 706, what power do courts have to find the decision "mistaken"?
A proper analysis of the type of issue presented in Chevron—a challenge to an agency legislative rule as inconsistent with the agency's statute—begins with a frank acknowledgment that the APA does indeed require courts to decide all issues of law. That does not end the matter, however, because an agency may have a general delegation to make substantive rules. The APA itself equates rulemaking with the power to "prescribe law." 5 U.S.C. Section 551. Thus, even a court reviewing legal issues de novo must take account of the agency's substantive rulemaking power, and one quite plausible interpretation of such a power is that it allows an agency to "fill in the details" where other portions of the statute are silent or ambiguous. As Judge Posner notes, statutory authorizations for agencies to make substantive rules are like administrative "necessary and proper" clauses.18
In Chevron itself, the EPA had a general delegation to create substantive rules, and the government's brief explicitly relied on that delegation as the source of the agency's power. Furthermore, the Chevron Court described the EPA as "an agency to which Congress has delegated policymaking responsibilities," 467 U.S. at 865, and the Court has since made clear that "policymaking responsibility" flows from a substantive rulemaking power, see Martin v. OSHRC, 499 U.S. 144, 157 (1991). If the Chevron Court did not explicitly limit its holding to cases involving substantive rulemaking, the Court can be excused because of the terrible briefing on the issue—the respondents mentioned the standard of review only in a footnote conceding that an agency is entitled to deference in interpreting its statute unless the meaning of the statute is clear. Thus, given the context of the case, Chevron provides a incredibly weak basis for a general principle of deference that defies the plain command of Section 706 of the APA.
Reconciling Chevron with the original meaning of the APA will not produce a different result in every case. Where an agency has properly exercised a rulemaking power (as in Chevron itself), reviewing courts will have limited power to overturn the agency's rule: The court's power to "say what the law is" will be effectively limited by the agency's congressionally conferred power to make the law. But there are cases where the statutory and common law approaches diverge. One example is the current circuit split on whether agency interpretative rules are entitled to Chevron deference. Fidelity to the original meaning of APA would dictate de novo review; but the reasoning of Chevron might suggest deference. (Other examples of differences are considered in my longer article in the Texas Law Review.)
How any of these issues will be resolved is less important than how the courts approach the issues. If they regard these as matters to be resolved solely by reference to their own precedents and their own notions of good policy, then the courts will be taking a step back to the theory of the New Federal Common Law and undermining important limitations on the judiciary's role in the constitutional system. If, as the trend has been, they decide these issues with renewed respect for the work of the 79th Congress, they will be continuing the modern renaissance of the APA and restoring traditional restraints on federal judicial power.
1. * Associate Professor of Law, Cardozo School of Law, Yeshiva University. Visiting Assistant Professor of Law (Fall 1999), Marshall Wythe School of Law, College of William and Mary. The longer version of this article, which was published in the Texas Law Review (Administrative Common Law in Judicial Review) 77 Tex. L. Rev. 113 (1998), received the 1999 Scholarship Award of the ABA's Section of Administrative Law and Regulatory Practice.
2. Staff of Senate Comm. on the Judiciary, Legislative History of The Administrative Procedure Act, S. Doc. No. 79-248, at 368 (1946) (statement of Rep. Walter) [hereinafter APA Legislative History]. For more complete citations and additional authority, please see John F. Duffy, Administrative Common Law in Judicial Review, 77 Tex. L. Rev. 113 (1998).
3. APA Legislative History at 354 (statement of Rep. Walter); id. at 326 (statement of Sen. McCarran).
4. Kenneth Culp Davis, Administrative Law Treatise ` 2:18, at 140 (2d ed. 1978).
5. Louis L. Jaffe, Judicial Control of Administrative Action 329 (1965).
6. Id. at 424.
7. Montgomery v. Rumsfeld, 572 F.2d 250, 253-54 (9th Cir. 1978).
8. Davis, Administrative Law Treatise (2d ed.) ` 26.12, at 468-69.
9. Wheaton v. Peters, 33 U.S. (8 Pet.) 591, 657-58 (1834). Swift v. Tyson, 41 U.S. (16 Pet.) 1 (1842), was no exception to this approach. The Court there applied not federal common law but "general" common law, which, at least in theory, derived from custom and was "in a great measure, not the law of a single country only, but of the commercial world." Id. at 19.
10. Attorney General's Manual on the Administrative Procedure Act 101 & 93 (1947). See also id. at 93 (declaring that the Act "leaves the mechanics of judicial review to be governed by other statutes and by judicial rules").
11. See AG's Manual at 93 n.1, 96 & 108 (all relying mainly on statements and letters by the Attorney General to support the view that the APA's judicial review provisions are a mere restatement of prior law).
12. Richard A. Posner, The Rise and Fall of Administrative Law, 72 Chi.-Kent L. Rev. 953, 955 n.1 (1997).
13. Daniel B. Rodriguez, Jaffe's Law: An Essay on the Intellectual Underpinnings of Modern Administrative Law Theory, 72 Chi.-Kent L. Rev. 1159, 1160 (1997).
14. See SEC v. Cogan, 201 F.2d 78, 87 (9th Cir. 1951).
15. See, e.g., Velasquez-Tabir v. INS, 127 F.3d 456, 458 n.9 (5th Cir. 1997); DuBois v. USDA, 102 F.3d 1273, 1284 (1st Cir. 1996); Smith v. Office of Civilian Health & Med. Program of the Uniformed Serv., 97 F.3d 950, 955 (7th Cir. 1996); Stupak-Thall v. United States 70 F.3d 881, 886 (6th Cir. 1995) (all citing Section 706 as requiring de novo review on issues of law).
16. APA Legislative History at 370.
17. Antonin Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke L. J. 511, 4.
18. North American Telecommunications Ass'n v. FCC, 772 F.2d 1282, 1292 (7th Cir. 1985).