68 Am. U. L. Rev. F. 1 (2018).

*University Professor of Law and Political Science, Vanderbilt University.

Regulatory analysis, which includes cost-benefit calculations, is based on a modern conception of law.  It can only be understood—and certainly can only be improved—if that conception is fully recognized.  In the pre-modern era, law was conceived as a set of normatively-based limits on human behavior designed to maintain public order.  In the administrative state, law is treated as a means of implementing consciously-chosen policies designed to achieve public welfare.  Regulatory analysis is based on this administrative conception of law.  It does not address questions about whether an enactment is a correct interpretation of principle, but rather whether the enactment is likely to achieve some intended public welfare improvement.  Professor Cary Coglianese has proposed that two elements of regulatory analysis, namely cost-benefit analysis and evaluation of regulatory consequences, be imposed by statute on independent agencies.  This is an excellent suggestion, but the modern concept of law on which it is based suggests that we should go further, for both independent and executive agencies.  We should not only require agencies to conduct cost-benefit analysis, a component of policy analysis, but also require that they follow, and document, the whole policy analysis approach.  We should not only require them to evaluate their regulations, but we should also incorporate funding for evaluation into virtually all budgetary allocations and fund the agency to experiment with different versions of a regulation as part of that evaluative process.

Introduction

This Response is about a methodology referred to as regulatory analysis and also about a phenomenon that has no accepted designation but can be described as political immaturity.  Regulatory analysis is the methodology or set of methodologies that can be used to determine whether a regulation (a generally applicable, mandatory pronouncement by an administrative agency or, in other words, a law) is likely to achieve its desired purpose.  Political immaturity is the failure to employ regulatory analysis for every major regulation and the failure to restructure modern governance around that effort.  The prominence of regulations in our system of government springs from the essential and unavoidable necessities of modern, urbanized, and industrial society.¹See Edward L. Rubin, The Constitution and the Administrative State, in The Cambridge Companion to the United States Constitution 292 (Karen Orren & John W. Compton eds., 2018); Edward L. Rubin, From Coherence to Effectiveness:  A Legal Methodology for the Modern World, in Rethinking Legal Scholarship:  A Transatlantic Dialogue 310–11 (Rob van Gestel et al. ed., 2017) [hereinafter From Coherence to Effectiveness].  Modern regulation began during the Progressive era as a response to the rapid industrialization and urbanization of the nation.  See generally John Whiteclay Chambers, II, The Tyranny of Change:  America in the Progressive Era, 1890–1920 (2006); Eric F. Goldman, Rendezvous with Destiny:  A History of Modern American Reform (1952); Richard Hofstadter, The Age of Reform:  From Bryan to F.D.R. (1955); Michael E. McGerr, A Fierce Discontent:  The Rise and Fall of the Progressive Movement in America, 1870–1920 (2003); Nell Irvin Painter, Standing at Armageddon:  The United States, 1877–1919 (1989); Stephen Skowronek, Building a New American State:  The Expansion of National Administrative Capacities, 1877­–1920 (1982); Robert H. Wiebe, The Search for Order, 1877–1920 (1967).  Political immaturity springs from an unwillingness to recognize those necessities.

Cary Coglianese has proposed that independent agencies should expand their use of regulatory analysis, both prospectively through the use of cost-benefit analysis and retrospectively through the use of evaluation and assessment techniques.² Cary Coglianese, Improving Regulatory Analysis at Independent Agencies, 67 Am. U. L. Rev. 733 (2018). He then suggests some ways by which Congress could achieve this goal.³See id. at 737–38.  It could expand prospective regulatory analysis by independent agencies, while preserving their present degree of insulation from the executive, by extending the scope of the cost-benefit requirement in the Unfunded Mandates Reform Act of 1995 (UMRA).⁴Unfunded Mandates Reform Act of 1995, Pub. L. No. 104-4, 109 Stat. 48 (codified at 2 U.S.C. §§ 1501–1571 (2012)).  It could expand retrospective analysis by these agencies through a statutory requirement that an evaluation plan, reviewable by courts on procedural but not substantive grounds, accompany every major regulation, and by providing adequate funding for the resulting evaluative efforts.⁵See Coglianese, supra note 2, at 750–51.

This is an excellent suggestion on the merits.  Professor Coglianese demonstrates UMRA’s virtues in a convincing manner and has delineated a clear and pragmatic path toward its implementation.  One might argue, however, that this good proposal comes at a bad time.  Within the executive branch, the President has used his executive order authority to promulgate various ill-advised and sloppily drafted provisions,⁶See, e.g., Exec. Order No. 13,769, 82 Fed. Reg. 8977 (Feb. 1, 2017) (restricting refugee admissions to the United States and suspending admissions from seven designated countries).  For the legal and technical defects in this order, see Washington v. Trump, 847 F.3d 1151, 1168 (9th Cir. 2017). including a childishly over-simplified effort to reduce the level of regulation by focusing on the arbitrary feature of the way in which these rules are numbered.⁷See, e.g., Exec. Order No. 13,771, 82 Fed. Reg. 9339 (Feb. 3, 2017) (requiring agencies to repeal two existing regulations for every one regulation it enacts).  The mere number of regulations, as opposed to their extent or quality, is the subject of ignorant political oratory, not serious regulatory reform.  The ways in which this Order might be manipulated, either to circumvent its effects or to undermine the agency’s statutory goals would be a relatively easy assignment in a junior high school civics class.  Many of the agencies are now headed by incompetent, corrupt officials who are hostile to their agency’s basic mission; the unprecedented rate at which they have resigned or been dismissed has produced conditions of chaos and demoralization at institutions that carry out basic and essential functions of our government.⁸See Gregory Krieg, The Trump Administration’s Scandals and Embarrassments, in One Place, CNN (Sept. 5, 2018, 5:10 PM), https://www.cnn.com/2018/04/27/ politics/trump-administration-scandals-and-embarrassments; Dan Mangan, Trump’s Cabinet Has Been Rocked by a Number of Ethics Scandals—Here’s a Complete Guide, CNBC (Sept. 5, 2018, 5:11 PM), https://www.cnbc.com/2018/02/15/trump-cabinet-officials-in-ethics-scandals.html; Peter Overby, Trump’s Cabinet Scandals:  Is Abuse of Office Contagious?, NPR (Sept. 25, 2018, 5:12 PM), https://www.npr.org/2018/ 03/08/591524460/trumps-cabinet-scandals-is-abuse-of-office-contagious (noting that “[a]t least seven current and former officials in Trump’s [twenty-four]-member Cabinet have faced accusations of abusing the perks of their office”).  A partial list of turnovers includes Secretary of Health and Human Services Tom Price, who resigned following accusations of corruption; Secretary of State Rex Tillerson, who was dismissed; Secretary of Veterans Affairs David Shulkin, also dismissed; Director of the Center for Disease Control and Prevention Brenda Fitzgerald, who resigned following accusations of corruption; Director of the FBI James Comey, who was dismissed; and numerous members of the President’s staff, including Chief of Staff Reince Priebus.  Meanwhile, many members of the legislative branch appear determined to undermine the very agencies they are ostensibly tasked with sustaining or improving.⁹See Krieg, supra note 8 (providing a complete listing).  They have proposed various bills whose purpose, revealed or disguised to varying extents, is to disable federal agencies from carrying out their statutory responsibilities.¹⁰See, e.g., Regulations from the Executive in Need of Scrutiny Act of 2017 (“REINS Act”), S. 21, 115th Cong. (2017) (requiring that all major regulations be approved by affirmative vote of Congress in order to become effective); Separation of Powers Restoration Act of 2016, H.R. 4768, 114th Cong. (2016) (overruling Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837 (1984)).  Given Congress’s current inability to act on anything, and the extensive opportunities that the system provides for blockage and delay by determined majorities, the REINS Act is nothing more than an effort to impede the regulatory process without acknowledging that goal.  Chevron instructs reviewing courts to defer to executive agency interpretations of statutes on the ground that Congress intended the agency to decide.  Chevron, 467 U.S. at 864–66.  The decision has been criticized on various grounds, but it is clearly grounded on the separation of powers and enforces that principle.  The idea that it violates the separation of powers is absurd.  To recommend any proposal for institutional reform in this environment could be seen as opening the door to truly counterproductive measures that will require decades to undo.  It might be a better strategy for anyone who is truly interested in improving our governmental system, in the way that Professor Coglianese’s proposal would almost certainly achieve, to ally with those who favor consistency or tradition and oppose any alteration of the status quo until the current situation changes in favor of responsible administration.¹¹See Coglianese, supra note 2, at 739–66.

The major impediment to implementing Professor Coglianese’s suggestions, however, is a deeper one and will remain, even if the political landscape changes in a desirable direction.  This impediment is not strategic but conceptual.  The virtues of his proposal emerge from a perspective that is essentially different from the one that prevails at present, even among responsible political leaders and government officials.  This perspective represents a serious effort to promote effective governance that pays heed to the realities of modern industrial society and seeks a management strategy with citizen’s interest in mind.  Denying these realities, to either wish them away or deny the need to rethink familiar but outmoded ideas in response to them, is behavior worthy only of children; in other words, it is political immaturity.

Part I of this Response will explicate the nature of this conceptual issue, that is, the shift in our concept of law from a set of normatively-based limits on human behavior designed to maintain public order to a means of implementing consciously-chosen policies designed to achieve public welfare.  Part II applies the modern concept of law to Professor Coglianese’s proposal to extend cost-benefit analysis to regulations issued by independent agencies and increase their evaluation and assessment efforts of those regulations.  It concludes that these proposals are desirable precisely because they reflect this modern conception.  Part III then suggests that the modern conception of law should lead to further recommendations along the lines Professor Coglianese suggests.  We should not only require independent agencies to conduct cost-benefit analysis, a component of policy analysis, but also require that they follow, and document, the whole policy analysis approach.  We should not only require them to evaluate their regulations, but also incorporate funding for evaluation into virtually all budgetary allocations, and fund the agency to experiment with different versions of a regulation as part of that evaluation process.

I.  Law as Policy Implementation:
The Conceptual Basis of the Proposal

The momentous transformations that occurred in the Western world at the end of the eighteenth century redefined the scope and character of government.¹²For documentation of this account, see Edward L. Rubin, Beyond Camelot:  Rethinking Politics and Law for the Modern State 29 (2005) [hereinafter Beyond Camelot]; Edward L. Rubin, Soul, Self, and Society:  The New Morality and the Modern State 116–29 (2015).  Previously, the main domestic task of government had been to maintain internal order, that is, for the central government to assert its control over the entirety of the geographic area that fell within its jurisdiction.¹³See Beyond Camelot, supra note 12, at 26–27.  This was not an easy thing to do, either conceptually or pragmatically.¹⁴See id. at 25–26.  It took nearly five centuries after the decline of Roman rule for Western Europeans to realize that the imperial mode of governance could not be revived.¹⁵See Soul, Self, and Society, supra note 12, at 34–36 (discussing the changes across Europe that saw to the Empire’s reluctant demise).  Once Charlemagne’s rickety attempt to do so had ended in obvious failure,¹⁶See Roger Collins, History of Europe:  Early Medieval Europe, 300–1000 318–30 (1991); Chris Wickham, The Inheritance of Rome:  Illuminating the Dark Ages, 400–1000 392–404 (2009).  For an intriguing account of the conflict between the desires of European rulers and the realities they faced, see generally Paul Edward Dutton, The Politics of Dreaming in the Carolingian Empire (1994). the political units that emerged set to work on building nation states,¹⁷See Robert Bartlett, The Making of Europe:  Conquest, Colonization and Cultural Change, 950–1350 90–96, 300–14 (1994); James B. Collins, The State in Early Modern France 28–70 (2d ed. 2009); J.H. Elliott, Imperial Spain, 1469–1716 77–129 (1963); Wickham, supra note 16, at 451–52. but they needed another five centuries or so to envision this alternative and develop its essential features.¹⁸On the European state-building process generally, see Thomas Ertman, Birth of Leviathan:  Building States and Regimes in Medieval and Early Modern Europe ch. 3 (1997); Stein Rokkan, State Formation, Nation-Building, and Mass Politics in Europe:  The Theory of Stein Rokkan 153–90(Peter Flora et al. eds., 1999); Samuel E. Finer, Military Forces and State Making, in The Formation of Nation States in Western Europe (Charles Tilly ed., 1975).  By the seventeenth century, the so-called Age of Absolutism, the task was largely done—the independent barons were transformed to dependent courtiers while the now-fragmented church became subordinate to the state.¹⁹See generally 3 Norbert Elias, On the Process of Civilisation:  Sociogenic and Psychogenic Investigations 183, 362–63 (Stephen Mennell et al. eds., Edmund Jephcott trans., 2012).  Elias describes this process as “the courtization of the warriors.”  Id. at 387.

This promising solution to the problem of governance was destabilized by the conceptual development of the Enlightenment, the political developments of the French Revolution, and the economic and technological impact of industrialization.  With traditional and essentially non-governmental systems of local law enforcement, trade regulation, education, and poverty relief in disarray, national governments moved to develop what we now describe as the administrative state:  a centralized, bureaucratic system that provides policing, education, health and welfare services to the citizenry and consciously manages and attempts to improve the economic and social systems of society.²⁰See generally Christopher A. Bayly, The Birth of the Modern World, 1780–1914:  Global Connections and Comparisons 261–67 (2004); Eric J. Hobsbawm, The Age of Revolution:  Europe, 1789–1848 168–71 (1962); Max Weber, Economy and Society:  An Outline of Interpretive Sociology 956–62, 1070 (Guenther Roth & Claus Wittich eds., 1978) (describing bureaucracy as an ideal type and tracing its origins).

As a result of these changes, law—which can be regarded as the modality through which government acts in a generalized and systematic manner—acquired a different role.  In the pre-modern state, its role was to implement the maintenance of internal order.²¹See Weber, supra note 20, at 957–59 (explaining that role of bureaucrat has changed from broad control of internal order to specialized functions).  To achieve this, it established limits on personal behavior.²²Id. at 959–62 (discussing how the transition from rule by divine ordainment transformed the role of bureaucrat from servant of the ruler to vocation).  The rationale behind this use of law was largely sacerdotal.²³See 2 St. Thomas Aquinas, Summa Theologica 956–57 (Fathers of the English Dominican Province trans., 1991).  Some of the limits law established came directly from the idea of natural law, that is, God’s universal rules about relations between human beings that were promulgated through the medium of human reason.²⁴Aquinas distinguished among natural law and eternal law, which governs the structure of the universe, and divine law, which governs the path to salvation and is revealed through Christ.  Id. at 996–1001.  He associated natural law with virtuous behavior.  Id. at 1010–11.  Other limits were acknowledged as being of human origin.²⁵Id. at 1013–22 (discussing human law).  But if those limits were derived from custom, they were seen as products of human reason as well, since the vagaries and even sinfulness of particular lawmakers would cancel out over time, leaving a substrate of rules that the common rationality of humans had generated and approved.²⁶Id. at 1022–25 (discussing changes in human law).  Thus, the law was conceived as a coherent and consistent system that established normatively-based rules for personal behavior.²⁷See Antony Black, Political Thought in Europe, 1250–1450 34–40 (1992); J.P. Canning, Law, Sovereignty and Corporation Theory, 1300–1450, in The Cambridge History of Medieval Political Thought, c. 350–c. 1450 454, 454–56 (J.H. Burns ed., 2007).  The positive enactments of the sovereign were seen largely as intrusions on this system, a view particularly prevalent in England where seventeenth century absolutism had been checked by revolution, and the customary or common law of the realm remained in force.²⁸See From Coherence to Effectiveness, supra note 1, at 310.

In the modern state, by contrast, the role of law is to implement the broad range of economic and social programs that the administrative state is charged with creating and maintaining.²⁹Id. at 310–11.  Law is used to prevent harm to employees and consumers, to protect the environment, to control the economy, to provide health and welfare services, to manage recreational facilities—the list goes on and on.  Agency-run programs are positive enactments of the sovereign or its delegates, conceived as carrying out consciously designed social policy.  Due to this positive character, and closely connected conceptual changes in the general society, the law is now seen as exclusively secular, the product of human will.³⁰Id. at 319.In addition, it is particularized; instead of forming a coherent and consistent system, it functions as a means for implementing each of the modern state’s individual enactments.³¹Id. at 321–22.  These enactments vary greatly in methodology and purpose, based on the will of the lawmaker (a shifting political majority in democratic regimes) and the requirements of the particular subject matter.³²See Beyond Camelot, supra note 12, at 191–226.

In short, the advent of the administrative state has changed the purpose and the meaning of law as well.  Law is no longer a coherent and consistent system of normatively-based rules for personal behavior; instead, it is an instrumentality for carrying out regulatory policy and is often specific to each policy.  A conceptual shift of this magnitude is both challenging and disconcerting.  It is challenging because time and coordination are needed to develop new ideas about governance; the shift from the imperial to the national model of governance took five centuries, as noted above.³³See Soul, Self, and Society, supra note 12, at 34–36.  It is disconcerting due to what John Kenneth Galbraith described as a vested interest in understanding, our attachment to familiar ideas and our reification of those ideas as necessary truths even when they have ceased to be necessary or even truthful.³⁴See John K. Galbraith, The Affluent Society 7 (1998) (stating that “[f]or a vested interest in understanding is more preciously guarded than any other treasure”).

Professor Coglianese’s proposal provides us with a prospective methodology for determining whether a given enactment, specifically an enactment by an independent administrative agency, is likely to be effective in carrying out a regulatory program.³⁵See Coglianese, supra note 2, at 737–39 (discussing and comparing retrospective and prospective analysis).  It also provides us with a retrospective methodology for determining whether a given enactment has been effective.³⁶See id.  But treating law in this manner embodies the modern view of law as a means of implementing public policy and rejects the view of law as a coherent system of normative rules.  Those who find the modern view of law to be challenging or disconcerting will resist adopting his proposal.  This is the major impediment that stands in the way of its adoption.

II.  Policy Implementation as Regulatory Analysis:
The Implications of the Proposal

These observations about the conceptual structure of modern law serve as both a caveat regarding the practicality of Professor Coglianese’s proposals and as an appreciation of their conceptual significance.  Consider first his suggestion that Congress amend UMRA to extend cost-benefit analysis to independent agencies.³⁷See id. at 749–51.  As he notes, cost-benefit analysis was first incorporated into the American regulatory process through an executive order issued by President Reagan.³⁸See Exec. Order No. 12,291, 46 Fed. Reg. 13,193 (Feb. 17, 1981) (requiring that all agencies submit a regulatory plan to the Office of Management and Budget (OMB) each year and that executive agencies submit all significant regulations to OMB accompanied by a cost-benefit analysis).  In doing so, Reagan’s primary goal was probably not the desire to improve the quality of regulation but rather to restrict its scope.³⁹See Richard L. Revesz & Michael A. Livermore, Retaking Rationality:  How Cost-Benefit Analysis Can Better Protect the Environment and Our Health 24–26 (2008).  Reagan ran as an opponent of regulation,⁴⁰See Martha Derthick & Paul J. Quirk, The Politics of Deregulation 30–35 (1985); Theda Skocpol, Social Policy in the United States:  Future Possibilities in Historical Perspective 232–33, 303–06 (1995). after all, and during his term took initiative to privatize government functions,⁴¹See Jon D. Michaels, Constitutional Coup:  Privatization’s Threat to the American Republic 86–98 (2017). appoint agency heads who were hostile to the agency’s mission, and reduce administrative control over protecting the environment, auto safety, civil rights, and other areas.⁴²See Ian Ayres & John Braithwaite, Responsive Regulation:  Transcending the Deregulation Debate 7–10 (1992); Bruce Bartlett, The New American Economy:  The Failure of Reaganomics and a New Way Forward 151–54 (2009) (explaining that economic ineffectiveness of benefits, cutbacks, and deregulation led to the failure of Reaganomics).  His famous quote—“the nine most terrifying words in the English language are:  I’m from the Government and I’m here to help”⁴³Ronald Reagan, U.S. President, The President’s News Conference (Aug. 12, 1986), https://www.presidency.ucsb.edu/documents/the-presidents-news-conference-957.—reflect a truly irresponsible disparagement of the institution he was supposed to lead and of the many individuals who had devoted their careers to government service.⁴⁴For a description of Reagan’s impact on civil servants, see generally Marissa M. Golden, What Motivates Bureaucrats?  Politics and Administration During the Reagan Years (2000).

Reagan’s cost-benefit executive order was continued by George H.W. Bush, not surprisingly since Bush had been part of the Reagan administration and shared its ideology.⁴⁵See Stephen Skowronek, The Politics Presidents Make:  Leadership from John Adams to George Bush 429–42 (1993).  The heading for this section is:  “George Bush as the Faithful Son.” Id. at 429.  One can readily identify a group of Presidents who served as Vice President to a transformative leader and followed that leader’s policies; the list includes John Adams, Martin van Buren, Andrew Johnson, William Howard Taft, and Harry S. Truman, in addition to Bush.  Id. at 430.  But it was continued again, and given its present form, by the Clinton administration,⁴⁶See Exec. Order No. 12,866, 58 Fed. Reg. 51,735 (Oct. 4, 1993) (establishing the current regime of mandatory cost-benefit analysis for regulations promulgated by executive branch agencies). and then continued, in only a slightly amended form, by the Bush II, Obama, and Trump administrations as well.⁴⁷George W. Bush amended it by issuing Executive Order No. 13,258, 67 Fed. Reg. 9385 (Feb. 28, 2002), and again by issuing Executive Order No. 13,422, 72 Fed. Reg. 2763 (Jan. 18, 2007), making minor changes in each case.  See Roger G. Noll, The Economic Significance of Executive Order 13,422, 25 Yale J. on Reg. 113, 113 (2008) (explaining that the only significant change of Exec. Order No. 13,422 was the economic analysis requirement for guidance documents).  Barack Obama then revoked both of George W. Bush’s Executive Orders, see Exec. Order No. 13,497, 74 Fed. Reg. 6113 (Feb. 4, 2009), and enacted his own slight revision in Executive Order 13,563, 76 Fed. Reg. 3821 (Jan. 21, 2011).  The survival of Reagan’s executive order through the political pendulum swings of the past twenty-five years suggests that cost-benefit analysis serves a purpose beyond Reagan’s hostility towards the functions of modern government.  A plausible hypothesis is that it provides the President with an instrument for controlling the federal government’s vast and multifarious agencies, and thus a means of implementing his own policies.⁴⁸See Elena Kagan, Presidential Administration, 114 Harv. L. Rev. 2245, 2246 (2001) (arguing that enhanced presidential control over administrative agencies can serve pro-regulatory objectives).  As Professor Coglianese notes, the Executive Order, in its present form, repeatedly demands that regulations must be consistent with the “President’s priorities.”⁴⁹See Coglianese, supra note 2, at 747 (explaining that the phrase, “the President’s priorities,” is used eleven times throughout Executive Order 12,866).  The Order enforces that demand by requiring agencies to submit their plans for regulations to the Office of Information and Regulatory Affairs (OIRA),⁵⁰OIRA was established by the Paperwork Reduction Act, Pub. L. 106–398, 114 Stat. 1654 (2000) (codified at 44 U.S.C. §§ 3501–3549), as a part of the OMB, with an administrator appointed by the President with the Senate’s advice and consent.  Id. § 3503.  Its initial mission was to “minimize the paperwork burden for individuals, small businesses, educational and nonprofit institutions, Federal contractors, State, local and tribal governments . . . .” Id. § 3501(1).  OIRA acquired a much more significant role in government when it became the unit within OMB assigned to implement cost-benefit review of executive agency regulations that President Reagan established by executive order.  See Exec. Order No. 12,281, 46 Fed. Reg. 7923 (Feb. 17, 1981).  In the current version of this provision, Exec. Order No. 12,866, 3 C.F.R. § 638, reprinted in 5 U.S.C. § 601 (1993), OIRA is explicitly identified as the “repository of expertise concerning regulatory issues,” id. § 2(b), and assigned to carry out the bulk of the provision’s functions.  For a general history, see James F. Blumstein, Regulatory Review by the Executive Office of the President:  An Overview and Policy Analysis of Current Issues, 51 Duke L.J. 851 (2001). by preventing agencies from issuing regulations without OIRA approval, and by referring conflicts between agencies regarding their regulatory policies to the Vice President and the President.⁵¹See Coglianese, supra note 2, at 746–47.

If cost-benefit analysis is viewed as either a component of Reagan’s hostility toward modern government or as a device for imposing the President’s priorities on administrative agencies, it would be a mistake to extend it to independent agencies by congressional action.  To be sure, Professor Coglianese accounts for this second concern by proposing that cost-benefit analysis be extended to independent agencies by congressional amendment of UMRA, rather than by presidential amendment of Executive Order 12,866.⁵²Id. at 749–51. In addition to resolving all doubts about the legality of the proposal, this alternative approach would keep the cost-benefit analysis of the independent agencies out of the clutches of OIRA.

That might not resolve the most serious problem with cost-benefit analysis, however.  If the governmental mechanisms are judged by their uses, the fact that cost-benefit analysis has been used for regulatory rollbacks and policy control by political actors might suggest that it is not a particularly promising mechanism to apply to independent agencies.⁵³See Rachel A. Potter, How the Trump Administration can use Benefit Cost Analysis to Justify Deregulation, Brookings (Aug. 1, 2017), https://www.brookings.edu/research/ how-the-trump-administration-can-use-benefit-cost-analysis-to-justify-deregulation (discussing how the Trump administration is presently using cost-benefit analysis to fuel a politically motivated deregulatory agenda).  Independent agencies are supposed to regulate; that is the reason for their creation.  The executive agencies have various other purposes, such as foreign relations, national security, internal policing, immigration control, government financing, veterans’ benefits, and so forth, that cannot be defined as truly regulatory.⁵⁴See The Executive Branch, White House, https://obamawhitehouse.archives. gov/1600/executive-branch (last visited Oct. 17, 2018) (describing the various executive departments and agencies, as well as their functions).  But most independent agencies are created by Congress for the specific purpose of regulating defined sectors of the economy.⁵⁵See Harold H. Bruff, On the Constitutional Status of the Administrative Agencies, 36 Am. U. L. Rev. 491, 512–14 (1987).  Moreover, these agencies are supposed to regulate independently.⁵⁶See id. at 513 (noting that “there is no single elected officer to whom the independent agencies are accountable”).  If cost-benefit analysis is imposed by Congress, perhaps congressional committees will be encouraged or induced to increase their oversight of these agencies­—a perfectly legal approach but one that seems contrary to the rationale behind an independent agency’s creation.  If Congressional imposition of cost-benefit analysis on independent agencies does not increase congressional oversight, one might question the value of a mechanism that has been designed for control when that control is not being exercised and is not supposed to be exercised.  In other words, the purposes that seem to lie behind the introduction of cost-benefit analysis might suggest that it is a mistake to extend that approach to independent agencies.

The reason it is not a mistake depends on the conceptual shift between the pre-modern idea that law is a declaration of norms and the modern idea that law is an implementation of social policy.  Administrative regulations, it is generally agreed, are supplementary laws, issued under the authority of the legislature as the primary lawmaker in our democratic system.⁵⁷See Administrative Law, Cornell Legal Info. Inst., https://www.law.cornell. edu/wex/administrative_law (last visited Oct. 17, 2018) (outlining the authority of administrative agencies).  If they are normative declarations, or ramifications of such declarations, then independent agencies should be able to declare them despite presidential disapproval or presidential control.  But if these regulations are means of implementing policies established by the legislature, then they should be subject to supervision to ensure that they will be effective in carrying out their intended mission.  The question of whether they achieve social benefits that justify the costs that they impose is a crucial component of effectiveness, and one that should generally be considered when promulgating regulations.  Subjecting this issue to supervision by an external reviewer is good administrative practice; administration is a form of management, and supervision is an important part of the management process.⁵⁸See, e.g., Edwin C. Leonard, Jr., Supervision:  Concepts and Practices of Management 9–10 (12th ed. 2013); John W. Newstrom, Supervision:  Managing for Results 2 (10th ed. 2013).

Once regulation is conceived in the appropriate administrative and managerial terms, it is possible to supervise the independent agency’s use of cost-benefit analysis without compromising its independence.⁵⁹For an interesting look at the difference between administrative supervision and democratic accountability, see Edward Rubin, The Myth of Accountability and the Anti-Administrative Impulse, 103 Mich. L. Rev. 2073, 2073 (2005).  Accountability, generally defined as some way in which a government agency should be answerable to the public, is often treated as a means of supervision, but real supervision, in a bureaucratic setting, is generally exercised by a hierarchal superior.  See id. at 2134.  While many parts of the executive branch are less immediately connected to the President than OIRA—which is one of the relatively small number of agencies that work directly with the President⁶⁰See generally Bradley H. Patterson, Jr., The White House Staff:  Inside the West Wing and Beyond (2001); Nelson W. Polsby, Congress and the President 76–84 (4th ed. 1986); Beyond Camelot, supra note 12, at 46–48; Justin S. Vaughn & José D. Villalobos, Czars in the White House:  The Rise of Policy Czars as Presidential Management Tools (2015); Chris Whipple, The Gatekeepers:  How the White House Chiefs of Staff Define Every Presidency (2017); Justin S. Vaughn & José D. Villalobos, Conceptualizing and Measuring White House Staff Influence on Presidential Rhetoric, 36 Presidential Stud. Q. 681 (2006).—any Article II supervision seems best avoided.⁶¹See Office of Management and Budget:  Information and Regulatory Affairs, White House, https://www.whitehouse.gov/omb/information-regulatory-affairs (last visited Oct. 7, 2018).  The federal courts are certainly independent of the President, but their established basis for review is legal, not managerial or economic, and they are already criticized for rigidifying or “ossifying” the regulatory process.⁶²See, e.g., Thomas O. McGarity, Some Thoughts on “Deossifying” the Rulemaking Process, 41 Duke L.J. 1385 (1992) (arguing that the hard look doctrine and other amplifications of judicial review impede the efficiency and flexibility of the rulemaking process); Richard J. Pierce, Jr., Seven Ways to Deossify Agency Rulemaking, 47 Admin. L. Rev. 59, 60 (1995) (explaining how agencies over the past decade have been increasingly turning to informal methods for promulgating policies, which does not provide for adequate opportunity for public participation in the policymaking process).  But see Jason W. Yackee & Susan W. Yackee, Administrative Procedures and Bureaucratic Performance:  Is Federal Rule-making Ossified?, 20 J. Pub. Admin. Res. & Theory 261, 261 (2009) (arguing that the “ossification” thesis lacks merit as empirical data suggests that procedural constraints may actually speed up the promulgation of regulations).  The alternative is a reviewing institution that is part of the legislative branch but independent of its political process.  A number of nations use an ombudsperson’s office for this purpose.⁶³See, e.g., Walter Gellhorn, Ombudsmen and Others:  Citizens’ Protectors in Nine Countries xi–xii (1966) (discussing several countries that employ ombudsmen including:  Denmark, Finland, New Zealand, Norway, Sweden, Poland, and Japan).  The term refers to a wide variety of institutions, but its classic structure is an independent office answerable to the legislature that can receive complaints about administrative agencies, provide advice to those agencies, and intervene to some extent in its proceedings.⁶⁴See Edward L. Rubin, Bureaucratic Oppression:  Its Causes and Cures, 90 Wash. U. L. Rev. 291, 327–32 (2012).  For a descriptive account on the ombudsmen enterprise, see Trevor Buck et al., The Ombudsman Enterprise and Administrative Justice (2011); Gellhorn, supra note 63.  Another possibility would be to assign the supervisory function to the Government Accountability Office (GAO), an acknowledged non-partisan agency that provides investigative reports on the administrative apparatus to Congress.⁶⁵GAO is an Article I agency, like the Congressional Budget Office and the Congressional Research Service, within the Library of Congress.  See Bruce Bimber, The Politics of Expertise in Congress:  The Rise and Fall of the Office of Technology Assessment 93–99 (1996) (describing the ability of Article I agencies to adopt and maintain a non-partisan stance).  Given its present role, and because it would be monitoring all the independent agencies, the GAO could readily develop expertise in cost-benefit analysis.

In addition to prospective efforts to ensure that the regulation is likely to be effective, Professor Coglianese recommends retrospective analysis to determine whether the regulation has actually served its intended purpose.⁶⁶See Coglianese, supra note 2, at 759–62.  He proposes that independent agencies be required to issue evaluation plans when promulgating regulations and that funding be provided so that these plans can be instantiated.⁶⁷Id. at 760–62.  This proposal, like the prospective one, reflects and relies upon the modern concept of law as policy implementation.  If a statute or a regulation is seen as the declaration of a norm, then no retrospective evaluation is required; the only question is whether the enactment has correctly stated the norm.  Subsequent interpretations of the language may produce normatively undesirable effects, of course.  One response, probably the more common one, is to assert that these are improper interpretations of the enactment.  Another response, however, is to argue that the statutory language is unclear, that it does not state the norm with sufficient clarity.

But evaluation is distinctly different from interpretation.  The question is not whether the enactment is stated correctly, but whether it achieves the desired results in the real world.  This treats regulations as means of implementing the social policies established by the legislature.  Whether the regulation, and whether the statute that authorizes it, combines with other statutes and regulations to form a consistent or coherent body of law is largely irrelevant.  Such consistency and coherence might be a minor convenience to regulated parties, and an aid to codification, but it will be no more than that.  In each area or issue, the political process generates a particular policy, one that is designed to achieve society’s goals in that area.  In transportation, the free market may work best; in financial services, extensive supervision may be needed; in agriculture, economic subsidies may meet society’s preferences.  The overarching value, for regulations at least, is whether those varied policies are being effectively implemented.  This is the question that evaluation is designed to answer.

III.    The Possibilities of Regulatory Analysis:
Beyond the Proposal

Once Professor Coglianese’s proposal is placed in the context of modern government and our modern conception of law, it becomes possible to suggest some ramifications and extensions of the proposal that would institute more far-reaching changes in our mode of governance.  Such extensions, admittedly, have the disadvantage of being less likely to be implemented.  The purpose of proposing them is to provide insights about the nature of modern government and the possibilities for its ultimate improvement.  It might be noted, however, that Professor Coglianese’s proposal, convincing and delimited as it may be, will probably not be implemented in the current political environment either, given that both the executive and legislative branches seem to have become dysfunctional.  If it were implemented, it would probably be in the wrong way and for the wrong reasons.

To begin with, the extension of cost-benefit analysis to independent agencies would probably improve the quality of those agencies’ decision making to some extent.  But cost-benefit analysis is not the best that we can do in the effort to make regulations more effective.  It might be compared to a hammer, the most familiar tool in twentieth-century philosophy,⁶⁸See Martin Heidegger, Being and Time 98–99 (John Macquarrie & Edward Robinson trans., 1962); Michael Polanyi, Personal Knowledge:  Toward a Post-Critical Philosophy 59, 175 (1974) (using a “hammer” as a metaphor in explaining that one tool is not enough to solve a problem, but instead, one must utilize many different tools).  which was developed in response to a particular quandary or dilemma.  Perhaps there was a wooden stake that needed to be driven into hard ground to tether some primeval genius’s cow, and all that was available was an irregularly shaped stone or the genius’s head.  Cost-benefit analysis is similarly directed to one purpose, which is the possibility that heavy costs will be imposed on private business for relatively little public benefit.⁶⁹See Amit Narang, The Stunning Triumph of Cost-Cost Analysis, The Reg. Rev. (Feb. 19, 2017), https://www.theregreview.org/2017/02/19/narang-stunning-triumph-cost-cost-analysis (discussing how cost-benefit analysis from Reagan to Trump focuses on easing business burdens).  Like many other enactments, it was probably a response to a particular event or group of events, apparent horribles that galvanized the relevant decision makers.  The triggering event may have been the snail darter controversy, which unfolded through the 1970s.  Its upshot, which occurred in 1978, was the Supreme Court’s decision that the massive and costly Tellico Dam,⁷⁰Tenn. Valley Auth. v. Hill, 437 U.S. 153, 193–95 (1978). although nearing completion, had to be dismantled under the Endangered Species Act⁷¹Endangered Species Act of 1973, Pub. L. No. 93–205, 87 Stat. 884 (codified as amended at 16 U.S.C. §§ 1531–44 (2012)). to preserve a small unprepossessing fish whose presence in the river was not known at the time construction began.⁷²See generally Kenneth M. Murchison, The Snail Darter Case:  TVA Versus the Endangered Species Act 1–6 (2007); Zygmunt J.B. Plater, The Snail Darter and the Dam:  How Pork-Barrel Politics Endangered a Little Fish and Killed a River 1–5 (2013).

In other words, cost-benefit analysis may be a valuable tool, but what is truly needed is a tool box.  In addition to knowing the economic value of the costs and benefits of a particular regulatory strategy, we need to know how to define the underlying problem in a clear and manageable way; what existing approaches are available to solve it; whether new approaches can be developed; what institutional capacities for implementation presently exist or need to be established; what distributional effects might flow from particular approaches; what the risk of failure is for each approach, how severe will the consequences of failure be; and whether they will be reversible.  To adopt any regulation that will have a significant impact on our society without at least trying to make these prospective judgments is irresponsible, a form of political immaturity.

In fact, we have such a tool box, although it is a somewhat rudimentary one—the methodology of policy analysis.⁷³See, e.g., Eugene Bardach & Eric M. Patashnik, A Practical Guide for Policy Analysis:  The Eightfold Path to More Effective Problem Solving xv–xix (5th ed. 2016) (proposing eight steps that beginners should take in problem-solving and policy analysis); William N. Dunn, Public Policy Analysis:  An Introduction 5–7 (5th ed. 2012); Stuart S. Nagel, Handbook of Public Policy Evaluation xi (2002); Malcolm K. Sparrow, The Regulatory Craft:  Controlling Risks, Solving Problems, and Managing Compliance 1–14 (2000); Edith Stokey & Richard Zeckhauser, A Primer for Policy Analysis 3–7 (1978).  Its essential features are to define the problem in pragmatic terms, to generate an array of alternatives that seem like possible solutions, to determine which of these possible solutions seems to offer the greatest promise of success, to implement that alternative, and then to evaluate the result.  This approach is supported by well-developed theoretical considerations⁷⁴See, e.g., Frank Fischer, Reframing Public Policy:  Discursive Politics and Deliberative Practices 76–81 (2003); Aaron Wildavsky, Speaking Truth to Power:  The Art and Craft of Policy Analysis 1–9, (1979); Elinor Ostrom, Institutional Rational Choice:  An Assessment of the Institutional Analysis and Development Framework, in Theories of the Policy Process 21, 35 (Christopher M. Weible & Paul A. Sabatier, eds., 4th ed. 2018). and has been implemented in a variety of settings, both public and private.⁷⁵See Sparrow, supra note 73, at 181–93 (describing considerations that underlie effective regulatory programs).  See generally John Friedmann, Planning in the Public Domain:  From Knowledge to Action (1987) (describing developments in planning theory).  It has been subject to extensive criticism⁷⁶Perhaps the best-known critique is Charles E. Lindblom, The Science of “Muddling Through,” 19 Pub. Admin. Rev. 79, 79–80 (1959) (arguing that policy analysis methods prescribed in the literature are impractical and unrealistic due to constraints in time, resources, and intellectual capacities).  For additional critiques, see David Braybrooke & Charles E. Lindblom, A Strategy of Decision:  Policy Evaluation as a Social Process 21–22 (1963); Charles E. Lindblom, The Intelligence of Democracy:  Decision Making Through Mutual Adjustment (1965); and Charles E. Lindblom, Still Muddling, Not yet Through, 39 Pub. Admin. Rev. 517, 517–18 (1979). and is far from a definitive decision making protocol, but it is certainly a start—the kind of conscientious effort that politically mature and responsible public officials should be expected to employ.⁷⁷Administrative agencies are structured to facilitate this kind of decision making when they promulgate regulations.  The informal rulemaking procedures of § 553 of the Administrative Procedure Act, 5 U.S.C. §§ 551–559, 701–706 (2012), do not require any such protocol, but they create space for it to occur, since the agency is not required to publish its preliminary rule until the rule has been designed.  Statutes should also be subjected to a policy analysis of this sort, but the institutional structure of the legislature is much less conducive to this approach.  For an effort to find space for it nonetheless, see Edward L. Rubin, Statutory Design as Policy Analysis, 55 Harv. J. on Legis. 143 (2018).

As is apparent, cost-benefit analysis is one aspect of one component of policy analysis.  It becomes relevant when the alternative solutions are being considered; at that point, it responds to crucial concerns about the imposition of unnecessary expenses on the individuals or firms who will be subject to the program.  It does not, however, answer all the questions that are relevant even at this stage, such as what institutions can implement the program and whether they have the necessary resources, or what the distributional effects of the program will be.  Still more seriously, all these considerations are only one component of the general methodology.  They may tell us how to assess a given alternative, but they are of little help in generating alternatives, that is, developing new ideas to solve the problem that may have a greater chance of success than the more traditional or familiar ones.

If regulatory analysis is to be extended to independent agencies, it would be preferable to require the agency to demonstrate that it followed the full methodology, rather than imposing one component of that methodology.  There is nothing impractical about such a requirement.  All the agency would need to do is to document the process that it should be following as a matter of good practice, and without being required to do so.  In fact, the agency should document that process as well, so the only additional requirement would be to publish the relevant documentation.  Judicial review would impose a significant burden on the agency, but UMRA does not authorize such review, as Professor Coglianese notes.⁷⁸See Coglianese, supra note 2, at 749–51 (discussing the UMRA).  The disciplining force of the requirement would reside in the obligation to publish the agency’s documentation, and the public scrutiny of the agency’s decision-making process that would result.

To extend the required procedure from cost-benefit analysis to policy analysis in general is to recognize the nature of modern law as a means of identifying and implementing regulatory policy for a nation, that, in the words of Jean Genet, is “no longer childlike, but severe.”⁷⁹Jean Genet, The Thief’s Journal 72 (Bernard Frechtman trans., 1964).  Genet is describing a sailor who had been sentenced to death for turning over information to the enemy:  “I am not talking about an act of treason causing the loss of a naval battle which is slight, hanging from the wings of a schooner’s sails, but of the loss of a combat of steel monsters wherein dwelt the pride of a people no longer childlike, but severe . . . .”  Id.

This same recognition that the meaning of law has changed as a result of contemporary circumstances might also lead to an expanded version of the retrospective part of Professor Coglianese’s proposal.  There can be no argument with the idea that regulations should be evaluated for their effectiveness.  This should be an invariable part of agency rulemaking practice, required by the Administrative Procedure Act,⁸⁰See Administrative Procedure Act, Pub. L. No. 89–554, 80 Stat. 381 (codified as amended in scattered sections of 5 U.S.C.). and the first issue that a court addresses when considering anything other than a facial challenge to a regulation.  Designated funding for evaluation should be part of the budgetary allocation of every agency that is authorized to issue regulations, and the greater the desire to subject the agency to cost-benefit analysis, the more extensive the funding for evaluation should be.  The fact that none of this is true is a further indication that we are still thinking about law—and specifically the subsidiary law promulgated by agencies—as a normative declaration, and not as an effort to implement defined policy.

Again, the realities of modern government suggest that we should go considerably further than mere retrospective evaluation.  If we truly comprehend the nature of regulation in our administrative state, then, as Rob van Gestel and Gijs van Dijck have proposed, we should require that regulations, barring special circumstances, be implemented on a trial or experimental basis and evaluated on the basis of that trial.⁸¹See Rob van Gestel & Gijs van Dijck, Better Regulation Through Experimental Legislation, 17 Eur. Pub. L. 539, 542–46 (2011) (proposing that governments adopt an experimental approach when enacting new laws or regulations, requiring them to first be tried on a small number of the population for a limited period of time).  Incorporated into the policy-making process, this procedure would allow the agency to test the range of alternatives that it had generated and found plausible and practicable.  The procedure would not be necessary if the agency were certain that its proposed regulation was the best solution, and it might not be advisable in cases of severe emergency.  But in ordinary cases, it would seem to be an essential component of a serious, mature approach to regulation.

Consider an example that is free from our traditional approach to law.  Suppose a fatal, infectious disease has become rampant in society, and scientists have developed several different vaccines that they think might prevent people from contracting it.  Each vaccine has demonstrated success with laboratory animals, but each has potentially serious side effects for a small percentage of the animals on which it has been tested.  It would be truly monstrous to choose one of these vaccines, without further information, and administer it to the entire population of the nation.  The responsible approach would be to try each vaccine on a delimited group to see which one provided the best protection with the fewest side effects—in other words, which vaccine had the best ratio of benefits over costs.  Even in a situation that can be described as an emergency, and even one that involves life or death, the value of testing or experimentation is apparent.

An experimental procedure of this sort does not implicate any of the moral violations that are involved in the idea of “experimenting on human beings.”  That phrase, and its invocation of Dr. Mengele, applies when people are subjected to a potentially harmful treatment that is of no benefit to the people themselves but only for the sake of gaining general information.⁸²See Edwin Black, War Against the Weak:  Eugenics and America’s Campaign to Create a Master Race 324, 356­–59 (2d ed. 2012) (describing experiments at Buchenwald and Auschwitz); Eugen Kogon, The Theory and Practice of Hell:  The German Concentration Camps and the System Behind Them 143–68 (Heinz Norden trans., 1950); Robert Jay Lifton, The Nazi Doctors:  Medical Killing and the Psychology of Genocide 269–302 (1986).  But experimental regulation, like the hypothesized vaccine, does not raise these concerns.  It is an effort to benefit people, recognizing that we cannot know in advance whether the expected benefits will be realized, and whether negative consequences will result.  Far from using human beings as “guinea pigs,” it treats them with respect and caution, avoiding, or more precisely, minimizing the harm that may result from imposing an untested treatment on the general populace in an effort to secure a necessary or desired benefit.

Conclusion

It is time for us, as a society that takes pride in governing itself, to grow up.  We need to embrace the fact we have a new form of government, the administrative state, one that is a necessity in the modern world and that we ourselves have chosen.  We need to recognize that this new form of government involves a new concept of law.  Professor Coglianese’s proposals are an important contribution to that recognition, in the context of a group of agencies that have been too often designed or managed in pre-modern terms.