Is the bureaucracy a deep state run by unaccountable bureaucrats? Most of us in administrative law think not, but when push comes to shove, we must concede that we have precious little idea of how agency staff operate on the ground to fulfill their statutory commands.
In The Accountable Bureaucrat, Anya Bernstein & Cristina Rodríguez (hereinafter B&R) step into this void and, in a manner reminiscent of the big reveal in the Wizard of Oz, expose the “man behind the curtain.” By asking the bureaucrats themselves how they transform “abstract statutes into concrete rules that govern conduct” (P. 1679), the authors make significant advances in our understanding of this important, but rarely studied world of bureaucratic discretion. Armed with a pre-prepared list of questions that structure their open-ended interviews, the authors (Bernstein holds a PHD in anthropology) canvass thirty-nine current and former employees (both political and career) from eleven different agencies, ranging from the Department of Homeland Security to the Environmental Protection Agency. (Pp. 1685-86.) B&R then code the transcripts and notes with a detailed rubric in search of patterns and behaviors that cut across the different regulatory programs.
What do they find? In an article overflowing with goodies for administrative law nerds, B&R make two particularly significant contributions—the first empirical and the second theoretical. On the empirical side, B&R provide a rich description of what is going on inside the black box of agency decision-making in the 21st century, supplementing the few dated books published on the topic. One of their several stand-out findings includes the discovery that—despite the rise of the unitary executive—the Chief Executive is rarely controlling agency policymaking or even providing clear directions in some cases. Instead, B&R find that generally the President and his appointees provide only an “overarching policy orientation,” and even this direction is negotiated dynamically throughout the decision-making process and mediated with other sources of input and influence. (Pp. 1626, 1628, 1630.) Indeed, they find that the “normal” mode of interaction between appointees and career staff was generally not one of conflict, but instead of “interdependence.” (P. 1631).
Another stand-out finding is B&R’s depiction of how, in the absence of top-down direction, agencies actually exercise this discretion. The authors portray agencies as operating essentially like massive vacuums—drawing in input from congressional offices, executive orders and speeches, stakeholders, experts, and other agencies in the process of crafting rules and policies. Characterizing what one interviewee called a “spider web”, the agencies then synthesize all this information through “multilateral negotiations among teams of officials with diverse roles and levels of seniority.” (P. 1640.) As a result, agency rules often end with a decision that is quite different from where they started.
B&R’s contribution to administrative law theory is even more impactful. The authors offer a conceptual model—consisting of several basic “scaffolding” elements—that captures the recurring behaviors the authors observe in their data. This scaffolding includes:
- decision-making that is highly diffused among diverse staff;
- decisions that are not hierarchical (top-down) but instead involve a kind of all-hands-on-deck, negotiated quality; and
- agencies that are highly responsive to stakeholder input, feedback, and engagement throughout the decision process.
(P. 1606.) Based on their findings, B&R argue that this model of bureaucratic decision-making provides a more accurate description of how most rules and policies are actually made, as contrasted with the caricature of a command center run by the President and his henchmen.
However, B&R do not stop with this new, empirically grounded model of how bureaucrats often exercise policy discretion but proceed to engage their findings with normative theory as well. After canvassing the theoretical literature, which emphasizes how “accountability consists of a reasoned deliberation that considers many options; an inclusivity that takes into account a wide range of interests; and a responsiveness to both affected publics and conditions on the ground” (Pp. 1668-69), the authors observe how the default bureaucratic processes they have unearthed generally hit this normative target. (Pp. 1605, 1671, 1674.) Indeed, they suggest that the “responsible bureaucrat” model may not only be superior to the unitary executive in meeting these indicia of accountability, but the model might be used to gauge the accountability of individual agency decisions going forward. (P. 1609.)
Given the dearth of empirically grounded work on bureaucratic discretion and accountability, one cannot overstate what the authors accomplish in this one article. However, it is precisely because of the inevitable future import of this piece that the authors’ normative claims in particular deserve further attention. Perhaps the “accountable bureaucrat” title is intended to convey only a relative virtue—namely that the type of bureaucratic decision-making model they discover is at least as accountable as the unitary executive model. But just in case the authors mean to suggest that the model might also serve as an ideal template for bureaucratic accountability more generally, added troubleshooting is warranted.
One concern is that—even when followed to the letter—there are at least a few settings in which the model will not produce accountable decision-making. As one example, when the model is applied to the agencies’ scientific expertise it seems poised to undermine the reliability of the agency’s expert work rather than enhance it. The all-hands-on-deck negotiating style of the accountable bureaucrat means that everything, including the underlying syntheses of the scientific literature, is up for grabs. But this means that basic principles of scientific integrity—including the scientists’ independence from influence, the need for transparent reasoning, and rigorous expert peer review—may not only be ignored in the process of incorporating expert work but violated.
Second, while responsiveness to stakeholders is certainly a vital ingredient to accountable agency decision-making, we know in practice that stakeholder engagement is frequently one-sided—especially in high stakes, technical rules dominated by well-financed industry. To suggest that bureaucrats who respond to this machine-gun fire from regulated industries are “accountable” is perhaps factually correct (that is all the Administrative Procedure Act and the reviewing courts legally require of bureaucrats), but as a normative matter it seems wrong to consider this agency responsiveness publicly accountable when it considers only a subset of affected parties. Yet from my reading, that is where the current model leaves us.
Finally, because so much agency decision-making occurs outside of the public eye and is not transparent—thanks in part to the deliberative process privilege—the “accountable bureaucrat” still resides, at least to some extent, in a kind of “deep state” or, in the words of Peter Strauss, “shade.” Even when an agency process closely maps against the model’s scaffolding, the underlying deliberations may still be afflicted with serious blind spots, flawed assumptions, or staff biases that are invisible to outsiders and lead to systematic substantive problems in the resultant decisions. Given this lack of transparency, added refinements are needed to ensure that agency decisions are truly accountable.
Thankfully, the authors are already hard at work on some of these (and other) challenges. In two subsequent articles that mine this same data set (one of which is forthcoming in our own Texas Law Review), B&R elaborate on the implications of their findings and suggest ways to shore up the model to make it even more accountable. Thus, The Accountable Bureaucrat is not the end, but only the beginning of what this dynamic duo has in store for us.
