WEBGALLEYS.DOC 05/05/99 5:38 PM



    Peter H. Schuck*

I know David Schoenbrod, and he is no Owen Fiss. Yet, if the United States Supreme Court were to resurrect the nondelegation doctrine as Professor Schoenbrod proposes, it would radically in-crease judicial power over vast areas of American life at the ex-pense of the “political” branches (as we quaintly call them). Fiss, my beloved but occasionally misguided colleague, would probably applaud this change, but Professor Schoenbrod, usually so sensible about such things, would surely deplore it. Fortunately, the Court is most unlikely to adopt Professor Schoenbrod’s perverse pro-posal. Besides, even if the Court did revive the nondelegation doc-trine, it would surely want to neuter it. Like Buddy, our First Dog, the Court lacks the balls to do what the doctrine would require. So much the better for the Court—and for the rest of us (though not for poor Buddy).

In explaining why this is so, I have—we all have—the advan-tage of Jerry Mashaw’s recent and excellent book, Greed, Chaos, and Governance.1 There, Mashaw reviews the arguments for and against delegation using the public choice literature and Professor Schoenbrod’s book2 as analytical foils. I do not accept all of Mashaw’s claims about delegation. He argues, for example, that voters can more readily discern and police a legislator’s prefer-ences through statutory standards like “protect the public health” and “fair and reasonable” than they can through statutory lan-guage that prescribes more specific tradeoffs of competing values. This claim is quite implausible as a general matter, although it is surely correct in some subset of cases depending on the particular statutes being compared. Additionally, his statement that his point has not been “to decide the nondelegation doctrine issue

conclu-* Simeon E. Baldwin Professor, Yale Law School; John Marshall Harlan Visiting Pro-fessor, New York Law School (Spring 1999).




WEBGALLEYS.DOC 05/05/99 5:38 PM

776 CARDOZO LAW REVIEW [Vol. 20:775

sively one way or the other”3 strikes me as disingenuous, for his analysis plunges a long, sharp knife deep into the doctrine’s heart, leaving it near death’s door.

Nevertheless, I accept almost all of Mashaw’s arguments. Our agreement, I presume, is not due to something in the New Haven water supply. Instead, it reflects administrative law scholars’ fa-miliarity with a wide variety of regulatory schemes and public ad-ministration arrangements. It also reflects our common under-standing that the ubiquity of broad delegations denotes much more than the undoubted desires of politicians to eat their cake and have it too. In my view, delegation—when backed (as it is in our system) by many powerful institutional and informal controls over agency discretion—constitutes one of the most salutary de-velopments in the long struggle to instantiate the often competing values of democratic participation, political accountability, legal regularity, and administrative effectiveness.

I wish to make some arguments against a robust nondelega-tion doctrine (and if it is not robust, there is no point talking about it) that Mashaw does not make.4 In the spirit of Dean Michael Herz’s observation that most broad delegations satisfy the formal requirements of Article I legislation and that the merits of a non-delegation doctrine must therefore turn on functional considera-tions,5 my arguments in favor of broad delegations in many cir-cumstances are functional in nature. I shall organize my arguments around four questions: What is the nature of the dele-gation problem?; What should be our goals in seeking to control delegation?; In the absence of a nondelegation doctrine, is agency lawmaking effectively constrained?; What would be the conse-quences of reviving the nondelegation doctrine?

My answers to these questions can be briefly stated. First, al-though it is always difficult and costly for a democratic citizenry to monitor, control, guide, and correct the conduct of its governmen-tal agents,6 this problem has not yet reached the level of serious political dysfunction nor is it the kind of problem for which courts (or even scholars, who are less constrained than judges) can devise an effective doctrinal solution in terms of the desired specificity of

3 MASHAW, supra note 1, at 12.

4 I think, however, that Mashaw would agree with them.
5 Michael Herz, Remarks at the Cardozo Law Review symposium “The Phoenix Rises Again: The Nondelegation Doctrine from Constitutional and Policy Perspectives” (Mar. 19, 1998) (transcript on file with the Cardozo Law Review).


WEBGALLEYS.DOC 05/05/99 5:38 PM


statutes. Second, a coherent nondelegation doctrine would not limit itself to the single, simple goal of “responsibility” that is the lodestar of Professor Schoenbrod’s analysis; it would also include other goals that exist in profound tension with that of responsibil-ity. Third, agencies are highly constrained to comply with legisla-tive intent as they understand it, and they are further impelled by other powerful forces to understand this legislative intent in much the way that Congress does.7 In truth, the freedom of agencies “to do as they please”8 is among the least of their problems—and of ours. Finally, the consequences of a robust nondelegation doctrine would be so pernicious that the Court will either never adopt it or will render it toothless.9


I think that Professor Schoenbrod has the problem wrong. The real problem with delegation is not a lack of political “respon-sibility,” a concept that he deploys frequently but never defines.10 I understand responsibility to be the accountability of elected offi-cials (and the indirect accountability of their appointed agents) to the electorate for significant policy choices. If anything, our politi-cal system produces too much of this kind of responsibility. Our system creates incentives for legislators (especially those with rela-tively short terms of office, like members of Congress) to think so obsessively about their immediate electoral prospects that they are unduly timorous, lacking the leeway that a more Burkean concep-tion of representation requires. Hence, they may neglect longer term social problems whose solutions require immediate sacrifices for delayed gains, problems that demand as much of the legisla-tors’ attention, prudence, and political courage as they can mus-ter.11

Whether we have struck the best balance between account-ability and stewardship that can be achieved in light of the realistic

7 The notions of congressional understanding and legislative intent are merely meta-phors, of course, and somewhat misleading anthropomorphizing metaphors at that; they are images of cognition and purposiveness, necessary fictions for ascribing relatively sim-ple, discrete goals to exceedingly complex institutions.

8 David Schoenbrod, Delegation and Democracy: A Reply to My Critics, 20 CARDOZO L. REV. 731, 762 (1999).

9 Such a doctrine moreover would also contradict Professor Schoenbrod’s generally sound jurisprudential and political values. He should therefore reject it.

10 See infra Part II.
11 See, e.g., DAVID R. MAYHEW, CONGRESS: THE ELECTORAL CONNECTION (1975). Reform of the Social Security system and global warming are current examples.

WEBGALLEYS.DOC 05/05/99 5:38 PM

778 CARDOZO LAW REVIEW [Vol. 20:775

constraints on democratic governance is a vital question to which no one really knows the answer. But, I feel quite certain that non-accountability in Professor Schoenbrod’s sense is, relatively speaking, a non-problem. The greater dilemma, I think, is that growing social complexity has made it far more difficult for legisla-tors (not to mention voters) to accurately predict the consequences of their choices so that they can reason their way to a conclusion as to the best policy choice. If I am right about this, we may need more delegation to agencies, not less.

I begin with the proposition that in comparative terms, the United States is a well-governed society. I say well-governed, not perfect—and I am comparing it to other large, diverse, and dy-namic societies faced with the characteristic problems of post-industrialism, not to Shangri-La or even to a relatively homogene-ous, corruption-free country like Denmark. I recently had occa-sion to adduce some empirical evidence to support the claim that the United States is much better governed than it was in 1965 in the sense that its political processes and policy outcomes are now much more democratic, just, and social welfare enhancing.12 In the two years since I presented this evidence, the comparative dimen-sion of my claim has also grown stronger. The United States has turned in an extraordinary economic performance (growth in pro-duction and wage levels with low inflation and unemployment), while responding to public demands for governmental reform and for fiscal prudence (a budget surplus) and boasting improved so-cial indicators in many policy areas such as crime, health, educa-tion, environment, standard of living, poverty, and minority group progress. These achievements are in sharp contrast to the con-tinuing process of Eurosclerosis13 and the recent freefall of almost all of the Asian economies, with grave threats to their nascent, of-ten fragile political democracies. Such developments belie the im-age of a dysfunctional lawmaking process that Professor Schoen-brod and many other commentators draw, as well as the persuasive substantive policy failures that public choice pessimists like McNollgast14 and the late Mancur Olson predict.15

12 See Peter H. Schuck, Against (And For) Madison: An Essay in Praise of Factions, 15 YALE L. & POL’Y REV. 553 (1997).

13 Eurosclerosis is a term that was coined in the early 1980s to describe the period of sluggishness in the development of the European Union when all issues, however minor, languished until all Member States could agree. See The Business of Europe, THE ECONOMIST, Dec. 7, 1991, at 63. It now applies more broadly to Europe’s slow growth, high unemployment, and policy rigidity.

14 See Mashaw, supra note 1, at 140.

WEBGALLEYS.DOC 05/05/99 5:38 PM


I can imagine Professor Schoenbrod responding in several ways. First, as he writes in his Article, “[l]arge majorities tell poll-sters that government has somehow eluded their control.”16 I do not doubt this claim, although much of its significance turns on how the pollsters phrased their question and how respondents un-derstood it. The more important point, however, is that Ameri-cans have probably always said this—and that what they say is in a sense true, at least at the level of the individual. What is less clear is that such polling data tell us anything about the desirability of the nondelegation doctrine.

Second, he might argue that if the nondelegation doctrine constrained lawmakers more, the policy processes and outcomes would be even better. Perhaps he is right about this, but I seri-ously doubt it. At the very least, the post-1965 progress of Ameri-can society increases the already heavy burden of proof that Mashaw’s theoretical analysis imposes on Schoenbrod and the public choice pessimists.

Finally, Professor Schoenbrod might also insist that by dis-cussing policy outcomes, I am addressing the wrong question. The right question to pose about delegation, he might say, is not about its social consequences but about its effects on democratic legiti-macy. Here, however, I stand firmly with Dan Kahan,17 who main-tains, along with Mashaw, that these two criteria cannot be sepa-rated, that democratic legitimacy is a function of effective governance, desirable policy outcomes, and other political values.18


This leads me to my second point. Professor Schoenbrod has identified the wrong goal, or at least has fastened on one goal to the exclusion of other, equally attractive ones. His sole desidera-tum, it would appear, is what he calls “responsibility.” He uses the word no fewer than five times in his one-page introduction, and he mentions no other goal.19 As noted above, I am not certain what he means by this, although he does associate it with making “hard

15 Olson might object to being characterized as a pessimist. In a letter written to me six months before his death, Olson said “I don’t agree that subsequent experience has gone contrary to any ‘predictions’ in The Rise and Decline of Nations.” Olson also ex-pressed optimism about the role of ideas in improving social outcomes.

16 Schoenbrod, supra note 8, at 731.
17 Professor of Law at the University of Chicago Law School. See Dan Kahan, Democ-racy Schmemocracy, 20 CARDOZO L. REV. 795 (1999).

18 See id.

19 See Schoenbrod, supra note 8, at 735.

WEBGALLEYS.DOC 05/05/99 5:38 PM

780 CARDOZO LAW REVIEW [Vol. 20:775

choices.”20 The closest he comes to defining responsibility is to ob-serve that it is “more meaningfully accountable for what [the ad-ministration] does,”21 in contrast to “responsiveness,” which “will give voters what they want.”22 I confess that I thought that giving the voters what they want, and what they thought they voted for, is precisely what democracy is supposed to be about and what advo-cates of the nondelegation doctrine hope and suppose it will achieve. If that is not Professor Schoenbrod’s objective, then I do not know what is.

In any event, political responsibility is no more a self-defining term than is democracy.23 More to the point, responsibility is only one value among others. Let me suggest some additional constitu-tional and quasi-constitutional goals that any democratic, just, and effective lawmaking system should seek to both reify and advance. Lawmaking should encourage active, meaningful participation by individual citizens and groups affected by the law. It should facili-tate and reflect mature deliberation among members of the public and among the lawmakers themselves. Lawmaking should exhibit instrumental competence, in the sense that it implements a satis-factory level of legislative purposes. Lawmaking, as Mashaw notes, should promote justice in individual cases, not merely at wholesale. It should also achieve responsiveness to public prefer-ences, in Professor Schoenbrod’s sense of giving the voters what they (think they) want. Finally, of course, lawmaking in both its procedural and substantive aspects should exemplify and secure the rule of law.

The political responsibility that Professor Schoenbrod wants to achieve through more specific statutes must coexist with these goals and will sometimes conflict with them. Even if the nondele-gation doctrine would in fact promote political responsibility, which I very much doubt, it would also frustrate some or all of these other values. Mashaw explains, for example, how more spe-cific statutes can undercut both justice in the individual case and responsiveness to diverse local conditions.24

Professor Schoenbrod seems innocent of, or at least unim-pressed by, these poignant and inescapable normative and empiri-cal tradeoffs. He assumes that the legislature is the site where the

20 See id. at 740.

21 Id. at 750.

22 Id.
23 Professor Kahan emphasizes this point. See Kahan, supra note 17, at 795.
24 See MASHAW, supra note 1, at 50-80.

WEBGALLEYS.DOC 05/05/99 5:38 PM


virtues of responsible lawmaking are best achieved; it is there, he suggests, that the public’s values should be expressed and the hard policy choices made. He fails to see, however, that the particular attributes of the legislature’s delegation—its breadth, type, and level—are themselves fundamental policy choices. Moreover, these issues are hardly peripheral to legislative choice. Along with the closely related issue of the scope of the agency’s regulatory authority, they are almost always—and quite explicitly—at the heart of the political debates in Congress over the shape and con-tent of particular pieces of legislation. The optimal specificity and other delegation-related features of the legislation are among the questions on which almost all of the parties to these legislative struggles—congressional committees, legislative staffs, the White House, regulated firms, “public interest” groups, state and local governments, and others—tend to stake out clear positions, for they know the resolution of these questions may well determine the nature and effectiveness of the regulatory scheme being estab-lished.25 The issue of statutory specificity is not resolved sub silen-tio or by default, as Professor Schoenbrod suggests. Rather, it is a focal point of the political maneuvering in the legislature.

Legislation is only part of the process of responsible lawmak-ing, and it is becoming a less important part. In some important respects, this is for the better. Today, the administrative agency is often the site where public participation in lawmaking is most ac-cessible, most meaningful, and most effective.

The administrative agency is often the most accessible site for public participation because the costs of participating in the rule-making and more informal agency processes, where many of the most important policy choices are in fact made, are likely to be lower than the costs of lobbying or otherwise seeking to influence Congress. Moreover, the institutional culture of the administrative agency, despite its often daunting opacity, is probably more famil-iar to the average citizen, who deals with bureaucracies constantly and probably works in one, than the exotic, intricate, unruly (and “un-ruley”), insider’s culture of Congress.

The agency is often a more meaningful site for public partici-pation than Congress, because the policy stakes for individuals and interest groups are most immediate, transparent, and well-defined at the agency level. One can scarcely exaggerate the importance of this consideration to the legitimacy of democratic politics and to

25 See, e.g., Terry Moe, The Politics of Bureaucratic Structure, in J. CHUBB ET AL., CAN THE GOVERNMENT GOVERN? (1989).

WEBGALLEYS.DOC 05/05/99 5:38 PM

782 CARDOZO LAW REVIEW [Vol. 20:775

the substantive content of public policy. After all, it is only at the agency level that the generalities of legislation are broken down and concretized into discrete, specific issues with which affected parties can hope to deal. It is there that the agency commits itself to a particular course of action; because only there does it propose the specific rate it will set, the particular emission level it will pre-scribe, the precise restrictions on private activity it will impose, the exact regulatory definitions it will employ, the kinds of enforce-ment techniques it will use, the types of information it will collect, and the details relating to the administrative state’s myriad other impacts on citizens and groups. In short, it is only at the agency level that the citizen can know precisely what the statute means to her; how, when, and to what extent it will affect her interests; whether she supports, opposes, or wants changes in what the agency is proposing; whether it is worth her while to participate ac-tively in seeking to influence this particular exercise of governmen-tal power, and if so, how best to go about it; and where other citi-zens or groups stand on these questions. God and the devil are in the details of policymaking, as they are in most other important things—and the details are to be found at the agency level. This would remain true, moreover, even if the nondelegation doctrine were revived and statutes were written with somewhat greater specificity, for many of the most significant impacts on members of the public would still be indeterminate until the agency grappled with and defined them.

Finally, the agency is often the site in which public participa-tion is most effective. This is not only because the details of the regulatory impacts are hammered out there. It is also because the agency is where the public can best educate the government about the true nature of the problem that Congress has tried to address. Only the interested parties, reacting to specific agency proposals for rules or other actions, possess (or have the incentives to ac-quire) the information necessary to identify, explicate, quantify, and evaluate the real-world consequences of these and alternative proposals. Even when Congress can identify the first-order effects of the laws that it enacts, these direct impacts seldom exhaust the laws’ policy consequences. Indeed, first-order effects of policies usually are less significant than the aggregate of more remote ef-fects that ripple through a complex, interrelated, opaque society. When policies fail, it is usually not because the congressional pur-pose was misunderstood. More commonly, they fail because Con-gress did not fully appreciate how the details of policy implemen-

WEBGALLEYS.DOC 05/05/99 5:38 PM


tation would confound its purpose. Often, however, this knowl-edge can only be gained through active public participation in the policymaking process at the agency level where these implementa-tion issues are most clearly focused and the stakes in their correct resolution are highest.

If the rule of law is a central goal (as Professor Schoenbrod clearly thinks, although usually calling it “responsibility”), then our
desideratum should not be statutes of a certain specificity with that level of specificity enforced by courts as a matter of constitu-tional law. Instead, we should seek to assure that bureaucratic power is checked and effectively bent to the legislative purpose, that the agency, in Professor Schoenbrod’s words, is not “free to do as it pleases.”26 Now, no one would suggest that this essential goal of constraining and guiding bureaucratic power is easily ac-complished. If it were, Congress could take longer recesses and most administrative lawyers would be out of a job. Developing agency cost theory and other approaches have painstakingly de-tailed the tensions between this goal and other important goals such as technical rationality, policy flexibility, procedural simplic-ity, speed of decision, justice in the individual case, individual dig-nity, and the like.27

Federal agencies, however, are hardly at liberty. They are surrounded by watchdogs with sharp, penetrating teeth. Indeed, what most clearly distinguishes the American administrative state from that of other countries is the pervasive public philosophy of mistrust of government bureaucracies and the subordination of bureaucracy to numerous, diverse, external, power-checking insti-tutions and processes. These institutions, moreover, are remarka-bly powerful; they routinely shape policy and delve into the intri-cate details of administration. Their broad array of inducements, both positive and negative, enable them to guide and often deter-mine the agency’s exercise of discretion. In this fundamental sense, the structural preconditions for democratic delegation are satisfied: the legislature is delegating power to a branch whose de-cisions the legislature and its other agents—for example, the courts

26 Schoenbrod, supra note 8, at 762.
27 See, e.g., JERRY L. MASHAW, BUREAUCRATIC JUSTICE: MANAGING SOCIAL SECURITY DISABILITY CLAIMS (1983); Peter H. Schuck, When the Exception Becomes the Rule: Regulatory Equity and the Exceptions Process in Federal Energy Policymaking, 1984 DUKE L.J. 163.

WEBGALLEYS.DOC 05/05/99 5:38 PM

784 CARDOZO LAW REVIEW [Vol. 20:775

and interest groups—can effectively influence, if not wholly con-trol.28 I do not claim that this control is complete, nor should it be if the advantages of technocratic administration are to be realized. Agencies enjoy some leeway and sometimes abuse it. The con-trols, however, are extensive.

Some of these external constraints on bureaucratic policy-making are: (1) Congress; (2) the Executive Office of the Presi-dent; (3) judicial review; (4) interest group monitors; (5) media; and (6) informal agency norms. It is important, moreover, to re-member that these and other constraints on bureaucracy’s free-dom to “do as it pleases” all operate simultaneously. My consid-eration of them can be brief, as a vast political science literature exists on each.

    A. Congress

Congress possesses numerous formal and informal controls over agency discretion. I shall mention six of them: statutory con-trols; legislative history; oversight; the appropriations process; statutory review of agency rules; and confirmation of key person-nel.

   1. Statutory Controls
Congress, of course, writes the statutes that confer and govern agency authority. In doing so, it prescribes the substantive content of that authority, the structure of agency decisionmaking, the pro-cedures through which it occurs, and the informational, budgetary, and other controls to which the agency will be subjected. As Terry Moe,
29 McNollgast,30 and many other political scientists have shown, Congress uses these controls to shape the administrative

28 The desired level of delegation may vary according to a number of political factors, including the degree to which Congress and the executive branch are controlled by the same political party and the degree of party unity that exists. This is one reason why many New Deal statutes, enacted during a period of unified government, delegated power to the agencies in very broad terms, while many of those enacted during the early 1970s, when government was divided, contained many more controls over agency discretion. See DAVID R. MAYHEW, DIVIDED WE GOVERN: PARTY CONTROL, LAWMAKING, AND INVESTIGATIONS (1990); JAMES SUNDQUIST, THE DECLINE AND RESURGENCE OF CONGRESS (1981); see also Peter Strauss, Comments at the Cardozo Law Review sympo-sium “The Phoenix Rises Again: The Nondelegation Doctrine from Constitutional and Policy Perspectives” (Mar. 19, 1998) (transcript on file with the Cardozo Law Review).

29 Senior Fellow at the Hoover Institution and Professor of Political Science at Stan-ford University.

30 McNollgast is an anagram of Matthew McCubbins, Roger Noll, and Barry Weingast. The authors coined this term for their own joint work, and it has been adopted by others.

WEBGALLEYS.DOC 05/05/99 5:38 PM


process in ways that serve its electoral and policy interests and make it difficult for agencies to threaten those interests.

   2. Legislative History
Even when Congress enacts vague statutes, it often clarifies their meaning through legislative history. Agencies understand that the committee reports and floor debates serve as important controls on agency discretion. Justice Scalia frequently reminds us, of course, that legislative history can be indeterminate and is often used strategically by members who cannot muster the votes to get their preferences inscribed in the statutory language itself.
31 As Judge Harold Levanthal famously put it, using legislative his-tory is like looking out over a crowd in order to find one’s friends.32 Despite these abuses, or rather because of them, legislative history can dictate which policies agencies may and may not adopt.

    3. Oversight
Congressional oversight of administration is one of the central pillars of the constitutional schemes of checks and balances—or in Richard Neustadt’s phrase, “separate institutions sharing pow-ers.”
33 While the nature, quality, and intensity of legislative over-sight vary from committee to committee, it is often used to signal congressional preferences on agency policy issues and to extract policy commitments from agency officials. Agencies fear intrusive oversight and their decisions and behavior often reflect what po-litical scientists refer to as “anticipatory reaction” to those con-trols.

   4. Appropriations

The appropriations process sharply constrains the authority and discretion of agencies. These constraints are imposed through the language of the funding legislation, through formal committee and subcommittee oversight hearings, and through the frequent in-formal interactions between members and agency officials. De-spite congressional rules against including substantive legislation in

31 See, e.g., Morales v. Trans World Airlines, Inc., 504 U.S. 374, 383 (1992); United States v. Nordic Village, Inc., 503 U.S. 30, 37 (1992); see also Fort Stewart Schs. v. Federal Labor Relations Auth., 495 U.S. 641, 649 (1990) (“There is no conceivable persuasive ef-fect in legislative history that may reflect nothing more than the speakers’ incomplete un-derstanding of the world upon which the statute will operate.”).

32 See Conroy v. Aniskoff, 507 U.S. 511, 519 (1993) (quoting Judge Leventhal).

WEBGALLEYS.DOC 05/05/99 5:38 PM

786 CARDOZO LAW REVIEW [Vol. 20:775

appropriations bills, it is nonetheless a common practice for ap-propriations committees to engage in this practice and for agencies to acquiesce and abjectly obey. Indeed, substantive controls on agency policymaking are often included even in the increasingly common omnibus budgetary reconciliation legislation.

   5. Statutory Review

In addition to the constitutional power to override presiden-tial vetoes, Congress has over the years enacted a grab-bag of pro-visions requiring some agency rules to run the gauntlet of various forms of legislative veto before they could become effective. The Supreme Court decision in INS v. Chadha34 invalidated certain forms of legislative veto but left most others, including report- and-wait provisions, in place. In 1996, Congress extended this form of control to all rules of all federal agencies, as defined in the Ad-ministrative Procedure Act (“APA”),35 although the legislation subjects “major rules” to a more intensive review than other rules.

Professor Schoenbrod makes much of the fact that this Con-gressional Review Act,36 while comprehensive, has not led to any votes of disapproval by Congress; this inaction shows, he says in a vivid simile, that “legislators react to responsibility as vampires do to garlic—they flee.”37 But it shows no such thing. Numerous em-pirical studies of the operation of the various kinds of legislative vetoes, both before and after Chadha, have found that they gener-ate strong anticipatory reactions by agency officials, often includ-ing intensive discussions between those officials and committee members or staff that cause the agency to conform their rules to the wishes of the committee.38 Indeed, recognition of this powerful informal process of congressional review of agency policymaking outside the statutory procedures of the APA has long caused many consumer, environmental, and other “public interest” groups to oppose legislative veto provisions, including the Congressional Review Act, on the ground that it enables industry interests to use congressional staff to gain an extra bite at the regulatory apple.

34 462 U.S. 919 (1983).

35 5 U.S.C. 551-559 (1994).

36 Id. 801-808.

37 Schoenbrod, supra note 8, at 739.

WEBGALLEYS.DOC 05/05/99 5:38 PM


    6. Confirmation
Finally, Congress frequently uses its constitutional power to confirm or reject presidential nominees in order to shape agency policies. Again, much of this influence operates through the me-dium of anticipatory reaction, here on the part of the President who must consider congressional policy and personal preferences when determining whose name to send to Capitol Hill. But con-gressional influence also often operates through the confirmation hearings when members extract from the nominee explicit com-mitments to adopt or avoid certain policies. These policy com-mitments can be quite specific; indeed, they may be even more specific than the statutory commitments that Professor Schoen-brod would require of Congress under a revitalized nondelegation doctrine.

  B. Executive Office of the President
Even before agency policies and rules see the light of day, the significant ones are vetted with the Office of Management and Budget (“OMB”)—and, in a few highly controversial and delicate cases (e.g., FDA regulation of tobacco), with the President and his closest political and policy aides. This may take the form of budg-etary review, if the agency action would entail significant fiscal im-pacts; legislative review, if it requires statutory change; or regula-tory review, if it meets the OMB criteria. Once again, the greatest impact of these processes is not so much the reviews themselves as the anticipatory reactions of the agency officials and the reviewers. They are keenly mindful of the policy concerns of the relevant congressional committees and key members, who often are con-sulted informally as part of these processes and usually receive drafts of the review documents before they are released to the public. Although the review process works somewhat differently with respect to the independent regulatory agencies, their vaunted (but to some extent, illusory) independence is designed to make them even more dependent on Congress and responsive to its pol-icy priorities.

   C. Judicial Review
In exercising their review of agency “actions” (not just rules) under the APA, the courts attempt to discipline the agencies by requiring that their actions conform to congressional intent and, where they enjoy a delegated discretion, that the discretion is not abused. Although I and other scholars have questioned the effec-

WEBGALLEYS.DOC 05/05/99 5:38 PM

788 CARDOZO LAW REVIEW [Vol. 20:775

tiveness and value of this review in many cases, none of us doubts that agencies invariably fear it, that they seek to tailor their actions (at least those that are reviewable) in anticipation of it, and that the courts do attempt to police agency departures from congres-sional policy choices, which must often be inferred from general statutory language and other interpretive materials. Although it is true that the Chevron39 doctrine in principle increases the defer-ence to agency interpretations of their governing statutes, many commentators have observed that the courts can, and often do, manipulate this doctrine in order to preserve much of their influ-ence over agency decisions, including enforcing agency fidelity to congressional intent.

Toward the end of his book, Professor Schoenbrod quotes then-professor Scalia to the effect that the nondelegation doctrine amounts to a judicial self-denying ordinance. Justice Scalia, of course, was writing before the Supreme Court in Chevron curbed judicial authority to interpret statutes without regard to the agen-cies’ own interpretations. In any event, we certainly do not need a nondelegation doctrine to enable Congress to protect its legislative prerogatives from judicial incursion, which was Justice Scalia’s concern. As my colleague Bill Eskridge has carefully documented, Congress possesses ample power to overrule judicial rulings and exercises that power frequently.40 The fact that Congress can do so, however, does not mean that the nondelegation would there-fore not create mischief.41 It would indeed, as I discuss in the final section.

It is noteworthy that the Congressional Review Act expressly provides that the courts may not review any congressional action involved in Congress’s review of agency rules,42 and further pro-vides that Congress’s failure to disapprove an agency rule shall not raise any inference concerning “any intent of the Congress.”43 These provisions are the clearest indication, if any were needed, of Congress’s determination to maintain its close control over agency rulemaking as against both the courts and the Executive Branch. Indeed, these instruments of congressional control create a

para-39 Chevron v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984).
40 See William N. Eskridge, Jr., Overriding Supreme Court Statutory Interpretation De-cisions, 101 YALE L.J. 331 (1991).

41 Among other things, the politics of restoring a judicially transformed statutory meaning will differ from the polities that produced the original enactment, and may pre-vent such restoration. See infra.

42 See 5 U.S.C. 805 (1994).

43 Id. 801(g).

WEBGALLEYS.DOC 05/05/99 5:38 PM


doxical situation: By preserving and extending its oversight and veto power over agency decisions, Congress is more willing to delegate power broadly to the agencies, knowing that it can always retrieve and discipline that power if need be.

   D. Interest Group Monitors

Agencies’ freedom to “do as they please” is further con-strained by the close surveillance of their actions by the constella-tion of interest groups that invariably cluster around them. As students of Congress have pointed out, these groups police the agencies by signaling Congress in a variety of ways whenever the groups think that the agencies are deviating from the policy path that they prefer and that they believe Congress has chosen, re-gardless of how general or specific the statutory language may be. These groups, of course, further shape agency policies through their participation in formal and informal agency decisionmaking and through their ability to mobilize opposition to these policies. To be sure, these efforts sometimes seek to push the agency to ex-ercise its discretion in a way not intended or permitted by Con-gress, a circumstance that underlies all concerns about broad dele-gations. The burden of my comments here, however, is to show that their power to accomplish this end is a highly constrained one.

    E. Media

Every agency is surrounded by organs of public communica-tion that focus on its decisions. Some of these media report to a mass audience while others serve highly specialized audiences, in-cluding regulated firms, trade associations, consultants, “public in-terest” groups, Congress, administrative lawyers, and the like. In addition to using conventional journalistic methods of investiga-tion, these media enjoy a legal right of access to agency proceed-ings under sunshine laws, the Federal Advisory Committee Act,44 the Freedom of Information Act,45 and other such laws. In effect, the media help to police the activities of agencies in ways that en-able both interest groups and Congress to keep the agencies in line.

Agencies, then, know that their actions are likely to come un-der public scrutiny and must tailor their conduct and decisions ac-cordingly. This is not to deny, of course, that agencies would often

44 Id. app. 1-15 (1994).

45 Id. 552.

WEBGALLEYS.DOC 05/05/99 5:38 PM

790 CARDOZO LAW REVIEW [Vol. 20:775

prefer to conduct their business in secret and that they sometimes succeed in doing so. It is to say, rather, that intensive media cov-erage of agency actions makes it almost impossible to shield from public notice the important decisions that deviate significantly from congressional intent.

   F. Informal Agency Norms
Finally, agencies are subject to internal limits on their ability to “do as they please.” Or to put the point more precisely, they ordinarily are pleased to do what they think Congress has re-quired—and not only because of the legal and political constraints discussed above. In most if not all agencies, there exists an organ-izational culture committed to the rule of law. Indeed, as political scientist James Q. Wilson46 has observed, most of the classic com-plaints about public bureaucracies are really criticisms of agencies for being too legalistic (too rigid, unimaginative, process-oriented, etc.) in their strict adherence to the statute, at least as they under-stand it. In addition, management controls are ordinarily designed to reinforce these tendencies. Moreover, agencies are increasingly staffed by individuals who are professionally trained and whose professional norms, which include a commitment to the rule of law, often contradict or transcend the narrow political or bureau-cratic interests that might otherwise lead decisionmakers astray.


The consequences of a robust nondelegation doctrine would not be pretty. First, it would greatly strengthen the power of the federal courts relative to that of Congress and the agencies. Ironi-cally, it would do so in the name of the separation of powers while undermining that very principle. Such a massive shift of power to unelected federal judges should trouble us deeply.47 It is particu-larly obnoxious when, as in this case, it is not essential to the

vindi-46 James Q. Wilson is a professor of management and public policy emeritus at Uni-versity of California at Los Angeles.

47 In this sense, as Martin Redish points out, an invigorated nondelegation doctrine would be tantamount to a return to substantive due process, another essentially standard-less judicial doctrine that reached its zenith at roughly the same time. Redish notes that in other respects, the two doctrines are quite dissimilar, if not opposite. While the nondele-gation doctrine is concerned with the form of legislation, substantive due process focuses on its content. And while the nondelegation doctrine demands clearer legislative policy choices, substantive due process denies to legislatures the power to make certain choices at all. See Martin Redish, Comments at the Phoenix Rises Again: The Nondelegation Doctrine from Constitutional and Policy Perspectives (Mar. 19, 1998) (transcript on file with the Cardozo Law Review).

WEBGALLEYS.DOC 05/05/99 5:38 PM


cation of enumerated constitutional rights, and when the courts would be guided by Professor Schoenbrod’s principles, which are as unhelpful as “Congress need only state the law”48 and as slip-pery, manipulable, and epistemologically incoherent as the distinc-tion between “lawmaking and law interpretation.”49 How general is too general, how specific is specific enough—these are, contrary to Professor Schoenbrod’s claim, questions of degree, not kind. They are preeminently questions of politics and of policy that courts are poorly equipped to answer as a functional matter and that they are disabled from resolving as a constitutional matter. Resolution of these questions, moreover, depends entirely on con-text—or as Professor Schoenbrod recognizes, on a “number of other factors.”50

To inscribe the nondelegation doctrine in the Constitution, ei-ther textually or interpretively, would be little better than having the courts make it up as they go along.51 This is evident once one tries to imagine how one would draft such an amendment. The line-drawing problems are simply insuperable, which is why the Supreme Court—whether liberal or conservative, textualist or in-terpretivist—has resisted any robust nondelegation doctrine dur-ing more than two centuries of public law adjudication, and espe-cially during the age of statues spanning the last sixty-five years.

It is true, as I noted earlier, that Congress could overturn a court decision invalidating a statute under the nondelegation doc-trine by enacting another version of the statute with the requisite degree of specificity. Ironically, of course, the Supreme Court would probably avoid specifying precisely what degree of specific-ity would pass constitutional muster under a revived nondelega-tion doctrine, both because it would not know the answer—never having had to grapple with the political and policy problems that animated Congress to draft the statute as it did—and because stra-tegically it would not want to commit itself in advance. Vagueness, it turns out, has its legitimate uses for courts as well as for legisla-tures. Indeed, the most charitable way to view the doctrine is as an effort to encourage Congress—through a judicially enforced and orchestrated dialogue—to take a second, harder look at the dele-gation issue. In this view, the nondelegation doctrine would

func-48 Schoenbrod, supra note 8, at 752.

49 Id. at 764.

50 Id. at 755.
51 Much the same can be said for the balanced budget amendment, but that is another subject.

WEBGALLEYS.DOC 05/05/99 5:38 PM

792 CARDOZO LAW REVIEW [Vol. 20:775

tion as a kind of “clear statement” doctrine analogous to the provi-sion of the Canadian Charter of Rights and Responsibilities, which permits Parliament to override certain judicial rulings invalidating statutes on constitutional grounds once the ruling has forced Par-liament to confront the constitutional objection squarely and ex-plicitly.52

Such a second look requirement certainly has its attractions, but its apparent benignity can be misleading. It means that the court in effect is forcing Congress to bear the burden of inertia, which is always a crucial, and often determinative, factor in legisla-tive politics. It is not enough to say that if Congress enacted the statute once, it can readily do so again—now free of the constitu-tional taint of inadequate specificity. The political situation may have changed, and it may be impossible to reassemble the winning coalition. Professor Schoenbrod might reply that this is precisely the point; if Congress could not enact the more specific version of the statute, it should not be permitted to enact the less specific one. This argument, however, entirely begs the question before us—whether the nondelegation doctrine is justifiable on other grounds in the first place. To say that it is more justifiable because it can always be overridden by Congress is not simply wrong em-pirically (the political burden of inertia will have shifted, perhaps decisively, to the supporters of the statute), but is also bootstrap-ping.

By magnifying the already great uncertainty surrounding the legislative process, the doctrine would further increase the strate-gic opportunities of politicians and organized interests that hope to shape, derail, or delay new legislation. The risks that courts would invalidate statutes would be vastly greater yet still utterly unpre-dictable. From the more detached, public interest perspective of one concerned solely with the legitimacy and effectiveness of leg-islative politics, much would certainly be lost. It is hard to see what, if anything, would be gained.

In the end, then, the nondelegation doctrine is a prescription for judicial supervision of both the substance and forms of legisla-tion and hence of politics and public policy, without the existence

52 See Canadian Charter of Rights and Freedoms 15 (1994). In a similar spirit, Beth Garrett notes that the rules permitting members to make points of order in the congres-sional budgetary process enable opponents of majoritarian legislation to encourage the discussion of under-enforced norms such as federalism. She notes further that these rules also give minorities the power to stop the legislation in its tracks for any reason, whether principled or not. See Elizabeth Garrett, Accountability and Restraint: The Federal Budget Process and the Line Item Veto Act, 20 CARDOZO L. REV. 871 (1999).

WEBGALLEYS.DOC 05/05/99 5:38 PM


or even the possibility of any coherent, principled, or manageable judicial standards. This leads us to three other ironies.53 The non-delegation doctrine utterly lacks the intelligible standards that it demands of legislation. It invites and empowers judges to render the legislative process even more chaotic and opportunistic than the doctrine’s advocates think it already is. And it does so at a cost to democratic politics—the very touchstone of those advocates— that can scarcely be imagined.

53 Peter Strauss notes yet another. Were a robust nondelegation doctrine in place, Congress would find it even more necessary than it already is to delegate its legislative power to its own subcommittees and staff, particularly when it enacts so much substantive legislation through the mechanism of massive budget reconciliation bills. In this connec-tion, Strauss quotes Gordon Crovitz to the effect that members of Congress can walk around these statutes and touch them; the one thing that they cannot do, however, is read them. See PETER L. STRAUSS ET AL., ADMINISTRATIVE LAW: CASES AND COMMENTS 195 (9th ed. 1995).