IT ISN'T EASY BEING A BUREAUCRATIC EXPERT: CELEBRATING THE EPA'S INNOVATIONS.

• Author: Wagner, Wendy
• Position: The EPA at Fifty Symposium

CONTENTS INTRODUCTION I. A HOSTILE INSTITUTIONAL ENVIRONMENT FOR AGENCY EXPERTISE A. Agencies Are Subjected to Political, Top-Down Direction B. Notice-and-Comment Rulemaking and Judicial Review Can Further Undermine an Agency's Incentives for Producing Scientifically Grounded Rules C. Staffing the Expert Bureaucracy Is also Subject to Political Control II. EXPERT INNOVATIONS AT THE EPA III. THE PATH TO MORE CONSISTENT BUREAUCRATIC EXPERTISE CONCLUSION INTRODUCTION

Fifty years ago, the Environmental Protection Agency (EPA) was created to serve as the nation's expert on all issues environmental. The EPA's 1970 authorizing charter delineates four roles for the new agency, two of which direct the fledgling agency to serve as the nation's expert authority on environmental protection. (1)

Because it is a core element of EPA's mission, (2) providing high quality expertise remains one of the legally established benchmarks against which we can judge the EPA's success. (3) While the demarcation of science from other ways of knowing is intrinsically fuzzy, (4) it seems uncontestable that, at the very least, agencies should make full and rigorous use of the scientific literature in informing policymaking. (5) Regulatory decisions are more effective and efficient when informed by the very best scientific and technical information. Moreover, if staffed with excellent professionals, agencies are well situated to provide rigorous, policy-relevant scientific information that is difficult to recreate at a grand scale in the private sector. Gillian Metzger argues that the Constitution may even require the administrative state to be staffed with "professional and expert government employees ... insofar as such career staff are necessary to ensure expertise and institutional stability in agencies." (6)

Yet, over the last fifty years, the integrity of the EPA's expertise has been under almost continuous attack. Industry, members of Congress, executive officials, public stakeholders, and all facets of the media have focused their criticisms--sometimes in unison--at the EPA's use of "bad science" and its failure as the nation's environmental expert. (7)

In this Article, I take a close look behind the media coverage and legal challenges to locate some stunning innovations in the EPA's ability to provide quality scientific advice. (8) On the Agency's fiftieth anniversary, I spotlight these expert accomplishments, which occur despite considerable political pressure and public controversy. The EPA in these cases has not only fulfilled its mission as a nation's expert, but it has pioneered models for other agencies with regard to how to provide this essential, but hotly contested science advice.

Of course, the EPA is not a monolith, and some of its expert work has, from time to time, rightly sparked public condemnation. (9) Indeed, even the accomplishments discussed here are full of fits and starts. But extracting some of the EPA's successes in providing expertise tells us something--perhaps a lot--about expert agencies; it also offers valuable lessons for the future. The Trump Administration's targeted, and thus far still only partial, efforts to dismantle some of the EPA's expert-based programs only further underscore the important role of credible, scientific expertise, as well as the remarkable resiliency of the EPA, even in response to wholesale executive attack. (10)

In discussing this expert role for the EPA, it is nonetheless important to be crystal clear about what is and is not meant by "bureaucratic expertise." The expert role of the EPA explored here concerns only the Agency's capacity to provide rigorous scientific information and advice. It does not include the ability of agency experts to select the best regulatory standard or outcome. In the literature, these two very different roles are sometimes conflated: an "agency expert" is portrayed as one that chooses a preferred policy, as well as conducts the underlying analysis. (11) That is emphatically not the scope of the expert role considered in this Article. Instead, the focus is simply on the first, advisory step: is the agency able to provide rigorous expert analyses to the nation or is even this advice-giving role undermined by existing administrative procedures? (12)

This Article's celebration of the EPA's periodic success as an expert bureaucrat begins in Part I by reconstructing the institutional architecture within which expert agencies must operate. While on the one hand, a significant role for the EPA is tied to the Agency's ability to provide rigorous scientific expertise to inform policy; on the other, we have placed a number of legal constraints on the EPA's ability to carry out this goal. Rather than constructing an executive branch that nurtures and encourages excellent expert analysis by agencies, there are key architectural features of administrative process that make it difficult, and sometimes impossible, for agencies to deliver this much-needed expert advice. Part II then considers the EPA's ability to rise above these institutional impediments at various points in its fifty-year history and nevertheless provide rigorous scientific expertise in spite of the less-than-ideal legal structure for doing so. Part III closes with some thoughts for the future of expert agencies.

1. A HOSTILE INSTITUTIONAL ENVIRONMENT FOR AGENCY EXPERTISE

   The scientific research that informs policy is a natural target for political and stakeholder pressure when the findings are inconvenient. Accounts of ends-oriented manipulations of policy-relevant science are legendary. (13) It should thus come as no surprise that a great deal of scholarly and governmental effort has been dedicated to devising ways to protect applied scientific research and analyses from this type of biasing. Science journals in particular have devised various methods to help protect the integrity of research from external influence. (14) These methods benchmark the rigor of a scientific study or analysis in part by how well the researcher is afforded autonomy and is able to approach the research in a skeptical, open-minded way. (15) Science journals, for example, insist on full disclosures of conflicts of interest, as well as certifications of authorship that include assurances that the author enjoyed full control over all aspects of the research. (16)

   In designing expert bureaucracies, this same "independence" has become a central structural theme as well. In a series of recent reports, the Organization for Economic Cooperation and Development (OECD) reiterates independence from political and stakeholder control as the most important institutional design feature to advance agency expertise. (17) The essential role of agency autonomy has also been spotlighted by other commenters as a central design principle for administrative design. (18) Indeed, without some agency autonomy in conducting the foundational scientific analyses that inform policy (such as a literature review, for instance), there will be no limit to the invisible, ends-oriented tinkering that could occur to bias the analyses. (19) This is especially true if there is no record of these sources of influence.

   But administrative processes that govern agency expertise in the United States move in precisely the opposite direction from these best practices. Several key institutional features may, in fact, prove antithetical to nurturing and protecting expert advice within our agencies. These features are not ephemeral or "Trumpian," but have become hard-wired into the structure of how we design administrative processes. Despite the centrality of expertise to bureaucracy and public administration, then, the administrative state's architecture presents a series of impediments to this goal of autonomous expertise.

   1. Agencies Are Subjected to Political, Top-Down Direction.

      Political appointees within the agencies and sometimes in the White House have long enjoyed control over all aspects of their agency's work, including its scientific analyses. (20) In some agencies, the very first science-intensive report that is made public is jointly offered between the scientific staff and political officials. (21) In other cases, even if a political appointee delegates authority to the technical staff and is not engaged in the analytical work, it is technically his or her prerogative to intervene and hold up, and, in some cases alter, that technical analysis if he or she so chooses. (22)

      Moreover, if there are internal disagreements between the scientific staff and the policy officials, the fact of these disagreements--not to mention their merits and ultimate resolution--are routinely classified as deliberative process and protected from public disclosure under the Freedom of Information Act ("FOIA"). (23) In these internal government discussions, outsiders cannot trace the role that politics might play in an analysis that is presented as technical in nature. (24)

      It may seem initially disconcerting to allow for this type of top-down, nontransparent, political control over all aspects of an agency's work, but the overarching logic is sound. (25) Since aspects of these technical analyses almost always involve some modicum of policy discretion--and since an agency must speak with one voice and ensure the scientific integrity of its analyses--individual staff should not be allowed to operate with complete independence. (26) Misleading technical information could surface in the public sphere before it has been adequately vetted. Rogue agency staff would accordingly enjoy too much power and might confuse and terrify the public. Official agency clearance is thus needed at some level to ensure the integrity of that agency's work.

      When carried too far, however, this institutional design can allow for political manipulation of the scientific record in ways that escape detection. (27) Indeed even Justice Kagan, who is considered a proponent of presidential...