By: Sidney Shapiro*
Elizabeth Fisher**
Wendy Wagner***

“Expert discretion is the lifeblood of the administrative process . . . .”[1]

The history of administrative law in the United States constitutes a series of ongoing attempts to legitimize unelected public administration in a constitutional liberal democracy.[2]  It is a history of many twists and turns in which public administration and understandings of its legitimacy have coevolved.[3]  It is also a history that took a very wrong turn with what Richard Stewart described in 1975 as a “reformation” in the subject.[4]  The reformation provided a “surrogate political process,” one that was intended to ensure a “fair representation of a wide range of affected interests in the process of administrative decision.”[5]  Even at the time, Professor Stewart was skeptical the reformation could effectuate an interest-group process that produced “outcomes that better serve society as a whole.”[6]  In this he was prescient.  Empirical evidence indicates that industry interests dominate the rulemaking process in a number of important areas of social regulation, with no public interest representation at all in many rulemakings.[7]

The reformation was an effort to strengthen the relationship between administrative law, accountability, and legitimacy following a basic model later dubbed the “rational-instrumental” paradigm.[8]  Under this model, agency accountability is ensured by deploying various external scientific, technical, and legal oversight processes to prevent agency staff from exercising discretion.  Besides failing at its own goal, however, the reformation and the rational-instrumental reforms that have followed it have had four adverse consequences for public administration.

First, this discretion-free or rational-instrumental paradigm treats public administration as a simple agent of the legislature, rather than a substantive institution in its own right, even though this understanding has always been at odds with regulatory and legislative realities.  Nevertheless, procedure after procedure has been added in a vain effort to eliminate discretion.  The result has been the ossification of rulemaking to the point where important and controversial rules usually take five or more years to make and sometimes even a decade or longer.[9]

Second, and less noticed, has been the impact on public administration.  Because the rational-instrumental paradigm refuses to acknowledge the legitimacy of administrative discretion, it facilitates the bureaucracy bashing that is all too common in the political system.  The result is that we have made it unattractive for the very professionals who are necessary for public administration to want to work for the government.[10]

Third, the rational-instrumental paradigm has resulted in a distinction being drawn between the scientific and participatory aspects of administrative decision making.  Not only is this dichotomy a false one, it has undermined regulatory science in a number of ways, including hiding the dangers involved in the interaction between science and interest representation.[11]  It has also resulted in both expertise and participation being treated as monolithic concepts when in fact there are many different versions of each.  Administrative law scholars’ understandings about the nature and potentialities of the administrative state have therefore been unduly narrowed.

Finally, the focus on the rational-instrumental paradigm has deflected attention away from the deliberative-constitutive paradigm, which dates back to at least the Progressive Era and has greatly influenced the development of public administration.  This paradigm relies on expertise, deliberation, and reason giving to establish the legitimacy of public administration.

For enlightenment, administrative law must develop “a constitutional design that accepts the need for supplementary bureaucratic lawmaking in the ongoing regulatory enterprise but self-consciously confronts the serious legitimization problems involved.”[12]  Scholars and lawyers must broaden their intellectual and conceptual vision.  Enlightenment requires recognition of the role of expertise and discursive decision making in the legitimization of administrative discretion.  To put the matter a different way, we need to look inside the agency for administrative legitimacy.

Contemporary administrative law scholarship and practice is so deeply enmeshed in rational-instrumental accountability that it is difficult for administrative lawyers to imagine that there is a complementary approach to legitimacy.  Yet, the history of administrative law, in this and other jurisdictions, highlights the significance of the deliberative-constitutive paradigm.[13]  The deliberative-constitutive paradigm embraces, rather than rejects, the professionalism of agency staff; agency professionalism is viewed as a positive attribute that helps ensure the integration of technical expertise in rulemaking and serves as a buffer against undue influence by highly interested stakeholders.  Indeed, in light of the demise of interest-group pluralism in rulemaking and the scholarly dead end in which we find ourselves, it is time to recognize and develop the deliberative-instrumental paradigm.

Our case for looking inside the agency for legitimacy proceeds in five steps.  Part I introduces the concept of “administrative constitutionalism,” which encompasses the debate over what should be the role and nature of public administration to ensure its legitimacy.  It then lays out the elements of the rational-instrumental and deliberative-constitutive paradigms and explains how they contribute to administrative constitutionalism respectively from the outside-in and the inside-out.  Part II provides a brief history of administrative constitutionalism, which reveals there have been ongoing tensions between two paradigms—and thus between outside-in and inside-out accountability—since the 1880s.  Part III elaborates on our argument that the current emphasis on the rational-instrumental model has made administrative constitutionalism unsustainable.  Part IV argues that acknowledging and developing the deliberative-constitutive paradigm will strengthen administrative constitutionalism by admitting the existence of agency discretion and by looking for realistic ways to make it accountable.  Finally, Part V offers a case study in how the deliberative-constitutive paradigm can contribute to administrative constitutionalism.

I.  Administrative Constitutionalism

The history of administrative law in the United States has been a continuing effort to fit the “‘round peg’ of administrative government into the ‘square hole’ of the nation’s constitutional culture.”[14]  In this history we see continuous debates over what should be the role and nature of public administration so as to ensure it is legitimate—a discourse Elizabeth Fisher describes as “administrative constitutionalism.”[15]  Legitimacy is a notoriously treacherous concept.  Here, we use it to refer to procedural concepts that have the quality of “worthiness to be recognized”[16] within a polity committed to liberal constitutional democracy.

Therein lies the dilemma.  While constitutional principles, such as the rule of law and the separation of powers, are relevant in thinking about the role and nature of public administration, they are wholly inadequate by themselves to address, in full, the issue of how unelected administrative power should be constituted and limited.  In these circumstances, what has emerged is a debate that is semiautonomous from constitutional law over the role and nature of public administration.  Administrative constitutionalism is about the normative nature of law and different understandings of administrative legitimacy.[17]  As such, it is an “essentially contested concept,” akin to concepts such as democracy and the rule of law.[18]

Two different paradigms have fed the debate over administrative legitimacy.  By paradigms, we mean “descriptions of how our political world is organized and how it works.”[19]  Each paradigm manifests a different understanding of accountability.  A rational-instrumental paradigm underlies outside-in accountability,[20] while a deliberative-constitutive paradigm underlies inside-out accountability.  This Part describes the two paradigms and how the conception of the role of administrative law is different under each.

A.     The Rational-Instrumental Paradigm and Outside-In Accountability

The rational-instrumental paradigm envisions a Weberian bureaucracy, which is expected to implement, but not to develop, government policy and values.[21]  For Weber, the essence of the bureaucratic organization was its capacity to rationally pursue its intended purposes, its “purposive-rationality.”[22] In public administration, the Weberian bureaucracy serves as a “transmission belt” for legislative decisions[23]—an instrument of the legislature whose task is strictly to obey the preordained democratic will as it is expressed in legislation.

The rational-instrumental paradigm looks to three institutional elements to limit the discretion of public administration.  First, legal frameworks of scientific and social-scientific methodologies are used to police administrative discretion.  These frameworks are understood to “guide discretion and allow it to be easily assessed.”[24]  The expectation is that by objectifying decision making, these methodologies “will act as a constraint on administrative discretion.”[25]

Second, a fair, pluralistic participatory process is employed to the extent that an agency confronts policy issues for which there are no objective resolutions because value choices are involved.[26]  The goal, however, remains the same—it is to eliminate administrative discretion.  As Fisher has pointed out, interest-group pluralism “is a way of gaining an account of the ‘will of the people’ and the role of the [agency] is simply to be an umpire overseeing the process.”[27]  As a surrogate legislative process, interest-group pluralism is “also a way of gaining a more accurate understanding of how a prescription should apply in a certain circumstance.”[28]  That is, like the employment of rational decision-making methodologies, public participation identifies appropriate resolutions of the questions presented, thereby removing agency discretion to decide on a resolution.

Third, strict political oversight and judicial review is used to connect the regulatory process to representative democracy.  While bureaucracy may be necessary to effectuate government, the paradigm relies on outside-in accountability to ensure discretion is controlled by democratic elements in the government or is controlled through processes that enforce that democratic will.

Political oversight reduces administrative discretion by giving elected leaders more influence over agency decision making, thereby establishing what Emmette Redford described many years ago as “overhead democracy.”[29]  Under this familiar concept, the public chooses its leaders in competitive elections; the leaders assume office with the power and responsibility to enact and execute policy, including overseeing the bureaucracy; successful leaders are rewarded with reelection, thereby ratifying their actions, including the oversight of agencies; and unsuccessful leaders are replaced.[30]

Judicial review reduces administrative discretion by verifying that public administration has in fact been an instrument of legislative will, as it is expressed in the agency’s mandate.  Judges determine whether an agency has chosen a policy that is within its legal discretion (defined in terms of analysis), whether the decision is a rational interpretation of its statutory mandate, and whether the agency has followed applicable administrative procedures.[31]  This last function is in aid of the rational-instrumental paradigm because it ensures that agencies cannot ignore their legal obligations to rationalize decision making, and, where this is not possible, judicial review ensures that agencies have also paid attention to interest-group pluralism.

B.     The Deliberative-Constitutive Paradigm and Inside-Out Accountability

The deliberative-constitutive paradigm rejects the basic premise of the rational-instrumental paradigm.  In recognition of the factual and normative complexities of administrative decision making, it grants to public administration substantial and ongoing problem-solving discretion.  Under this paradigm, legislation is understood to set out a series of general principles and parameters for the exercise of discretion.[32]  Public administration is therefore not an “agent” of the legislature but instead is an institution constituted by the legislature to use its best judgment.[33]

The paradigm accepts administrative discretion both as unavoidable and as necessary.  Discretion is unavoidable because the methodologies of the rational-instrumental model cannot eliminate it, as much as the proponents of the paradigm might try to do so.  Given this reality, the paradigm seeks to make a virtue of necessity.  It employs administrative expertise, deliberation, and reason giving to reach appropriate decisions.

Moreover, the paradigm rests on an understanding that legally imposed frameworks of scientific and social-scientific methodologies do not make decision making “objective” in practice.  These methodologies, particularly cost-benefit analysis, have not displaced the operation of politics.  They also lack accuracy and are subject to being manipulated according to an analyst’s policy preference.[34]  Moreover, claims about objectivity are simply untenable in light of post-empiricism.  Economics and other social sciences at best are a mixture of empirical data and social construction.[35]  Postempiricism acknowledges that science and social science can be important sources of knowledge, but the limits of these disciplines must be recognized.  Following on from this, experts are not limited to persons trained in scientific methodologies but include other professionals, particularly lawyers and public administrators, who rely on qualitative analysis to identify and justify regulatory solutions.  As discussed below, professionalism has an important role in the operation of expertise and the application of specialist knowledge.

This paradigm also understands its role regarding the public differently.  The job of public administration is not limited to aggregating the preferences of interest groups when normative issues present themselves.  Instead, as Brian Cook points out, public administration must be a “political institution” that “help[s] to create, to express, and to realize a nation’s public purposes.”[36]

Having constituted public administration as responsible for resolving discretionary issues, the paradigm looks to deliberation and reason giving as the modes of collective problem solving.  Deliberation is the means by which regulatory issues are defined, the relevance of information and expertise is established, and potential solutions are vetted.  It can involve a wide array of actors or a small group, depending on the problem at hand.

Unlike the reliance of social science methodologies on revealed preferences, this paradigm does not understand individual preferences to be given.  A deliberative dialogue is transformative in nature because different actors can learn from the process and reconsider their perspectives.  This approach also maintains that dialogue and deliberation must be “insulated from the mainstream political process, which is over-responsive to particular political interests.”[37]  Finally, public administration is responsible for informing and directing deliberations as well as for making the final decision.

The deliberative-constitutive paradigm manifests itself in terms of inside-out accountability.  Or, to put the matter another way, it looks to the elements of the paradigm—substantive expertise, deliberation, and reason giving—as ways of legitimizing public administration.  Thus, inside-out accountability legitimizes public administration by fulfilling the constituent role that public administration has been assigned.[38]

Inside-out accountability rests on the potential of organizational culture to promote among civil servants a mission orientation, a sense of public service, and professionalism.[39]  The first two norms establish and reinforce the “other-regarding” motives of civil servants, making them less likely to engage in self-interested behavior that sabotages the public interest mission of their agency.[40]  This reduces the risk that allowing for administrative discretion will reduce democratic responsiveness.[41]

The last norm, professionalism, creates administrative legitimacy by promoting neutral (as distinct from objective) expertise, in which scientists, lawyers, and other professionals present to political appointees the scientific, policy, and legal options relevant to the decisions that the administrators must make.[42]  It also means these civic servants will carry out the preferences of political appointees once these are made known.[43]

II.  Administrative Constitutionalism in Historical Perspective

We do not contend that the deliberative-constitutive paradigm and inside-out accountability are “the answer” but instead that there has been a failure to recognize the significance of this approach to addressing the contemporary legitimacy problems of American administrative law.  The recognition of both paradigms in administrative scholarship and practice would ensure that administrative constitutionalism reflects all of the different possibilities relating to the role and nature of public administration.  As a result, debates over accountability would become more nuanced and textured.[44]

For this purpose, we turn to a history of administrative constitutionalism, which reveals the influence of both paradigms on public administration.  This history may be familiar, but we briefly focus on it because it reveals an ongoing debate between the two paradigms.  As we noted earlier, administrative constitutionalism is a contested concept.  This has been true since the beginning of the administrative state, and it remains true today.

A.     In the Beginning

The earliest regulatory agencies adopted new policies through the use of administrative trials, which had most of the elements of a common law trial except a jury.[45]  The Interstate Commerce Commission (“ICC”), for example, used trial-like adjudicatory procedures to set railroad rates, and the Federal Trade Commission (“FTC”) used these procedures to determine if an unfair or deceptive trade practice had occurred.[46]  This makes the ICC and the FTC the earliest manifestations of the rational-instrumental paradigm, but it quickly became apparent that regulators were also required to exercise discretion grounded in experience and professionalism.  For example, in reviewing an ICC decision, Justice Holmes noted:

But the action does not appear to have been arbitrary except in the sense in which many honest and sensible judgments are so.  They express an intuition of experience which out-runs analysis and sums up many unnamed and tangled impressions,—impressions which may lie beneath consciousness without losing their worth.  The board was created for the purpose of using its judgment and its knowledge.[47]

B.     The Progressives

Despite reliance on administrative trials, the Progressive Movement sought to develop administrative institutions on the basis of the deliberative-constitutive paradigm.  In 1887, Woodrow Wilson proposed a “science of administration” that would operate “outside the proper sphere of politics.”[48]  We will have more to say about Wilson’s aspirations in Part IV.  For now, we note that the Pendleton Civic Service Act of 1883, an early Progressive victory, reflected the goal of a nonpartisan, expert bureaucracy.[49]

In the New Deal, the deliberative-constitutive paradigm had as much, or perhaps even more, influence than its rational-instrumental counterpart.  Legislation such as the National Industrial Recovery Act (“NIRA”) adopted procedures far less formal than the administrative trial model.[50]  New Dealers, such as Thurmond Arnold, justified the lack of administrative procedures on the ground that traditional procedural concepts were outmoded and counterproductive.[51]  They believed that “[m]odern regulatory statutes can provide no more than the skeleton, and must leave to administrative bodies the addition of flesh and blood necessary for a living body.”[52]

When progressives sought to reconcile democracy and expertise, they turned to deliberation as the methodology.[53]  John Dewey, for example, took his inspiration from the scientific method, with its emphasis on critical thinking, experimentation, and ongoing debate, and he argued that democratic process should reflect the same principles.[54]  Thus, while expertise was necessary to bring about social change, the expectation was that there would be an ongoing debate and consultation concerning the creation and implementation of policy.[55]  Following Dewey, Felix Frankfurter argued that commissions of inquiry were important vehicles for “defining issues, sifting evidence, posing problems and enlightening the public mind.”[56]

Similarly, the founders of the field of “public policy analysis,” Harold Lasswell, Yehezkel Dror, and others, were advocates of a deliberative-constitutive approach.[57]  They argued that analysis should study both how policy decisions are reached and what constitutes good public policy, with the goal of improving both process and substance.[58]  The analysis would be normative because the goal was to improve the practice of democracy and to better society, not merely to produce new knowledge.[59]  According to Dror, “[t]he main test of policy science is better policymaking, which produces better policies; these, in turn, are defined as policies which provide increased achievement of goals that are preferred after careful consideration.”[60]  As such, the intention was that policy science would inform both citizens and governmental officials, and would therefore serve as a bridge between professionals and democracy.[61]

C.     The Administrative Procedure Act

In 1947, Congress passed the current Administrative Procedure Act (“APA”),[62] which reflects both paradigms.  While the APA relies on legal procedures to check administrative power, it also reflects the Progressives’ understanding that rigid legal procedures slowed government action and were unnecessary.  The compromise can be seen in two key features of the APA.  First, the APA establishes rulemaking as an alternative to adjudication for many administrative decisions, permitting agencies to promulgate rules for entire industries or groups of regulated entities and to do so with a procedural framework designed to promote deliberation rather than simply constraining discretion.[63]  Second, the APA permits varying degrees of procedural formality for both adjudication and rulemaking.[64]  Further, formal trial-like procedures are only required for these functions if Congress requires such procedures in the agency’s statutory mandate.[65] For the most part, Congress has not required the use of formal procedures for either rulemaking or adjudication.[66]  Moreover, judicial review doctrine was reshaped on this basis—different grounds of review being deployed in relation to different rulemaking processes.[67]

D.    The Reformation

Between 1965 and 1975, public interest advocates were concerned that the many legislative victories they won in Congress would be lost in the halls of the agencies.[68]  Their concerns reflected a general disillusionment with the deliberative-constitutive paradigm that could be seen in reports and scholarship that highlighted problems of inefficiency, poorly trained personnel, and agency inertia.[69]  For the public interest movement, this suggested that administrative discretion invited corporate capture.[70]  The same conclusion was suggested by a number of revisionist histories, which argued that New Deal concepts of expertise merely shrouded agency capture and perpetuated maintenance of the status quo.[71]

The public interest movement turned to the courts to head off regulatory capture, and judges responded with the reformation, which empowered public interest groups to hold agencies accountable.[72]  What Stewart described as the “reformation” of American administrative law was part of this process.[73]  For their part, judges adopted the rational-instrumental understanding of interest representation pluralism.  The deliberative-constitutive model did indeed have its flaws; however, the response was not to recognize those flaws and see if they could be repaired but instead to shift understandings of the role and nature of public administration to the rational-instrumental paradigm.[74]  This was therefore an act of throwing the baby out with the dirty bathwater.

E.     Counterreformation

Beginning in 1980, the reformation was followed by what Shapiro has described as a “counter-reformation.”[75]  Once again, the goal was to reduce administrative discretion, but this time discretion was reduced using administrative, executive, and judicially imposed analytical requirements.[76]  Reformers pointed to a series of studies purporting to show that health, safety, and environmental regulations produced costs considerably in excess of regulatory benefits.[77]  They also believed that the government was overly responsive to irrational public demands for protection against health, safety, and environmental risks—demands that were inconsistent with expert judgments about the extent of those risks.[78]

Reflecting these understandings, the White House and Congress imposed numerous regulatory impact analysis requirements intended to require comprehensive and analytical decision making.[79]  In addition, the White House has escalated its control over agency government.[80]  It has sought to strengthen overhead democracy by increasing the number of political appointees in agencies.[81]  It has also required agencies to submit their significant rules and accompanying cost-benefit studies to the Office of Information and Regulatory Affairs (“OIRA”) for approval in an effort to reduce discretion.[82]

III.  The Unsustainable State of Contemporary Administrative Constitutionalism

There has been a decisive turn toward the rational-instrumental paradigm and outside-in accountability in the history of debates over administrative constitutionalism.  Despite its current popularity, this approach has left us with an understanding of administrative constitutionalism that is both incomplete and unsustainable.  For one thing, we can no longer depend, if we ever could, on interest-group pluralism to counter the disproportionate influence of regulatory entities over agency policy making.  This conclusion is highlighted by other articles in this special issue.

But this is not all.  The rational-instrumental paradigm has always been at odds with the regulatory and legislative reality of the administrative state.  Ignoring this dissonance weakens efforts to legitimize public administration, allowing for the demonization of bureaucrats and the defunding of agencies.  The turn toward the rational-instrumental paradigm has also resulted in a distinction being drawn between the scientific and participatory aspects of administrative decision making, a false dichotomy that has permitted the undermining of regulatory science.  Lastly, the preoccupation with the rational-instrumental paradigm has stunted thinking about how to develop inside-out accountability, a move that would allow us to offset some of the weaknesses of outside-in accountability.

A.     Rulemaking Without Pluralism

Empirical studies reveal that industry dominates the rulemaking process, both at agencies and at OIRA—a result that contradicts the premise of the reformation.  This dominance biases rulemaking and subverts judicial review.

1.     Rulemaking

Studies have found that business interests dominate rulemaking, whether measured by the number of rulemakings in which various interests filed comments or the relative number of comments that were filed in individual rulemakings.[83]  Wagner and her coauthors confirmed these results in a study of ninety hazardous air pollutant rulemakings at the Environmental Protection Agency (“EPA”).[84]  On average, industry filed over 81% of the comments submitted concerning a proposed rule;[85] public interest groups filed comments in less than 50% of the rulemakings;[86] and industry interests had an average of at least 170 times more communications with EPA staff (meetings, phone calls, letters, etc.) than did public interest groups during the period before the Notice of Proposed Rulemaking (pre-NPRM period).[87]

The same study found that the EPA mostly changed rules in the direction favored by industry.[88]  Other studies have found a similar result,[89] but there are also studies that have not found a connection between industry dominance and changes favorable to industry.[90]  Still, it is difficult to believe that business does not get an advantage from this asymmetry.  For years, knowledgeable observers have contended that asymmetrical information produces agency capture,[91] and this propensity is explained by the psychological literature.[92]

Furthermore, and even more ironically, the reformation has contributed to industry bias in rulemaking.  Because the courts expect an agency to respond to all significant comments,[93] the agency cannot “shield itself from this flood of information and focus on developing its own expert conception of the project.”[94]  This presents a problem of “filter failure” in which asymmetrical information overwhelms the agency and influences the outcome.[95]

2.     OIRA

OIRA oversight reinforces or even exacerbates existing pluralistic imbalances.  Rena Steinzor and her coauthors from the Center of Progressive Reform (“CPR”) found that 65% of the participants in 1080 meetings at OIRA were from industry interests, which was five times the number of attendees who represented public interest groups.[96]  An overwhelming number of the lawyers (nearly 95%), consultants, and lobbyists who attended these meetings represented business interests as compared to 2.5% who represented public interest groups.[97]  OIRA was also much more likely to meet alone with industry interests than with public interests.  Seventy-three percent of the more than 1000 meetings involved only industry interests, while a mere 7% involved only public interests.[98]

We know that OIRA habitually opposes stringent regulation,[99] although analysts have failed to provide empirical evidence linking this bias to industry dominance of White House meetings.[100]  This leads some analysts to doubt that the White House uses the review process to deliver benefits to powerful interests,[101] but we are less sanguine.  The countless meetings between industry interests and OIRA undoubtedly are about regulatory costs, and the public interest community lacks an equal opportunity to focus OIRA on regulatory benefits.[102]

3.     Judicial Review

Beyond the problem of biasing the agency, the lack of public interest involvement adversely affects judicial review, particularly when that involvement reflects the wide dominance of the rational-instrumental paradigm.  The reformation presumed public interest groups would put information in the record that supported the public’s interest in strong regulation, and that they could then hold an agency accountable in court if it ignored that evidence.[103]  Moreover, because an agency would know that it was vulnerable if it did not account for this evidence, a public interest group would have leverage to lobby the agency for stronger regulation.[104]  Obviously, both of these advantages are lost to the extent that no public interest group appears to file comments.[105]

Even if a public interest group files comments, it is still disadvantaged by the overwhelming number of industry comments.  Industry interests have been able to turn the requirement that agencies respond to all significant comments to their favor.  By filing so many comments, industry interests can use a blunderbuss attack during judicial review, which accuses the agency of neglecting their point of view because it failed to respond to some of the many comments that were filed.[106]  This gives rise to judicial review strategies driven by “analytical opportunism.”[107]

B.     Catch-22

An administrative constitutionalism rooted in the rational-instrumental paradigm cannot deliver true outside-in accountability if there is disproportionate industry influence, which there appears to be.  The rational-instrumental paradigm has another weakness that prevents it from successfully legitimizing agency government.  It aspires to provide regulatory government without bureaucratic discretion, although it cannot deliver on that promise.  Yet, without delivering on that promise, the paradigm cannot legitimize public administration.  This failure invites more efforts to control administration from the outside, which further slows the administrative process, prevents it from accomplishing its purposes, and brings little or no additional accountability.  This failure also leads to political demonizing of the bureaucracy, making it more difficult to recruit the professional expertise necessary to make public administration work.[108]

1.     Bracketing Discretion

Any realistic assessment demonstrates the inability of outside-in accountability to drain administrative discretion from the system.  Political oversight is unsystematic and ineffective, among numerous other weaknesses.[109]  As discussed earlier, analytical methodologies imposed from the outside are often unreliable in specific contexts and subject to manipulation.[110]  Judicial review is supposed to be deferential because unelected federal judges should not be influencing regulatory policies, an admonition that, if heeded, limits the extent to which judicial review can limit administrative discretion.  Judicial review has become less deferential since the reformation, but ultimately generalist judges are limited in second-guessing agency judgments because of institutional competence.[111]

In short, the rational-instrumental paradigm in practice brackets administrative discretion, reducing an agency’s discretion to some extent.  Clearly, agencies have less discretion than they would have otherwise if there were no presidential and legislative oversight and judicial review.  But, just as clearly, the rational-instrumental paradigm does not eliminate significant sources of agency discretion.  How much discretion exists varies from agency to agency, and even rule to rule, but no agency operates simply as a transmission belt for legislative directives.  The model has failed to fulfill its promise of largely discretion-free administrative law.

2.     Dismantling Administrative Capacity

As Abraham Maslow has pointed out, “it is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail.”[112]  Ignoring the potential of inside-out accountability, administrative reformers keep returning to the rational-instrumental paradigm to address the discretion that the previous set of reforms did not eliminate.  The result has not only led to the ossification of rulemaking,[113] but it has further weakened the legitimacy of public administration.  As agencies flounder in their efforts to address pressing regulatory problems, the perceived legitimacy of public administration declines even more.  After all, the legitimacy of the regulatory process depends not only on accountability and fairness, but whether agencies can efficiently carry out their statutory mandates—mandates that are the product of the democratic process.[114]

The reliance on the rational-instrumental paradigm has had another pernicious effect.  Because civil servants are understood merely as agents, and unreliable agents at that, they are understood to lack any inherent legitimacy.  This opens the door for the demonization of the bureaucracy by our political leaders.  John Kennedy, with this declaration, “[a]sk not what your country can do for you—ask what you can do for your country,”[115] is about the last political leader to consider government service as a noble calling.  Bureaucracy bashing in turn has discouraged professionals from joining the government and has made it difficult to retain them once hired.[116]  Moreover, even though government cannot be effective without professionalized administration, the political system regards civic servants with disdain, preventing efforts to reform the civic service system, to ensure adequate funding of agencies, and to take the other steps necessary to guarantee effective and efficient public administration.[117]

C.     The False Dichotomy

A further problem with the current state of affairs is that it leads to a false dichotomy.  Science and participation are understood to operate in separate institutional spheres rather than as closely interrelated constituent parts of public administration.  This is best seen in the context of risk regulation.

A common feature of health and safety legislation is a legislative authorization to act on the basis of anticipated harm, which makes scientific uncertainty an unavoidable aspect of regulatory science.[118]  In order for agencies to act in the face of uncertainty, agencies have come to evaluate the scientific evidence using certain overly formalized frameworks such as risk assessment, which are often portrayed as the result of expert judgment.[119]  This rational-instrumental paradigm approach to regulatory science is problematic in three ways.

First, particularly after the counterreformation, it encourages agencies to exaggerate the contributions made by science in resolving regulatory issues, a strategy that Wagner has characterized as a “science charade.”[120]  Because the goal of the rational-instrumental paradigm is to make agencies a transmission belt, it is in an administrator’s self-interest to claim that “science made me do it” as legal and political cover for a set of professional judgments.  This distorts an agency’s standard-setting mission and undermines transparency, driving up the costs of participation so that pluralism becomes even further beyond reach.  As Wagner notes, the “consequences range from administrative delays bordering on paralysis as experts debate incomplete science, to significant limitations on the ability of the public, the courts, and even public officials to participate in the policy choices embedded in scientific-sounding standards.”[121]

The failure to admit regulatory decisions are the product of professional judgment also opens the door for the “sound science” campaign of regulatory opponents.[122]  This campaign points to scientific uncertainty to contend that regulatory action is not based on “sound science,” but the real objection is with Congress’s decision not to wait for more definitive information about the extent of a risk before a regulatory agency can act to reduce that risk.[123]  Nevertheless, this argument persuades many due to the fact that public administration is not supposed to exercise the type of discretion that lies behind the use of regulatory science under the rational-instrumental paradigm.

The rational-instrumental paradigm also hides the efforts of regulatory opponents to bend science to their interests, particularly by manufacturing uncertainty.[124]  Regulated entities often point to scientific evidence they claim demonstrates that risks to people and the environment are significantly lower than agencies claim, and on several occasions have even commissioned research to raise doubts about the building scientific consensus.[125]  Regulated parties have also attacked research and researchers to undermine their credibility, sometimes in ways that are not scientifically credible but nevertheless prove effective in the political sphere.[126] Indeed, manufacturing uncertainty has been such a successful approach to the obstruction of protective regulations that the tobacco industry spearheaded an appropriations rider, the Data Quality Act, to provide a formal vehicle for stakeholders to challenge the reliability of research used by agencies at any point in the regulatory process.[127]

D.    Stunted Growth

Our last objection to the dominance of the rational-instrumental paradigm is that it has discouraged legal scholars from considering the potential of inside-out accountability to offset some of the weaknesses of outside-in accountability.  In 1903, Bruce Wyman made a distinction between external and internal administrative practice,[128] and, with some exceptions,[129] the “possibility that something is going on inside the agency that contributes to democratic accountability is never seriously considered.”[130]

The preoccupation with outside-in accountability has infected even the newest efforts to reform public administration.  Both civic republican and new governance reforms operate within the outside-in paradigm.[131]

1.     Civic Republican Reform

Legal scholars seeking to revive civic republicanism have recognized the significance of deliberation and reason giving for legitimating public administration, but they place civic republicanism in the context of outside-in accountability.  By portraying civic republicanism as a process that primarily engages those citizens who are not already engaged in civil service, legal scholars understand civic republicanism as a theory of democracy writ large rather than a theory of administrative constitutionalism.

Taking classic republicanism as their inspiration, these scholars look to make the administrative process more deliberative and less subject to pluralistic politics, looking to the courts to effectuate these changes.[132]  Even the most nuanced scholar advocating civic republicanism, Mark Seidenfeld, understands civic republicanism in this manner.  While he discusses the role of professionalism in public administration, he ultimately calls for heightened judicial review to turn rulemaking into a more civic republican process.[133]  Although Seidenfeld treats the transmission belt thesis as a myth, he does not trust professionalism to produce the civic republican dialogue that he seeks.  This places his efforts squarely within an outside-in approach.

No one argues that we can dispense with judicial review, but we need to consider that inside-out accountability can reduce the need for the type of outside-in judicial review that currently operates.  With this recognition, we can avoid the rulemaking ossification that comes with the aggressive judicial review of the type Seidenfeld supports.[134]

2.     New Governance Reform

This same complaint applies to the new governance scholars.[135]  In their search for alternatives to traditional standard setting, new governance scholars advocate other institutions that can create regulatory structures and foster a dialogue between stakeholders.  Stakeholder involvement is seen as crucial.  Orly Lobel, for example, notes new governance reforms are “based on engaging multiple actors and shifting citizens from passive to active roles,” thereby pluralizing the “exercise of normative authority.”[136]  It may be that new governance reformers can find policy networks in which all stakeholders are represented; the empirical evidence we reviewed earlier, however, suggests that the lack of pluralism may be a significant constraint on legitimizing alternative decision-making arrangements.[137]

Beyond this challenge, we see some new governance scholars as rejecting the deliberative-constitutive paradigm that we seek to promote alongside the rational-instrumental paradigm.  For them, the job of the bureaucracy is to steer policy networks towards solutions to regulatory problems,[138] making it an alternative version of the outside-in administrative pluralism endorsed by the reformation.

The new governance project has much to recommend it.  As compared to the reforms adopted in the counter-reformation, it recognizes and seeks to address the discretionary nature of public administration.  But, like the civic republican efforts, it cannot legitimize public administration because it distrusts, or at least fails to recognize, how the deliberative-constitutive paradigm contributes to administrative constitutionalism.[139]

IV.  Toward Enlightenment

We find ourselves stuck with an unsustainable understanding of administrative constitutionalism.  We therefore turn to a consideration of the potential of the deliberative-constitutive paradigm and inside-out accountability to foster administrative constitutionalism.  We begin by arguing that the Progressives were mostly correct about the potential for professionalism to contribute to the legitimacy of public administration.  We then consider how inside-out accountability contributes to legitimizing public administration.

We are not arguing that the deliberative-constitutive paradigm should replace the rational-instrumental paradigm.  Rather, our argument is that the enlightenment of American administrative law will only occur with the recognition of both paradigms and with a more wide-ranging understanding of public administration and administrative constitutionalism.  Indeed, the notion of “redundancy checks” is well known in engineering, particularly for potentially catastrophic technologies such as nuclear power plant safety.  We call for similar types of redundancy checks in administrative practice that utilize both outside-in and inside-out processes to enhance legitimacy and accountability.  Administrative law scholars and lawyers need to broaden their worldview.

A.     The Real Progressive Legacy

Doubts about Progressive claims for the legitimacy of public administration reflect two influences.  First, the Progressives were obviously mistaken that there could be a science of administration that operates outside the sphere of policy making.  Second, in light of public choice explanations of bureaucratic behavior, this failure makes it impossible to legitimize public administration from the outside-in.  These objections overlook a more complex argument made by Progressives on behalf of professionalism and accountability, one for which there is considerable evidence.

1.     Professionalism, Not Objectivity

The belief of the original Progressives that public administration would be democratic was based on three premises.[140]  Government employees would be hired on the basis of their technical skills and professional training.  Further, the political system would make the necessary policy choices when a law was passed, leaving it to the employees to use their training and expertise to find the best way to implement it.  Lastly, employees would refrain from making political decisions because of their training, personal preferences, and professional ethics.

The Progressives soon recognized, however, the difficulty with the second premise that there could be an objective science of administration.  Herbert Croly, for example, maintained that while an expert administrator must be a “social expert,”[141] expertise needed to “be kept articulate with the democracy.”[142]  He and other Progressives believed that this could be done when civic servants operated within the bounds of professionalism, the third assumption of the Progressive legacy.

The public choice challenge to inside-out accountability ignores the potential of professionalization and organizational culture to promote other-regarding behavior.  Professionals are trained to evaluate information on the basis of standards of evaluation external to the agency.  In doing so, they utilize the methodology that they have been trained and socialized to employ.  Peers reinforce this behavior by approving of those who follow it and distancing themselves from those who do not.[143]

When professionals act in this manner, they provide a balanced picture of information as part of professional behavior.[144]  This means, for example, that lawyers present to agency administrators the information necessary to decide on a course of action.  Likewise, scientists present an impartial reading of the evidence available to them because their professional training and self-identity dictate this behavior.

As noted earlier, the deliberative-constitutive paradigm took a beating in the 1970s from both the political left and right,[145] but contemporary understandings of professionalism indicate that this complete distrust was misplaced.  Lamont’s powerful study of interdisciplinary academic funding panels in the United States is a case in point.[146]  Lamont highlights the roles of expertise, preparation, discourse, socialization, and pragmatism in this context.  Her ethnography is not to romanticize the role of professionalism in academia but to show how it produces workable results.[147]  Sennett’s recent works on socialized expertise[148] and dialogic cooperation[149] furnish additional examples of the significance and reliability of professionalism.  In the more specific area of risk regulation, the National Research Council’s discussion of “analytical-deliberative” approaches to decision making also provides an important blueprint.[150]

Much of the academic work on professionalism can be found in the public administration literature, which puts professionalism at the center of its traditional concept of the civil service.[151]  We now turn to the evidence about professionalism in that literature, which indicates the potential of professionalism in agencies that are properly managed and have a culture of professionalism.

2.     The Evidence

Herbert Kaufman’s study of the United States Forest Service remains the classic study of how professionalism defeats self-interest.[152]  At the time of Kaufman’s study, which was before modern advances in communication, the far-flung physical locations of Forest Service officers made it difficult to monitor and direct them.  Concerned that this would result in inconsistent policies and possible corruption, Gifford Pinchot, the first director of the Forest Service, set out to build an organization composed of professional foresters, using various methodologies.[153]  These efforts produced what Kaufman characterized as “voluntary conformity” to the goals and public purposes of the Forest Service.[154]  Kaufman found “almost no charges of administrative sabotage by frustrated leaders, for example; comparatively few accusations of local favoritism and discrimination by the clientele of the national forests; [and] no discoveries by Congressional investigators of scandalous field collusion with special interests.”[155]

More recent work corroborates Kaufman’s findings.  John Brehm and Scott Gates, for example, compared surveys of government and local employees that indicate how the employees view their own behavior and how it is viewed by their fellow workers, supervisors, and outside persons.[156]  The data consistently indicated “bureaucrats devote the majority of their time to working, rather than to shirking or to sabotage . . . .”[157]  In other words, Brehm and Gates found no evidence that bureaucrats exercised their self-interest by failing to work (shirking) or by working to defeat the policies of their political bosses (sabotage).  Brehm and Gates attributed this behavior to professional influences because they found that supervisors had little ability to influence behavior by financial rewards or the threat to fire employees.[158]  In light of these constraints, they believed the results were attributable to rewards such as “recognition from others, accomplishing worthwhile things, [and] serving the public interest . . . .”[159]

A similar result is found in Marissa Golden’s study of how upper-level civil servants in four agencies responded to the election of Ronald Reagan.[160]  The situation provided a good test of the reliability of government employees because, in each of the agencies Golden studied, the Reagan administration attempted to “turn agency policy 180 degrees from its past.”[161]  Nevertheless, Golden found that “career civil servants were, for the most part, responsive to this change in elected leadership.”[162]  Golden attributes this loyalty in part to internal hierarchical controls, noting that civic servants “did not want to be demoted or banished, and sought to advance their careers.”[163]  She also found, however, that civic servants understood that their role was to present information to political appointees and help them decide how best to carry out the President’s policy preferences.[164]

B.     Democratic Legitimacy

Various arguments have been made for how the deliberative-constitutive paradigm burnishes the democratic legitimacy of the bureaucracy.  The boldest claim is that historical documents relating to the framing of the Constitution indicate that the framers anticipated that administrators would implement government, giving administration a constitutional legitimacy.[165]  More modest claims start with the progressive formulation that professionalization makes the bureaucracy a reliable servant of political administrators for the reasons expressed in the last section.  Public administration, however, is also understood to add positive democratic value.

There are different perspectives concerning how bureaucracy adds to democratic legitimacy.  Some public administration scholars see this effort as communitarian, in which administrators facilitate purposive action among their fellow citizens.[166]  Other scholars advocate that the bureaucracy can represent their fellow citizens if the public service reflects different populations in the community and presents their point of view.[167]  Still others believe that the bureaucracy serves democracy when it speaks truth to power.  This occurs when professionals challenge political appointees by pointing out how policies that they favor are inconsistent with scientific and policy evidence.[168]  While ultimately civic servants will defer to political appointees, this role supports democratic legitimacy by warning administrators that they may not be faithfully executing the statutes they administer.  Finally, another group of scholars sees the role of bureaucracy as discursive, with administrators reaching out to individuals and organizations to dialogue about solutions for public problems.[169]

Whatever the merits of the other approaches in other contexts, we see the last two ideas as being best suited for rulemaking.  The role of speaking truth to power takes advantage of having a professionalized staff.[170]  Rulemaking, especially for health, safety, and environmental issues, relies on scientific, economic, and engineering data.  This means civic servants are in a position to be honest brokers concerning the implications of this evidence.  Moreover, even as to nontechnical issues, civic servants are in a position to speak truth to power.  As long-time employees, they constitute the institutional memory of the agency, particularly in a world in which political administrators come and go, sometimes as often as every few months.  As the repository of the institutional memory, civic servants can bring that wisdom into the rulemaking process, for example, pointing out approaches than have or have not worked in the past.

Proponents of a discursive role for civil servants contend that the discursive process itself legitimizes the outcome of the process through debate and deliberation.  For postmodernists, it is only possible to construct a legitimate policy through such vetting.[171]  But it is not necessary to endorse this viewpoint to understand the value of a discursive approach.  With the demise of pluralism in rulemaking, the bureaucracy can offset or mitigate industry dominance by reaching out to individuals and organizations in policy networks with differing points of view, rather than relying passively on whatever information comes in through the rulemaking process.  Professional values serve to further enhance the assimilation and processing of this evidence into regulatory products.

The type of dialogue that this process creates addresses the limitations of interest-group pluralism discussed earlier.  An agency would have the advice and insight of multiple perspectives rather than having input only from business interests.  If a proposed rule results from a thorough discursive process, the agency is also less likely to fall prey to filter failure, in which it is so busy responding to comments filed by business interests that it loses sight of the bigger picture.

V.  An Example from U.S. Practice

An example of a regulatory approach that epitomizes these deliberative, inside-out characteristics helps crystallize this more abstract discussion.  In this Part, we turn to a high profile rulemaking process used in the United States to set national ambient air quality standards (“NAAQS”).  This standard-setting is highly science intensive, but it also has extraordinary policy costs and consequences.[172]  Because of its social significance, the EPA, which sets these standards, has been under intense public, political, congressional, and judicial pressure over the last four decades concerning each of its proposed revisions.[173]  The standard-setting process emerging from these heated battles, in our view, exemplifies what the deliberative-constitutive paradigm can contribute to administrative constitutionalism.

A.     The NAAQS Process

Section 109 of the Clean Air Act requires the EPA to review, at five-year intervals, the standards for six criteria or general pollutants that the EPA identifies under Section 108 of the Act.[174]  These NAAQS must be set at a level “requisite to protect the public health” with “an adequate margin of safety.”[175]  While the Supreme Court interprets the statute as allowing the EPA to consider only scientific, not economic, factors in setting the primary health standards for these criteria pollutants,[176] there is still a great deal of room for technical maneuvering within scientifically plausible options.  Indeed, the remaining agency discretion is so great that a majority in one D.C. Circuit judgment concluded that this section of the Act violated the nondelegation doctrine of the Constitution.[177]

Given their national role in specifying the lowest acceptable air quality for any region in the United States, the selection of the NAAQS has significant regulatory consequences.  Elaborate regulatory permits and state implementation plans ensure that these specific NAAQS are met.[178]  Billions of dollars, both in health protection and compliance costs, hinge on adjustments as small as even one-thousandth of a part per million for the standard for any given criteria pollutant.[179] Presidential elections also can turn on, or at least be affected by, an administration’s decision to make NAAQS more stringent or not.[180]

1.     The Old Way

Until a recent change of direction, the EPA’s approach to setting the NAAQS largely followed the rational-instrumental paradigm.  From the 1980s to mid-2005, the EPA produced assessments that grew increasingly voluminous and were considered relatively impenetrable to anyone other than air quality experts, and even these experts were challenged by the document.[181]  Producing these reports was so unwieldy that the EPA itself could not complete them in a five-year time frame as required by statute and was perpetually at risk of being in contempt of court.[182]  The agency also found itself under constant attack for the judgments reached in its decisions.[183]

2.     The New Way

This all changed in 2006 when the EPA redesigned the NAAQS process.  The current approach involves five separate analytical steps and products.

a.  The Planning Report

The first step sets the stage for the integration of scientists, stakeholders, public health advocates, and professional agency staff by convening a “kick-off” workshop[184] that is followed by a staff-authored report that articulates the overarching policy questions that will guide the process.[185]  The report is reviewed by the “Clean Air Science Advisory Committee” (“CASAC”), a statutorily required standing committee of top scientists chartered under the Federal Advisory Committee Act (“FACA”), and by the public before it is final.[186]

The resulting final planning report is thus a professional, staff-authored document that has been reviewed iteratively by the public and external scientists.[187]  This planning report, moreover, is integral to enhancing transparency of the NAAQS review process.[188]  By framing the relevant science-policy questions, the planning report focuses the EPA’s subsequent NAAQS review, which stretches over a four-year process.[189]

b.  Integrated Scientific Assessment Report

At the next step of the NAAQS review process, the EPA compiles an integrated scientific assessment (“ISA”) that reviews all of the scientific evidence.[190]  In stark contrast to the EPA’s earlier version of this assessment in previous NAAQS processes, the new and improved ISA is more concise and focuses the assessment on the specific questions framed in the planning report.  More detailed information is reserved for annexes, which can sometimes be longer than the body of the report itself.

The document is prepared in a way that is roughly equivalent to a large team-authored scientific review paper.  Academics generally are contracted to draft the individual chapters of the ISA, with multiple points of review (at least three) from intra-agency reviewers, CASAC, and the public before the ISA is considered final.[191]  Like other NAAQS documents, the ISA includes a detailed list of the EPA executive staff, authors, contributors, and peer reviewers.[192]  Authors and peer reviewers outside the agency are also listed by name and affiliation in the front matter.[193]  Staff members who disagree with the scientific analysis may remove their names.  At the same time, those who agree are held accountable for the contents.[194]

c.  Risk/Exposure Assessment Report

Based on the analysis of the scientific evidence in the ISA, the EPA staff then prepares a separate risk assessment report that applies this evidence to predict the effects of alternate standards on public health.  The goal at this stage is to employ multiple models to produce quantitative risk estimates, accompanied by expressions of the underlying uncertainties and variability for various endpoints, such as the impacts of a pollutant on susceptible populations and ecosystems.[195]  The risk assessment process itself begins with a planning/scoping stage, which again involves CASAC review and public comment, followed by two more periods of intra-agency, CASAC, and public comment on the draft risk assessment reports.[196]

d.  Policy Assessment Report

The last document in the process is a policy assessment that “bridges” these more science-intensive (ISA and risk assessment) reports with the policy questions at hand.  In summarizing the evidence in a way that relates to the overarching policy question, the report offers alternative health protection scenarios and standards, accompanied by discussions of unknowns and uncertainties.  The policy analysis also identifies questions for further research.  The policy assessment is, in and of itself, an extensive document (in the EPA’s review of the particulate matter standard, the policy assessment was over 450 pages in length, including appendices),[197]but the discussion is written for laypersons who do not have an extensive background in the relevant science.

The policy assessment is reviewed by internal EPA staff and by CASAC, sometimes several times, to ensure that important scientific information is not lost in translation.[198]  It is worth noting that even at this late stage, CASAC review and comment is rigorous and extensive.  For example, the second CASAC review of the EPA’s Policy Assessment for the Review of the Particulate Matter (“PM”) NAAQS consisted of over seventy pages of single-spaced comments.[199]

e.  The Proposed and Final Rulemaking Process

Based on this wealth of deliberative science-policy work, the EPA management identifies a standard and prepares a proposed rule that is cleared through the Office of Management and Budget (“OMB”) and then published in the Federal Register.  At this point, the outside-in model kicks in.  Stakeholders appreciate that if they wish to preserve their challenges for judicial review, they must submit comments that raise every issue of concern.[200]  After notice and comment and further inter-governmental deliberations, the EPA promulgates a final rule.

B.     The Deliberative-Constitutive Elements

The NAAQS review process exemplifies how the deliberative-constitutive process can work to employ professionalism in the pursuit of democratic accountability from the inside-out.  We see five aspects of the NAAQS process that not only illustrate this potential but suggest a pathway for bolstering the deliberative-constitutive approach in other regulatory contexts.

1.     The Process is Professional

Perhaps the most critical aspect of the EPA’s process is the role that the agency’s professional staff plays in this deliberative exercise.  Professional EPA experts and academics, not agency managers, author the reports.  While EPA management is briefed through information sessions on the contents of these reports, there is no editing of the report by management.[201]  Indeed, at least some of the draft NAAQS reports contain the disclaimer that the “opinions, findings, conclusions, or recommendations” reflect those of the authors and do not necessarily represent the views of the EPA.[202]  The team of staff authors is also generally listed by name in the acknowledgments section of the final report, and their names are linked to specific contributions in individual chapters.[203]

The use of staff scientists to prepare all the foundational assessments underlying the NAAQS review is a deliberate feature in the design of the process.  For a short time, responsibility for authorship of the policy assessment was shifted from EPA staff to management and published as an Advanced Notice of Proposed Rulemaking (“ANPR”).[204]  The management-drafted policy assessment was harshly criticized by both the EPA’s Office of Research Development and by CASAC.[205]  CASAC, in particular, noted that the management-drafted policy assessment departed from the scientific recommendations of agency scientists, did not connect the policy options suggested to the scientific evidence, and presented options as equally plausible, despite their very different scientific underpinnings.[206]  In response to this controversy, Administrator Lisa Jackson ultimately returned responsibility of the policy assessment to the EPA professional staff.[207]  Employees and CASAC report a high level of satisfaction with this change.[208]

Agency staff authorship is valuable for several reasons.  Not only does authorship provide agency staff with well-deserved credit for their work, but it creates accountability for the quality of the final technical product.  Those who are part of a consensus report take responsibility for the contents.  Respected scientists within and outside the agency also scrutinize the staff’s work.  This scientific oversight further enhances the staff’s commitment to professionalism and rewards quality.  Authorship also sharpens internal discussions, ensures that revisions are generally on the merits rather than politically convenient, and even provides signals to the outside world of the backgrounds of the individuals doing these important analyses.  Finally, authorship helps separate the report—figuratively and literally—from management, an important attribute we consider again below.

This staff authorship is easily adapted to other rulemakings, particularly those that are initiated with an initial staff scientific report.  Even the preamble of a proposed rule could provide attribution or acknowledgments, however.  Regardless of its form, by providing this attribution, the agency affords the staff not only well-deserved credit but provides the staff the opportunity to opt out or even dissent from an agency’s public scientific analysis.

2.     The Process is Iterative and Discursive

The agency’s interactions with experts, the public, and other technical staff throughout the NAAQS process are iterative and discursive.  Four separate staff reports are reviewed multiple times—a total of at least seven back-and-forths with the public, agency staff, and CASAC.[209]  The comments are logged into the public record, as are the EPA’s responses to comments.  The staff also works closely with CASAC, which enhances the scientific rigor and credibility of the report.  The resulting interactions lead to a scientifically respected process for an otherwise very controversial and socially important standard.[210]

Each of the reports in the NAAQS process is also prepared with the goal of communicating the findings in a way that is accessible and clear, which further advances the deliberative quality of the process.  Reports that are succinct and accessible invite a wide range of participants into the deliberative process, which, in turn, provides a powerful method of accountability.[211]  The success of the NAAQS process in producing reports that are in fact succinct and accessible was spotlighted in a National Academy of Sciences report that identified the NAAQS reviews as a model with respect to providing a sophisticated, yet cogent, review and interpretation of the available evidence and models.[212]

While such an elaborate process is not possible in most rulemakings because of limited time and resources, many of the iterative features of the NAAQS review process can nevertheless be adapted to other rulemaking settings.  For example, a kick-off workshop that identifies the policy questions and surveys the available evidence is a relatively common feature of a number of rulemakings.[213]  Separating out these processes may add only a few weeks at most to a rulemaking exercise, but the value of doing so is considerable since it allows the agency to dedicate a separate, deliberative step to what is often the most important part of the rulemaking exercise—framing the assignment.  Subsequent, iterative stages of feedback in the NAAQS process could be collapsed by making initial reports publicly accessible, without a formal comment period, so that comments can be wrapped into the later public comment process.  A science advisory board might even be involved early and at multiple points in the evolution of a proposed rule to ensure that the end product is as technically accurate as possible.

3.     Inside-Out Accountability

Beyond the professional and deliberative qualities of the NAAQS review process is its relative insulation from both the political process and aggressive interest representation driven judicial review, a feature that is consistent with inside-out accountability.  This insulation derives in part from the fact that the iterative reports precede the separate, proposed rule process and also by concerted efforts by agency staff to keep staff and management roles separate and distinct.  The goal in all four reports is to characterize the scientific record and policy options as clearly as possible.

Political managers are involved in this scientific phase only via informational briefing sessions, and OMB is not involved at all in any of these four reports.[214]  Although this insulation, combined with the commitment to staff authorship and professional quality, cannot ensure “neutral” advice, the NAAQS review process simulates scientific review and thus comes about as close to neutrality as a regulatory process can.  Later on, at the proposed rule stage, management will struggle publicly with how to portray this information to justify its decision.

4.     Support of Outside-In Accountability

The final policy assessment, which bridges the scientific findings with the policy questions, begins a shift to outside-in accountability.  With the basic research and assumptions laid bare, stakeholders and others can then engage in the process in a more meaningful way.  Indeed, one can view the NAAQS review process as essentially beginning with an inside-out process, embodied in the four reports, that then switches over to the outside-in process at the time the agency begins drafting its proposed rule.  At this point, familiar pluralistic oversight takes over.  OMB will engage with the EPA in identifying and supporting a preferred standard.  Interest groups will fill the record with comments that serve as placeholders for possible litigation.  And ultimately, the courts may review the agency’s work based on the record the EPA has created.

This outside-in phase is constrained, potentially significantly, by the rigorous deliberative record prepared by staff, however.  Stakeholders may find that the kinds of criticisms and alternatives they can offer are much more limited because of the extensive analysis that forms the basis for the NAAQS proposal.

The credible threat of judicial review is also disciplined by the types of professional discussions documented in the record.  Specifically, courts may be inclined to defer to the careful analysis embodied in the four reports.  Thus, a rule that goes through an inside-out process before it is subjected to stakeholder comments and judicial review may be at serious risk of reversal only when the agency’s final rule deviates in material ways from the building scientific consensus captured in its administrative record.[215]

5.     Speaking Truth to Power

In recent years, two Presidents have actually set aside the agency’s robustly recommended NAAQS standards.  Presidents George W. Bush and Barack Obama both decided to effectively set aside the recommendations of CASAC and the EPA’s scientific staff with regard to setting revised standards for the NAAQS.[216]  It should be further noted that both of these presidential overrides were conducted on scientific recommendations that resulted from the EPA’s deployment of the new and improved NAAQS process.[217]

Ironically, however, this trumping of the work of agency professionals may be evidence of the great success of the inside-out approach.  Because the agency’s analyses were effectively ironclad, the dissenting Presidents could not pretend that “the science made me do it.”  Nor could they suggest the science was done improperly.  Instead the Presidents’ decisions conceded that the more stringent air quality standards were simply too costly.[218]  The resulting policy decisions were exposed for all to see in large part because the scientific deliberations were so complete and well documented.  Ironically, recognizing that there is a close interrelationship between science and democracy provides a more explicit and robust framework for the scientific and policy basis of decisions.

Agency staff members may prefer that their recommendations be followed, but a rigorous inside-out process ensures that when this does not occur, the decision—influenced by politics, courts, and other “outside” factors—is made against the backdrop of a professionally compiled scientific and policy record.  The role of the agency in this way is not so much to identify the answer as to ensure transparency in putting that answer in context so that the decision makers can be held accountable.  Seen in this way, the role of the Fourth Branch is not to govern so much as to use deliberative, public administrative processes to inform and hold the decision makers publicly responsible for the choices they make.

Conclusion: The Enlightenment of Administrative Law?

We have sought to explain why the current scholarly and institutional vision of American administrative law scholars and lawyers has become too narrow.  We are not pursuing some simple set of adjustments, but rather we are maintaining that there is another way to think about the role and nature of legitimate public administration that produces workable models of administrative action.  In so doing, we are attempting to open up a new way for contemporary scholars to engage with and think about public administration, following in the footsteps of early scholars.

Recognition of the importance of administrative constitutionalism also provides greater possibility for addressing the current legitimacy issues of public administration.  We are not shackled to a model that has been shown to have serious flaws.  Indeed, as our analysis shows, there is much to learn from history.

Overall, we are also not arguing for the replacement of one understanding of administrative constitutionalism with another.  We are arguing for a new stage in administrative law scholarship in which there is a commitment to broadening intellectual engagement and debate and a willingness to employ multiple models of administrative accountability simultaneously.  We liken this to an “enlightenment” in which a commitment to reason flourishes, as does a simultaneous pursuit of scientific and social inquiry.  In saying this, we recognize the complexity of any “enlightenment process,”[219] but it is exactly that complexity on which administrative law scholars should be focusing.

        *   University Chair in Law, Wake Forest University.

        **   Reader in Environmental Law, Corpus Christi College, UL Lecturer, Faculty of Law, Oxford University.

        ***   Worsham Centennial Professor of Law, University of Texas.

        [1].   Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 167 (1962).

        [2].   See generally Brian J. Cook, Bureaucracy and Self-Government: Reconsidering the Role of Public Administration in American Politics (1996); James O. Freedman, Crisis and Legitimacy: The Administrative Process and American Government (1978); John A. Rohr, To Run a Constitution: The Legitimacy of the Administrative State (1986); Stephen Skowronek, Building a New American State: The Expansion of National Administrative Capacities 1877–1920 (1982).

        [3].   See generally Gerald E. Frug, The Ideology of Bureaucracy in American Law, 97 Harv. L. Rev. 1276 (1984) (describing four theories that fail to overcome the problems of managerial domination and personal alienation in corporate and administrative agencies).

        [4].   Richard B. Stewart, The Reformation of American Administrative Law, 88 Harv. L. Rev. 1667, 1669 (1975).

        [5].   Id. at 1670.

        [6].   Id. at 1760.

        [7].   See infra note 83 and accompanying text.

        [8].   Elizabeth Fisher, Risk Regulation and Administrative Constitutionalism 27 (2007).

        [9].   See infra note 113 and accompanying text.

      [10].   See infra note 108 and accompanying text.

      [11].   See infra notes 118 and 120 and accompanying text.

      [12].   Bruce Ackerman, The New Separation of Powers, 113 Harv. L. Rev. 633, 696 (2000).

      [13].   Fisher, supra note 8, at 30.  For a similar discussion, see also Carol Harlow & Richard Rawlings, Law and Administration ch.1 (2009).

      [14].   Sidney A. Shapiro, Pragmatic Administrative Law, in Issues in Legal Scholarship: The Reformation of American Administrative Law 2, 3 (2005) [hereinafter Shapiro, Pragmatic Administrative Law], available at .

      [15].   Fisher, supra note 8, at 3.  Eskridge and Ferejohn also refer to “administrative constitutionalism,” but they are referring to the body of fundamental public law principles developed by the legislature and executive.  William N. Eskridge Jr. & John Ferejohn, A Republic of Statutes: The New American Constitution 24–26 (2010).  While we understand our concept and theirs to be overlapping, our focus is upon the constituting and limiting of administrative institutions.  For further discussion, see Fisher, supra note 8, at ch. 1.

      [16].   Jürgen Habermas, Communication and the Evolution of Society 178–79 (Thomas McCarthy trans., 1979).

      [17].   See Thomas C. Grey, Constitutionalism: An Analytical Framework, in Constitutionalism: Nomos XX, at 189 (J. Roland Pennock & John W. Chapman eds., 1979) (noting that constitutionalism is a “cloudy” concept).

      [18].   W.B. Gallie, Essentially Contested Concepts, 56 Proc. of the Aristotelian Soc’y 167, 184 (1956); Jeremy Waldron, Is the Rule of Law an Essentially Contested Concept (in Florida)?, 21 L. & Phil. 137, 139–40 (2002).

      [19].   Jerry L. Mashaw, Greed, Chaos, and Governance: Using Public Choice to Improve Public Law 1 (1997).  Paradigms “give us mental images of what to look for in political life and what to expect from it . . . .  [T]he influence of pictures or themes is not just on what we expect and what we see, but also on what we demand or affirm.”  Id.

      [20].   Elements of the rational-instrumental paradigm and outside-in accountability can be found in various accounts of the administrative process.  See, e.g.,Cook, supra note 2, at 4–5 (instrumental model); Thomas O. McGarity, Reinventing Rationality: The Role of Regulatory Analysis in the Federal Bureaucracy 3–16 (1991) (comprehensive analytical rationality); Martin Shapiro, The Supreme Court and Administrative Agencies 67–91 (1968) (synoptic model); Stewart, supra note 4, at 1779 (interest-group pluralism).

      [21].   Max Weber, Economy and Society 957–59 (Bryan S. Turner ed., 1991).

      [22].   See Max Weber, From Max Weber: Essays in Sociology ch. 7 (1991).

      [23].   See Stewart, supra note 4, at 1675.

      [24].   Fisher, supra note 8, at 28.

      [25].   Id.

      [26].   Id. at 28–29.

      [27].   Id. at 29.

      [28].   Id.

      [29].   Emmette S. Redford, Democracy in the Administrative State 70 (Roscoe C. Martin ed., 1969).

      [30].   See Kenneth J. Meier & Laurence J. O’Toole, Jr., Bureaucracy in a Democratic State: A Governance Perspective 6 (2006) (describing the operation of overhead democracy).

      [31].   Sidney A. Shapiro & Ronald F. Wright, The Future of the Administrative Presidency: Turning Administrative Law Inside-Out, 65 U. Miami L. Rev. 577, 584 (2011).

      [32].   Fisher, supra note 8, at 30.

      [33].    Id.

      [34].   See Sidney A. Shapiro & Christopher H. Schroeder, Beyond Cost-Benefit Analysis: A Pragmatic Reorientation, 32 Harv. Envtl. L. Rev. 433, 450–59 (2008).

      [35].   Id. at 460–62.

      [36].   Cook, supra note 2, at 16.

      [37].   Fisher, supra note 8, at 31.

      [38].   Id.

      [39].   Barbara S. Romzek & Melvin J. Dubnick, Accountability in the Public Sector: Lessons from the Challenger Tragedy, 47 Pub. Admin. Rev. 227, 229 (1987).

      [40].   See infra notes 152–64 and accompanying text.

      [41].   See infra notes 152–64 and accompanying text.

      [42].   See infra notes 140–51 and accompanying text.

      [43].   See infra notes 158–62 and accompanying text.

      [44].   Elizabeth Fisher, The European Union in the Age of Accountability, 24 Oxford J. Legal Stud. 495, 497–98 (2004).

      [45].   Richard E. Levy & Sidney A. Shapiro, Administrative Procedure and the Decline of the Trial, 51 U. Kan. L. Rev. 473, 477 (2003).

      [46].    See id. at 475.

      [47].   Chicago, Burlington, & Quincy Ry. Co. v. Babcock, 204 U.S. 585, 598 (1907).

      [48].   Woodrow Wilson, The Study of Administration, 56 Pol. Sci. Q. 481, 494 (1941).

      [49].   Ari Hoogenboom, The Pendleton Act and the Civil Service, 64 Am. Hist. Rev. 301, 315–16 (1959).

      [50].   Nicholas S. Zeppos, The Legal Profession and the Development of Administrative Law, 72 Chi.-Kent L. Rev. 1119, 1126 (1997).

      [51].   Thurman W. Arnold, The Folklore of Capitalism 372–75 (1937).

      [52].   Walter F. Dodd, Administrative Agencies as Legislators and Judges, 25 A.B.A. J. 923, 925 (1939).

      [53].   Herbert Croly, Progressive Democracy 400 (1914); Walter Lippmann, Drift and Mastery: An Attempt to Diagnose the Current Unrest, at xviii–xix (1914);see also Charles Forcey, The Crossroads of Liberalism: Croly, Weyl, Lippmann, and the Progressive Era 1900–1925, at 156 (1961), available at

      [54].   Joseph G. Metz, Democracy and the Scientific Method in the Philosophy of John Dewey, 31 Rev. Pol. 242, 242 (1969).

      [55].   Shapiro, Pragmatic Administrative Law, supra note 14, at 4.

      [56].   Felix Frankfurter, The Public and Its Government 162–63 (1930).

      [57].   See, e.g., Yehezkel Dror, Design for Policy Sciences 51 (1971); Harold D. Lasswell, A Pre-View of Policy Sciences 1 (1971).

      [58].   Peter deLeon, Advice and Consent: The Development of the Policy Sciences 29 (1988) (noting that Lasswell and others were interested in both “knowledgeof and in the policy process”).

      [59].   Lasswell, supra note 57, at 3–4.

      [60].   Dror, supra note 57, at 51.

      [61].   Shapiro & Schroeder, supra note 34, at 438.

      [62].   5 U.S.C. §§ 500–596 (2006).

      [63].   Id. § 553.

      [64].   Id. §§ 553–554.

      [65].   Id. § 554(a).

      [66].   Id. §§ 553–554.

      [67].   J. Skelly Wright, The Courts and the Rulemaking Process: The Limits of Judicial Review, 59 Cornell L. Rev. 375, 382–84 (1974).

      [68].   See Michael Pertschuk, Revolt Against Regulation: The Rise and Pause of the Consumer Movement 5 (1982) (noting that Congress passed twenty-five laws regulating the business community between 1967 and 1973).

      [69].   See, e.g., Am. Bar Ass’n Comm’n to Study the Fed. Trade Comm’n, Report of the ABA Commission to Study the Federal Trade Commission 1 (1969) (claiming the FTC had failed to achieve the goals of its designers); Kenneth Culp Davis, Discretionary Justice: A Preliminary Inquiry 27, 54–55 (1969) (arguing that unchecked and uncontrolled discretion was not desirable and that there was a greater need for clear rules); Henry J. Friendly, The Federal Administrative Agencies: The Need for Better Definition of Standards 2 (1962) (characterizing federal regulation as a “Serbonian bog”); James Landis, Report on Regulatory Agencies to the President-Elect 11–15 (1960) (discussing problems of unethical conduct and institutional inertia and touching on problems of delay and the deterioration in the quality of personnel).

      [70].   See Michael W. McCann, Taking Reform Seriously: Perspectives on Public Interest Liberalism 44 (1986) (discussing the distrust of discretion in the public interest movement).

      [71].   See, e.g., Marver H. Bernstein, Regulating Business By Independent Commission 270–71 (1955); Gabriel Kolko, Railroads and Regulation: 1877–1916, at 2 (1965).

      [72].   Sidney A. Shapiro, Administrative Law After the Counter-Reformation: Restoring Faith in Pragmatic Government, 48 U. Kan. L. Rev. 689, 694–95 (2000) [hereinafter Shapiro, Counter-Reformation].

      [73].   Id. at 692.

      [74].   Id. at 706.

      [75].   Id. at 697.

      [76].   Id.

      [77].   Id. at 697–98.

      [78].   See Cass R. Sunstein, Law of Fear: Beyond the Precautionary Principle 126–27 (2005).

      [79].   See McGarity, supra note 20, at xiv–xv, 19–21.  See Shapiro, Counter-Reformation, supra note 72, at 707–09 (describing the impact analysis requirements).

      [80].    Elena Kagan, Presidential Administration, 114 Harv. L. Rev. 2245, 2246 (2001).

      [81].   See Shapiro & Wright, supra note 31, at 604–08 (documenting the increase in political appointees).

      [82].   Id. at 583.

      [83].   See Sidney A. Shapiro, The Complexity of Regulatory Capture: Diagnosis, Causality, and Remediation, 17 Roger Williams U. L. Rev. 221, 226 nn.16–17 (2012) [hereinafter Shapiro, Complexity of Capture] (describing studies of industry dominance).

      [84].   Wendy Wagner et al., Rulemaking in the Shade: An Empirical Study of EPA’s Air Toxic Emission Standards, 63 Admin. L. Rev. 99, 119 (2011).

      [85].   Id. at 128–29.  The number of industry comments also greatly outnumbered public interest comments for those rules where there were public interest comments.  Industry filed an average of 35 comments per rule, while public interest groups filed an average of 2.4 comments per rule.  Id.

      [86].   Id. at 128.

      [87].   Id. at 125.  Industry interests had an average of 84 contacts per rule as compared to an average of 0.7 contacts for public interest groups.  Id.

      [88].   Comments raised an average of 22 significant issues in each rulemaking, and EPA on average made changes to the final rule concerning about one-half of these issues.  Of the changes made, 83% of them weakened the rule in some manner.  Id. at 130.

      [89].   See Shapiro, Complexity of Capture, supra note 83, at 239 n.82 (describing the studies that reached a similar result).

      [90].   Id. at 240 n.85 (describing the studies).

      [91].   Id. at 241 (noting comments of observers).

      [92].   Id. at 238 (discussing psychological tendencies).

      [93].   See United States v. Nova Scotia Food Prods. Corp., 568 F.2d 240, 245 (2d Cir. 1977) (reversing an FDA regulation governing good practices for whitefish in part because the FDA failed to respond to an important technical comment in its final rule); see also Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983) (holding that an “agency must examine the relevant data and articulate a satisfactory explanation for its action”).

      [94].   Wendy E. Wagner, Administrative Law, Filter Failure, and Information Capture, 59 Duke L.J. 1321, 1325 (2010).

      [95].   Id. at 1328.

      [96].   Shapiro, Complexity of Capture, supra note 83, at 236–37.

      [97].   Id. at 237.

      [98].   Id.

      [99].   See Shapiro & Schroeder, supra note 34, at 450–51 (discussing evidence of OMB bias).

    [100].   See Steven Croley, White House Review of Agency Rulemaking: An Empirical Investigation, 70 U. Chi. L. Rev. 821, 877, 858–60 (2003) (finding that OIRA sought changes in politically controversial rules, but finding that the type of interest group that attended a meeting with OIRA officials did not predict whether OIRA would change the rule or accept it as is).

    [101].   See, e.g., id. at 858–60.

    [102].   See Shapiro, Complexity of Capture, supra note 83, at 240–41 (noting that industry lobbying tends to confirm an overemphasis on costs among OIRA analysts).

    [103].   Stewart, supra note 4, at 1761.

    [104].   Shapiro, Complexity of Capture, supra note 83, at 226.

    [105].   Also, public interest groups have been adversely affected by the Supreme Court’s more restrictive standing doctrine, which disables them from appealing some agency rules.  See Shapiro, Counter-Reformation, supra note 72, at 717–20 (describing standing restrictions).

    [106].   Wagner, supra note 94, at 1325, 1339, 1352–54, 1364.

    [107].   Fisher, supra note 8, at 89–124.

    [108].   See Rena Steinzor & Sidney Shapiro, The People’s Agents and the Battle to Protect the American Public: Special Interests, Government, and Threats to Health, Safety, and the Environment 126–29 (2010) (noting the almost universal tendency of presidents and other elected officials to engage in “bureaucracy bashing”).

    [109].   See Shapiro & Wright, supra note 31, at 581–85 (noting weaknesses of political oversight); id. at 608–17 (discussing weaknesses of White House oversight).

    [110].   See Shapiro & Schroeder, supra note 34 and accompanying text.

    [111].   See Shapiro & Wright, supra note 31, at 584–85 (discussing weaknesses of judicial review).

    [112].   Abraham H. Maslow, The Psychology of Science: A Reconnaissance 15–16 (1966).

    [113].   See Regulatory Accountability Act of 2011: Hearing on H.R. 3010 Before the H. Comm. on the Judiciary, 112th Cong. 6 (2011) (statement of Sidney A. Shapiro, University Distinguished Chair in Law, Wake Forest School of Law and Member Scholar and Vice President, Center for Progressive Reform), available at
_102511.pdf (demonstrating that, at a minimum, significant rules take approximately four to eight years to complete); Carnegie Comm’n on Sci., Tech., & Gov’t, Risk and the Environment: Improving Regulatory Decision Making 108 (1993) (reporting that the EPA said it takes about five years to complete an informal rulemaking); Richard J. Pierce, Jr., Waiting for Vermont Yankee III, IV, and V?  A Response to Beermann and Lawson, 75 Geo. Wash. L. Rev. 902, 919 (2007) (“It is almost unheard of for a major rulemaking to be completed in the same presidential administration in which it began.  A major rulemaking typically is completed one, two, or even three administrations later.”).

    [114].   See Paul R. Verkuil, The Emerging Concept of Administrative Procedure, 78 Colum. L. Rev. 258, 279 (1978) (contending that administrative efficiency is of equal concern with accountability and fairness in the design of administrative procedure); Paul R. Verkuil, The Ombudsman and the Limits of the Adversary System, 75 Colum. L. Rev. 845, 855 (1975) (stating that efficiency, accountability, and fairness are of equal concern).

    [115].   John F. Kennedy, Inaugural Address (Jan. 20, 1961), in Public Papers of the Presidents of the United States: John F. Kennedy 1, 3 (1962).

    [116].   Steinzor & Shapiro, supra note 108, at 129 (discussing studies showing the demoralization of the bureaucracy and the negative impact on hiring and retention).

    [117].   Id. at 194 (noting the adverse effects on bureaucracy of delegitimization).

    [118].   See Sidney A. Shapiro, OMB and the Politicization of Risk Assessment, 37 Envtl. L. 1083, 1089 (2007) [hereinafter Shapiro, Risk Assessment].

    [119].   Id. at 1087–90 (describing the interaction of law, science, and policy).

    [120].   Wendy E. Wagner, The Science Charade in Toxic Risk Regulation, 95 Colum. L. Rev. 1613, 1617 (1995).

    [121].   Id.

    [122].   Shapiro, Risk Assessment, supra note 118, at 1091.

    [123].   See generally Thomas O. McGarity & Sidney A. Shapiro, OSHA’s Critics and Regulatory Reform, 31 Wake Forest L. Rev. 587, 612–13 (1996) (explaining that the “sound science” campaign objects not to the quality of data collected but to regulations based on incomplete scientific information and conservative default rules).

    [124].   See Thomas O. McGarity & Wendy E. Wagner, Bending Science: How Special Interests Corrupt Public Health Research 157–161 (2008); David Michaels, Doubt is Their Product: How Industry’s Assault on Science Threatens Your Health 198 (2008).

    [125].   Michaels, supra note 124, at 137–38.

    [126].   McGarity & Wagner, supra note 124, at 165–168.

    [127].   Id. at 151.

    [128].   Bruce Wyman, The Principles of the Administrative Law Governing the Relations of Public Officers 4–23 (1903).

    [129].   See, e.g., Jerry L. Mashaw, Bureaucratic Justice: Managing Social Security Disability Claims 15 (1983).

    [130].   Shapiro & Wright, supra note 31, at 580.

    [131].   Orly Lobel, The Renew Deal: The Fall of Regulation and the Rise of Governance in Contemporary Legal Thought, 89 Minn. L. Rev. 342, 356 (2004); Mark Seidenfeld, A Civic Republican Justification for the Bureaucratic State, 105 Harv. L. Rev. 1511, 1514 (1992).

    [132].   See, e.g., Frank I. Michelman, Foreword: Traces of Self-Government, 100 Harv. L. Rev. 4, 66–73 (1986) (calling on the courts to define the values which underlie governmental policy and which are embodied in law); Cass R. Sunstein, Interest Groups in American Public Law, 38 Stan. L.  Rev. 29, 72 (1985) (seeking to revitalize Congress’s deliberative processes through more active judicial review).

    [133].   Mark Seidenfeld, A Civic Republican Justification for the Bureaucratic State, 105 Harv. L. Rev. 1511, 1547, 1556–57 (1992).

    [134].   See Thomas O. McGarity, The Courts and the Ossification of Rulemaking: A Response to Professor Seidenfeld, 75 Tex. L. Rev. 525, 529–30 (1997) (explaining why Seidenfeld’s approach to judicial review will increase ossification).

    [135].   See Lobel, supra note 131, at 344 (describing the program of new governance scholars).

    [136].   Id. at 373.

    [137].   Id.

    [138].   Id. at 359.

    [139].   Elizabeth Fisher, Unpacking the Toolbox: Or Why the Public/Private Divide Is Important in EC Environmental Law, in The Public Law/Private Law Divide: Une entente assez cordiale? 236, 238–40 (Mark Freedland & Jean-Bernard Auby eds., 2006).

    [140].   Gerald Garvey, Facing the Bureaucracy: Living and Dying in a Public Agency 20–25 (1993).

    [141].   Croly, supra note 53, at 361.

    [142].   Id. at 373.

    [143].   Shapiro & Wright, supra note 31, at 592–93.

    [144].   Id. at 588.

    [145].   See supra note 74 and accompanying text.

    [146].   See generally Michèle Lamont, How Professors Think: Inside the World of Academic Judgment ch. 5 (2009).

    [147].    See generally id.

    [148].   See Richard Sennett, The Craftsman 246–49 (2008).

    [149].   See generally Richard Sennett, Together: The Rituals, Pleasures and Politics of Cooperation (2012).

    [150].   See Nat’l Research Council, Understanding Risk: Informing Decisions in a Democratic Society 75–76 (Paul C. Stern & Harvey V. Fineberg eds., 1996).

    [151].   See, e.g., Diana Woodhouse, In Pursuit of Good Administration: Ministers, Civil Servants, and Judges 28 (1997).

    [152].   See generally Herbert Kaufman, The Forest Ranger: A Study in Administrative Behavior (1960).

    [153].   Id. at 85–86.

    [154].   Id. at 198.

    [155].   Id. at 204.

    [156].   John Brehm & Scott Gates, Working, Shirking, and Sabotage: Bureaucratic Response in a Democratic Republic 198 (1999).

    [157].   Id. at 98.

    [158].   Id. at 82–83.

    [159].   Id. at 80–83, 195.

    [160].   Marissa Martino Golden, What Motivates Bureaucrats?: Politics and Administration During the Reagan Years 151–54 (2000).

    [161].   Id. at 151–52.

    [162].   Id. at 152.

    [163].   Id. at 158–59.

    [164].   Id. at 155.

    [165].   See generally Rohr, supra note 2.

    [166].   See, e.g., Terry L. Cooper, An Ethic of Citizenship for Public Administration 139–42, 160–69  (1991).

    [167].   See, e.g., Robert B. Denhardt & Linda deLeon, Great Thinkers in Personnel Management, in Handbook of Public Personnel Management 21, 32–34 (Jack Rabin et. al. eds., 1993); Grace Hall Saltzstein, Representative Bureaucracy and Bureaucratic Responsibility: Problems and Prospects, 10 Admin. & Soc’y. 465, 466–69 (1979).

    [168].   See, e.g., Michael W. Spicer & Larry D. Terry, Legitimacy, History, and Logic: Public Administration and the Constitution, 53 Pub. Admin. Rev. 239, 245 (1993).

    [169].   See generally Charles Johnson Fox & Hugh T. Miller, Postmodern Public Administration: Toward Discourse (1991).

    [170].   Shapiro & Wright, supra note 31, at 616.

    [171].   Fox & Miller, supra note 169, at 11, 13.

    [172].   Zeke Miller, President Obama’s $90 BILLION Clean Air Regulation, Bus. Insider (Sept. 1, 2011),

    [173].   See, e.g., Scott Learn, Environmental News: Pressure’s on EPA to back off New Smog Regulation, OregonLive (July 18, 2011, 9:13 AM),

    [174].   Clean Air Act §109, 42 U.S.C. § 7409(d)(1) (2006).

    [175].   Id., 42 U.S.C. § 7409(b)(1).

    [176].   See Whitman v. Am. Trucking Ass’ns, Inc., 531 U.S. 457, 464–65 (2001).

    [177].   Am. Trucking Ass’ns, Inc. v. EPA, 175 F.3d 1027, 1034–40 (D.C. Cir. 1999).

    [178].    See Laws and Regulations, EPA (Nov. 4, 2010),
/oaqps001/permits/requirem.html; State Implementation Plan Overview, EPA (Mar. 27, 2012),

    [179].   See, e.g., Summary of the Updated Regulatory Impact Analysis (RIA) for the Reconsideration of the 2008 Ozone National Ambient Air Quality Standard (NAAQS) at S1–4, EPA,
(indicating a range of between about $12 billion and more than $20 billion in costs annually between an ozone standard of 0.070 ppm versus 0.075 ppm).

    [180].   See, e.g., John M. Broder, Obama Abandons a Stricter Limit on Air Pollution, N.Y. Times, Sept. 3, 2011, at A1 (describing President Obama’s decision to reject a more stringent ozone standard despite strong scientific evidence, including CASAC endorsement in its favor); Clinton Oks Tough Clean Air Standards: Congress has Power to Overturn New Rules, Cleveland Plain Dealer, June 26, 1997, at A1 (describing President Clinton’s controversial decision to back the EPA’s more stringent ozone standard).  Vice President Gore was even dubbed the “Ozone Man” by President George H. W. Bush during campaign season due to Gore’s advocacy of strong air quality rules.  ‘Ozone Man’, Baltimore Sun (Oct. 14, 2007),

    [181].   EPA, Review of the Process for Setting National Ambient Air Quality Standards 3–5 (2006), available at
(describing the process prior to 2006).  The summaries of the assessments found in the agency’s proposed rulemaking preamble were somewhat more accessible, but they were litigation oriented rather than presenting a dispassionate and frank discussion about limits in the evidence. See, e.g., Wagner, supra note 120, at 1629–31 (providing several examples of this).

    [182].   See, e.g., Cmtys. for a Better Env’t v. EPA, No. C 07–03678 JSW, 2008 WL 1994898, at *1 (N.D. Cal. May 5, 2008) (bringing suit to compel the EPA to perform its past due, mandatory review duties).

    [183].   See, e.g., Am. Trucking Ass’ns, Inc. v. EPA, 175 F.3d 1027, 1055–56 (D.C. Cir. 1999).

    [184].   The “kick-off” workshop is a major event during which the agency’s staff solicits comments from the public and scientific community (including invited scientists) about developments in the science and policy that should frame the EPA’s review.  The workshop focuses specifically on scientific discoveries and related developments occurring over the past five years that might suggest the need for a revised standard and hence deserve careful scientific review.  See, e.g., EPA, supranote 181, at 9–10.

    [185].   The primary purpose of this planning document is to frame “key policy-relevant issues that would generally be used to frame the science assessment, risk/exposure assessment, and policy assessment . . . .”  Id.  The report also sets a timetable for subsequent stages of the process.  Id.

    [186].   Id. at 14.

    [187].   For a sample planning document, see generally EPA, Integrated Review Plan for the National Ambient Air Quality Standards for Particulate Matter (2008),available at
.  In particular, see id. at 18–21 (listing policy-relevant questions for primary PM NAAQS that expand on the excerpts provided above in the text).

    [188].   See generally id.

    [189].    Id. at 18. (“The first step . . . is to consider whether the available body of scientific evidence . . . supports or calls into question the scientific conclusions reached in the last review regarding health effects related to exposure to fine and thoracic coarse particles in the ambient air.  This evaluation of the available scientific evidence will focus on key policy-relevant issues by addressing a series of questions . . . .”).

    [190].   Air Quality: EPA’s Integrated Science Assessments (ISAs), EPA (Feb. 2, 2012),

    [191].   See EPA, supra note 187, at fig. 4.1 (creating a descriptive flowchart for production of the scientific assessment).

    [192].   For an example, see Nat’l Ctr. for Envtl. Assessment, Integrated Science Assessment for Lead, at xxii–xxx (2011), available at

    [193].   Id. at xxii–xxviii.

    [194].   Id. at xxii–xxx.

    [195].   See, e.g., EPA, supra note 187, at 41 (describing this goal of the risk assessment).

    [196].   See, e.g., id. at 54.

    [197].   For a sample of a policy assessment, see generally EPA, Policy Assessment for the Review of the Particulate Matter National Ambient Air Quality Standards (2011), available at

    [198].   For a very brief summary of CASAC input, see id. at 2–100 to 2–101 (summarizing CASAC advice).

    [199].   For the second CASAC review of the EPA’s policy assessment for particulates, see Letter from Dr. Jonathan M. Samet, Chair, CASAC, to Lisa P. Jackson, Adm’r, EPA (Sept. 10, 2010), available at

    [200].   See, e.g., Marcia R. Gelpe, Exhaustion of Administrative Remedies: Lessons from Environmental Cases, 53 Geo. Wash. L. Rev. 1, 10–25 (1985) (outlining the rationale behind the exhaustion requirement).

    [201].   Interview by Wendy Wagner with EPA Staff, Nat’l Ctr. for Envtl. Assessment, Office of Research & Dev. (Jan. 18, 2012).

    [202].   See, e.g., EPA, Preliminary Draft Policy Assessment for the Review of the Particulate Matter National Ambient Air Quality Standards (2009), available at
.  Note in this document that the names of individual staff are not listed, however.  This is different from the final report, which includes a detailed acknowledgement section that lists staff and reviewers by name and identifies their specific contributions to the report.

    [203].   It is not clear whether an agency staff member has the right to remove his or her name from this acknowledgement section if he or she disagrees with the final version of the chapter (presumably the issue has not yet arisen), but if this is the case, then these acknowledgements provide an indicia of authorship.

    [204].   Because the assessment was published as an ANPR, it required OMB-clearance.

    [205].   Interview by Wendy Wagner with EPA Staff, Office of Air Quality Planning & Standards (Jan. 17, 2012).

    [206].   See Letter from Dr. Rogene Henderson, Chair, CASAC et al. to Stephen L. Johnson, Adm’r, EPA (Jan. 23, 2008), available at
(condemning the ANPR for lead prepared by EPA management as “unsuitable and inadequate” because it does not provide “the underlying scientific justification” for the “range of options for standard setting” that the agency is currently considering, and providing substantial details in the remainder of the letter regarding these concerns).

    [207].   See Memorandum from Lisa P. Jackson, Adm’r, EPA, to Elizabeth Craig, Acting Assistant Amd’r for Air & Radiation, EPA, & Lek Kadeli, Acting Assistant Adm’r for Research & Dev., EPA (May 21, 2009), available at (regarding the process for reviewing NAAQS).

    [208].   Interview by Wendy Wagner with EPA Staff, Nat’l Ctr. for Envt’l Assessment, Office of Research and Dev. (Jan. 18, 2012); interview by Wendy Wagner with EPA Staff, Office of Air Quality Planning and Standards (Jan. 17, 2012).

    [209].   See Jackson Memo, supra note 207.

    [210].   See, e.g., EPA, Safeguarding the Future: Credible Science, Credible Decisions 38 (1992), available at
/pdf;jsessionid=E038C35EEEF84C12537448A1DA3EC1ED.tobacco03 (noting the positive effect of CASAC on EPA’s decisions); Mark R. Powell, Science at EPA: Information in the Regulatory Process 43 (1999) (reporting on how persons interviewed for the study on science at the EPA “gave SAB and CASAC credit for improving EPA’s acquisition and use of science”).

    [211].   See Memorandum from Roger O. McClellan, Advisor, Toxicology & Human Health Risk Analysis, EPA, to Dr. George Gray, Assistant Adm’r for the Office of Research & Dev., EPA (Mar. 18, 2006), available at

    [212].   See, e.g., Nat’l Research Council, Review of the Environmental Protection Agency’s Draft IRIS Assessment of Formaldehyde 120–21 (2011).

    [213].   See Jackson Memo, supra note 207.

    [214].   See id. (indicating that there is no OMB involvement in the NAAQS review process).

    [215].   See, e.g., Mobil Pipe Line Co. v. FERC, 676 F.3d 1098, 1099 (D.C. Cir. 2012) (holding that FERC’s decision was arbitrary and capricious because “FERC’s expert staff,” which had found the petition to be a “slam dunk,” demonstrated that the Commission’s decision “was unreasonable in light of the record evidence”).

    [216].    See Arnold W. Reitz, Jr., The Intersection of Climate Change and Clean Air Act Stationary Source Programs, 43 Ariz. St. L.J. 901, 930 (2011) (“[O]n September 11, 2011, President Obama rejected EPA’s proposed changes to the ozone standard.”); Juliet Eilperin, Proposed Standards for Air Quality Criticized, Wash. Post (Dec. 21, 2005),

    [217].   Obama Blocks Stricter Ozone Standards, Troutman Sanders LLP: Wash. Energy Rep.,
/obama-blocks-stricter-ozone-standards/ (last visited Sept. 15, 2012).

    [218].   See, e.g., Eilperin, supra note 216 (describing the Bush Administration’s decision to reject a more stringent particulate standard despite strong scientific evidence, including CASAC endorsement, in its favor); Letter from Cass R. Sunstein, Adm’r, OIRA, to Lisa Jackson, Adm’r, EPA (Sept. 2, 2011), available at (returning the ozone standard in part because the President is reluctant to “impose significant costs on the private sector or on state, local, or tribal governments” during this “economically challenging time”).

    [219].   See Bruno Latour, We Have Never Been Modern 130–45 (Catherine Porter trans., 1993).