Legitimacy, selectivity, and the disunitary executive: a reply to Sally Katzen.

• Author: Bressman, Lisa Schultz
• Position: Response to Sally Katzen, Michigan Law Review, vol. 105, p. 1497

INTRODUCTION I. SELECTING EPA II. DEFINING ACCOUNTABILITY AND EFFICACY III. THE WIZARD(S) OF OZ IV. SELECTIVITY OF WHITE HOUSE INVOLVEMENT CONCLUSION INTRODUCTION

This reply addresses the thoughtful comments that former OIRA Administrator Sally Katzen has provided on our Article, Inside the Administrative State: A Critical Look at the Practice of Presidential Control. (1) Our Article is the first to investigate the agency perspective on White House involvement in agency rule-making. We interviewed 30 of the 35 top political officials in the Environmental Protection Agency ("EPA") during the George H.W. Bush ("Bush I") and the William J. Clinton Administrations during 1989-2001. (2) Prior to our study, empirical studies of White House involvement in agency rule-making had focused almost exclusively on the White House side, mainly analyzing White House documents and interviewing White House officials. (3) While these studies could describe how the White House exercises control of agency rule-making, they could not speak to how agencies experience such control. We began to remedy the imbalance and paint a more complete picture of presidential control. (4)

By engaging our study, Katzen has added a valuable perspective to the debate. She has also helped to confirm the value of empirical studies in this field. Empirical studies sharpen our understanding of the relevant facts and thus narrow the gaps in the theoretical debate. At the end of the day, Katzen's claims are based only on her own experience on the White House side and, as a result, are yet more anecdotes from that side of the executive branch. Nonetheless, her comments are important. They confirm the need to examine whether presidential control works from the agency side and the need, more than ever, to get additional data that might explain why her recollections seem strikingly different from those of the top EPA presidential appointees we interviewed for our study.

We use this brief reply to contextualize Katzen's comments about our data. Katzen not only comes at our data from the White House side, but she also views centralized rule-making review primarily from a practical standpoint. In her view, such review works well to manage the bureaucracy. (5) Practice is important and worth considering. But missing from that perspective, and central to our Article, is consideration of whether such review serves the role that presidential control theorists have assigned to it of legitimating agency decisionmaking and grounding agencies within our constitutional structure. (6) Missing from Katzen's account is the overarching importance of accountability and efficacy to executive branch decisionmaking. Our data suggest that there may be a problem on this front. (7)

1. SELECTING EPA

   Katzen first suggests that our data are biased. (8) She contends that EPA is not representative of other agencies and that we have used EPA data inappropriately to draw conclusions about a larger phenomenon--White House oversight of all agency rule-making. (9) Importantly, Katzen does not argue that our data are biased regarding EPA in the way that empirical social scientists use that term. Selection bias in the social science literature refers to selecting a sample that is not representative of the population on a variable of interest. (10) Katzen does not dispute that our sample (30 EPA presidential appointees subject to Senate confirmation, known as "PAS"s) is representative of the total population (35 PASs, of which 34 are living) regarding any of the variables we studied. Thus, she essentially raises issues as to whether EPA was a proper target for our study and whether we can draw inferences from our data regarding other regulatory agencies.

   We have two responses. First, even if EPA is not representative of all agencies, it is a sufficiently large and important regulatory agency that no theory of presidential control can afford to ignore. (11) Katzen acknowledges that major rule-makings were the focus of OIRA review under the executive orders in place during the Clinton and Bush I Administrations. (12) She also acknowledges that EPA generated the most major rule-makings during the Clinton Administration and was third during Bush I. (13) Whether EPA was first or third over any given period, it was--and is--clearly one of the most important rule-making agencies in the federal government. In addition, the total cost generated by an agency's regulations is arguably an even more important measure of importance than the number of major rule-makings. (14) On that score, EPA dwarfs other regulatory agencies. In 1990, for example, the Office of Management and Budget ("OMB") noted that the EPA accounted for roughly half ($70-$80 billion) of all direct costs on the economy imposed by all federal regulations ($175 billion). (15) More recently, an OMB report estimated that between 1987 and the first quarter of 1999, environmental regulations imposed $71 billion of the total $92 billion in costs from all federal social regulations. (16) Environmental regulations also provided $75-$145 of the $198-$274 billion in quantified, monetized benefits. (17) Moreover, we further note that EPA has influence far beyond the mandates of specific major rule-makings. For example, environmental regulations have played a remarkable role in shaping judicial review (18) and congressional oversight of executive branch decision-making. (19)

   Second, EPA has strong similarities to a sizeable number of executive branch regulatory agencies, many of which are sub-units of federal departments. (20) Examples include the Food and Drug Administration ("FDA"), the Occupational Safety and Health Administration ("OSHA"), the Fish and Wildlife Service, and the National Marine Fisheries Service. In the aggregate, agencies such as these generate large volumes of major and non-major rule-makings. It is likely that the responses of the top managers at FDA, OSHA, or other agencies will not differ a great deal from the responses of the top managers at EPA. Like EPA, these agencies have statutory mandates to achieve national health, safety, and environmental goals.

   Of course, EPA is not like all agencies. It is dissimilar from non-regulatory agencies (e.g., the Department of Defense), as well as agencies without statutory mandates regarding health, safety, and the environment (e.g., considerable portions of the Departments of Commerce, Treasury, and Agriculture). (21) Nevertheless, our EPA results may hold for an extremely wide swath of the regulatory state.

   Although we think it likely that our results regarding EPA are generalizable to many other regulatory agencies, we do not claim that our EPA results hold for these agencies. (22) For now, we can only speak to our data, which are solely derived from EPA. Our study demonstrates the importance of gathering data in a more systematic fashion from sources on both sides of the White House oversight process. (23) Katzen's comments help identify the types of agencies that should be subject to future empirical studies. We are pleased if the value of that data collection is clear. We note, however, that our study cannot simply be dismissed or marginalized because it is the first on the agency side or because it is the first to reveal a picture that does not square easily with the view from the White House side.

   Katzen also makes another claim of bias. (24) She argues that EPA officials are biased in favor of a "parochial" interest, while OIRA officials are focused on the "national" interest. (25) We dispute this claim at the theoretical and practical level. First, we dispute the general characterization of EPA's institutional interest as parochial and OIRA's interest as national. In theory, both have national interests at stake. Both pursue the goal of maximizing social welfare. We think it is incorrect to label as "parochial" an agency's statutory obligation to reduce human health and environmental risks.

   Second, we believe that EPA officials are no more or less likely to reflect a bias in favor of their mission than are OIRA officials to reflect a bias in favor of theirs. As a practical matter, the claim of bias cuts both ways. White House staffers, whether at OIRA or any other White House office, can become just as attached to the mission of their offices as agency officials. Katzen makes no bones about her high regard for and deep appreciation of the career staff of OIRA, (26) We do not doubt that the career staffers are the dedicated civil servants she believes them to be, but they are not immune from bias. Katzen goes out of her way to point out that both the Bush I and Clinton PASs at EPA often "had previously been environmental activists or state and local officials," without noting that many also came from and returned to positions in corporations or corporate law firms. (27) She also does not respond to a point that we highlighted, well expressed by the EPA respondent who noted that "the civil servants in OIRA, who had been there largely since the Reagan Administration ... were more conservative and suspicious of EPA regulations than the political appointees." (28) Perhaps...