It is both humbling and daunting to have your book reviewed by so many eminent scholars. I am deeply grateful to Chris Walker and the JREG crew for hosting this symposium, and especially to Bridget Dooling for organizing it.   When I first began writing Bending the Rules: Procedural Politicking in the Bureaucracy, my hope was to encourage scholars to think about all the different ways that politics can (and does) intervene in the notice-and-comment rulemaking process. This symposium is heartening because it suggests that the book has helped to foster that conversation.

The notion of procedural politicking that I lay out suggests that, in a contentious political environment, the success of an administrative process rests not just on the ideas associated with a particular policy but also with the particular procedures used to carry that policy to fruition. As a quantitatively-oriented political scientist, my intention was to bring data to bear on the question of procedural politicking.  Because I do not view any one empirical test as dispositive evidence of procedural politicking, I present many empirical tests and marry these with numerous case studies.

The participants have reacted differently to this approach.  They have offered incisive critiques, identified interesting wrinkles, and pointed out opportunities for extension.  The exciting part for me is that these perspectives draw from three different disciplines: law (Bell, Bremer, Dooling, Nou, Walker), public policy (Carrigan and Shapiro), and political science (Rudalevige and me).  Additionally, three of the contributors (Dooling, Nou, and Shapiro) are fellow alumni of the Office of Information and Regulatory Affairs (OIRA), giving them a practical perspective on regulatory politics.  Suffice it to say, I’m absolutely thrilled to have all of these varying perspectives on my book.

So, without further ado, I’ll dive right into responding (with as much brevity as possible!)

I. Bridget Dooling’s Introductory Post

Bridget Dooling is the mastermind behind this symposium.  Her introductory post is very gracious and, if you’ve been following along on this symposium, then you should definitely be following the exciting work of Professor Dooling and her GWU colleagues at the Regulatory Studies Center.  At a minimum, you should be getting their newsletter (you can sign up here), which offers weekly updates on happenings in the world of regulatory politics.

II. Bernard Bell’s Procedural Politicking or Bureaucratic Perfidy, Part I

Bernard Bell responds to my book in two parts (!!) and I am appreciative that he took the time to engage in such a detailed and thoughtful way.  This response takes up his first post, which deals with the empirical results in the fourth chapter of the book (I’ll respond to his second post later).

If I may succinctly cut to the heart of Bell’s commentary, it is that he thinks my findings about rule drafting are more consistent with sincere agency behavior than strategic behavior.  That is, what I present as evidence that agencies act strategically to insulate their rulemakings from political interventions by the president, the Congress, and the courts, he views as equally consistent with agencies working seriously to improve the quality of their rulemakings.  He concludes that the book’s empirical results on the rule drafting front are “ambiguous.”

Let me explain why Professor Bell and I will have to agree to respectfully disagree on this interpretation.  First, let’s be clear about exactly what sincere behavior might look like.  Imagine a case where an agency is trying to write the best possible proposed rule.  They might write a long, clear proposal that lays out their arguments so as to ensure that all of the relevant audiences understand exactly what they are trying to accomplish and why.  This is what I might call sincere behavior.  Crucially, there is no reason to think that this behavior should be correlated with the political environment.  I show, however, that it is.

Once the crafters of a rulemaking text factor in how political actors might respond to it, strategic behavior starts to creep in.  So, when an agency writes a longer preamble (for instance) to “satisfy exacting judicial review,” is that not strategic behavior in action?  While this is the exact example Professor Bell gives for sincere behavior, this is the type of strategic behavior I was trying to shed light on! And, in terms of the two measures I look at—preamble length and abstract readability­—it is what the analyses demonstrate (at least for the courts).   (Part of the disconnect may stem from Bell’s equating strategic behavior with nefarious behavior; these are not the same things, as I explain later in my response to Part II of Bell’s review.)

There are other smaller points we might quibble over.  For instance, Bell seems to attach particular meaning to the lack of strong statistical results with respect to one of the levers I study: the readability of abstracts.  The argument I make here is that agencies should write less readable abstracts when the political environment is unfavorable to their interests.  The results show that this happens with respect to the courts, but not the other two branches.  While Bell suggests this is evidence in favor of sincere behavior, I am more hesitant.  Data in the social sciences are inherently messy.  One cannot make strong inferences from null results; as my favorite statistics professor used to quip, “the absence of evidence is not evidence of absence.”  In my opinion, this sounds a call for other researchers to come in and apply clever research designs that can more carefully unpack what’s happening here, not a call for definitive conclusions one way or the other.

Additionally, he argues that abstract readability is not a good metric in the first place because “it seems unlikely that a court would be put off from engaging in rigorous judicial review by even the most incomprehensible of abstracts.”  Here I think we are in complete agreement in one sense.  I agree that abstracts should not be decisive for elite political actors; this is why my argument is conditional on interest groups; they are the ones that bring these rules to the attention of busy Members of Congress and that bring agencies to court. Interest groups—a broad term meant to encompass activists, lobbyists, and other organized interests—engage in the political sphere in what scholars long ago dubbed a “fire alarm” system.  These groups, who have many competing demands for their attention, may be more susceptible to this kind of procedural politicking.

There is more I could say, but in the interest of brevity let me conclude by saying that a good-natured debate on these points underscores the highly elusive nature of strategic behavior.   It is difficult to find a definitive smoking gun with strategic behavior of any sort—and procedural politicking is no different.  It can readily masquerade as sincere behavior (and that is the exact objective of those who employ it!)  and that is precisely why the design of the book features many different parts of the process and mixes quantitative approaches with case studies.

III. Jennifer Nou’s Bureaucratic Power

Jennifer Nou’s incisive essay takes the issue of bureaucratic power head-on.  Note that Professor Nou is an expert on this topic herself, having written extensively on bureaucratic resistance and the strategic behavior of bureaucrats (see, e.g., here and here).  She highlights the central role of the “second face of power” in the book, which she aptly defines as a kind of conflict avoidance: agencies use procedural tools to avoid conflict with political overseers and thereby enhance the probability of rules being successfully completed.

Nou makes two points about the role of this type of invisible power in the book.  First, she questions whether some of the measures I employ in the analyses—such as D.C. Circuit case volume—adequately capture real opposition from political overseers to agencies. She also notes that the analyses focus on hypothetical opposition and possibly not “genuine” opposition.  I agree that the measures I employ are not as tightly bound to “genuine” opposition as they might be.  In an ideal world, we might have survey data about perceived opposition from both the bureaucrats working on a rule and the political actors overseeing that particular rule.  These data, of course, are not available, particularly not on the scale that I study in the book.  Even with the imperfect measures of opposition that I use, however, I find a meaningful relationship between the times we might expect there to be genuine opposition and procedural politicking behavior.  In a world of second-bests, this weighs in favor of the existence of procedural politicking.

Second Nou suggests that agencies may be even more powerful than the book acknowledges.  Here she points to the so-called “third face of power” whereby agencies may be able to co-opt potential opponents by shaping their preferences.  She highlights the increasingly common public persuasion campaigns associated with rulemakings.  On this point I couldn’t agree more; agencies certainly exercise this type of power, although it is perhaps even more difficult to observe than its second-face cousin.  Indeed, in the book I give a few examples that might neatly fall under this heading–such as how rulemaking proposals are framed or how agencies sometimes have a heavy hand in how they manage social media outreach.  That said, I do not describe this behavior using this framework or in as eloquent of terms as Professor Nou has done—although I certainly wish I had!

I cannot close without addressing a point that Nou makes in the concluding paragraph.  She suggests that one way to limit the procedural politicking associated with rulemaking procedures would be to randomize how rulemaking procedures are implemented.  What a wonderful idea! (I hope someone at the Office of Evaluation Sciences is reading this!) Randomization has the potential to not only limit the amount of procedural politicking associated with the rulemaking process, but would also allow scholars to study the impact of different regulatory procedures.  This kind of high-quality evidence is crucial at a time when proposals to reform the regulatory process are routinely introduced in Congress.

IV. Christopher Carrigan’s Considering Regulators in Research on Regulation

It has always seemed to me that reading the literature on regulation is a bit like trying to put together a jigsaw puzzle with pieces from different boxes. That is, one thinks they are engaged in the activity of putting together a larger portrait (of how regulation works), but the pieces just do not fit together.  This owes largely to the different disciplinary perspectives on the topic: economists talk about regulatory capture and the role of money, political scientists talk about the role of political power and incentives (guilty as charged!), public administration scholars focus on participation in rulemaking and internal agency considerations, law scholars focus on the normative and legal implications of the process, and public policy scholars home in on the effectiveness of different policy tools designed to induce regulatory compliance or analyze regulatory alternatives.  These are, of course, broad generalizations and there is often overlap among the disciplines.  But the bigger point is that, while each of these perspectives is valuable in its own right, in the aggregate we don’t walk away with a broader “Theory of Regulation.”  The sum is less than its parts.

Symposia like this one—which engages scholars from a host of different backgrounds—are a useful starting point. And Professor Carrigan is an exemplar in point.  His background is in public policy, but his own work meaningfully engages with economics and political science.  In reviewing Bending the Rules, he argues that its key contribution is its bottom-up approach to understanding the politics of the regulatory process—that is, its focus on the incentives of bureaucrats doing the rule-writing.  This, he argues, stands in contrast to much of the mostly top-down focus on studies of regulation (from multiple fields) and offers insights on potential policy reforms to the notice-and-comment rulemaking process (i.e., he makes a policy connection).

Professor Carrigan’s generous comments raise two points.  First, while there are many ways to study regulation, there are tremendous benefits to breaking from our disciplinary silos when we can.  Second, he makes the case that perhaps the key takeaway of Bending the Rules is that those who would reform the rulemaking process ought to be paying much more attention to the incentives of the regulators themselves.  This ties back to Professor Nou’s earlier post calling for greater systematic study of rulemaking reforms, perhaps through a randomization process that would enable high quality research.  The good news is that we have forums that can (and do) help realize the connection between research and reform; I’ve already mentioned the potential for linking regulatory reform to the Office of Evaluation Sciences, but the Administrative Conference of the United States (ACUS) is another key resource. The more we can engage across fields about the cost and benefits of different potential reforms, the more valuable we as scholars can be to policy makers in selecting and implementing those reforms.

V. Andrew Rudalevige’s Bending the Discipline

Fellow political scientist Andrew Rudalevige’s post situates Bending the Rules within the broader “Presidency and Executive Politics” subfield of the discipline.  He explains how this subfield has evolved and how not so very long ago a book like mine would have been excluded from the subfield.  What changed was a newer line of research that envisions the executive branch as far from unitary. Instead, the evolved view is of the executive as an amalgamation of many actors, each with their own viewpoint.  Regardless of the cause, I am glad to be included in the fold.

Reading Professor Rudalevige’s recounting of the development of the field of presidential studies, I cannot help but be reminded that we are all creatures of the times we live in (or write in).  Where will the field of executive branch politics (at least how it is studied by political scientists) head in the next five or even ten years?  I suspect that many of the methodological trends that are occurring within the broader discipline of political science—such as causal inference and text analysis—are likely to head our way.

I am excited about these developments as they have the potential to yield closer ties between theories of executive power and empirical evidence.  I hope they don’t take us too far away from the questions that have motivated the field for decades.  More importantly, I hope they serve as a conduit to our brethren in public policy, law, etc., rather than a wedge. To this point, it is imperative that we make our findings accessible to audiences that may have different training and be less familiar with these tools.

Rudalevige also pleads for a new more nuanced understanding of OIRA, one that serves as a stumbling block not just for agencies (which is the standard grievance), but also as a stumbling block for the president to overcome.  As I’d venture to guess my fellow OIRA alumni on this symposium would agree, the president can’t always control what OIRA does largely owing to the office’s focus on neutral competence—just as Rudalevige suggests.  What a wonderful book this would be, as it would complicate many existing accounts of what OIRA is and how it functions. Any takers out there?

VI. Emily Bremer’s Making the Invisible Visible

Professor Emily Bremer’s complimentary post extends the logic of procedural politicking to agency adjudication and finds… that it doesn’t fit.  She argues that adjudication exists in what she calls the “white space”—it is not governed by the same stable set of procedural norms and expectations that govern rulemaking.  Instead, adjudication is governed by “exceptionalism.”

I think Professor Bremer’s analysis is spot-on in that these considerations limit how one can study adjudication, at least in terms of the systematic empirical approach I apply to rulemaking.  I don’t have much to add except to thank her for the thoughtful exposition. However, her post raises another question: if not adjudication, then where does the procedural politicking approach offer inroads?  In the book, I talk about potential applications to other procedure-bound areas such as contracting, regulatory enforcement, and agency hiring practices.  Agency guidance documents may also present an interesting opportunity for future scholars.  Until recently these documents might have rightly have been considered more akin to adjudication in terms of their procedural exceptionalism (e.g., a recent House report described them as “regulatory dark matter”).  Recently OIRA reasserted its authority to review these documents; this move is likely to improve compliance with existing procedural requirements for these documents and perhaps to establish new requirements.  Scholars would do well to pay attention to these developments and perhaps, once a new equilibrium is established, to apply the lessons of procedural politicking there.

 VII. Stuart Shapiro’s The Revenge of the Enacting Coalition

Stuart Shapiro very effectively links my procedural politicking argument to a broader literature on “existing coalitions.” This not only distinguishes bureaucrats from enacting coalitions (those policy actors who were critical at the time a policy program was created), but it also links them to other actors in the political sphere who are attempting to influence politics contemporaneously (e.g., interest groups).  This is a great point and I only wish I had included it in the book!

Shapiro also makes the case that the Trump administration is different from past administrations.  He points out that not following procedures has gotten the Trump administration into some trouble with the courts, with decisions repeatedly overturned because the administration has flagrantly disregarded procedures.  As I’ve pointed out before, this behavior has perilous consequences for “little d” democracy.

Professor Shapiro’s point, however, is that there are limits to how far procedural politicking can be stretched.  He’s absolutely right.  The developments of the Trump era underscore an important lesson about procedural politicking: bending the rules is a way for policy actors to wield power and influence, breaking the rules, on the other hand, is likely to backfire and is, therefore, not a winning strategy.  Put simply, there are limits to what procedural politicking can accomplish.  To be effective, procedural politicking must be subtle enough to avoid pushback.  This is why “smoking gun” evidence of it is so hard to come by.

Regardless, it is clear that the ground is shifting in the world of procedural politicking.  Shapiro highlights two ways this is occurring–the increasing use of litigation to enforce procedural violations and this administration’s distrust of career civil servants.  There are more too.  For instance, in the period I study in the book (1995-2014), the Congressional Review Act was used exactly once (read Professor Shapiro’s take on that case here).  Since Trump’s election it has been successfully deployed 16 times.

The point is that we are in a moment where rulemaking norms are changing and emerging at a rapid clip. These new features do not, in my opinion, cause procedural politicking to disappear or make it irrelevant; instead they change how and when it is likely to be employed.  Bending the Rules, Part II?

VIII. Chris Walker’s Procedural Politicking and Auer Deference

Chris Walker explores what procedural politicking means for the future of Auer deference, a doctrine that calls on the courts to defer to agency interpretations of their own regulations.  From the perspective of the findings on rule drafting (in Chapter 4), he draws an important and logical connection between my focus on how bureaucratic incentives push agencies toward longer and perhaps less coherent rules and the application of Auer deference.  Walker’s post raises some terrific points—all of which resulted in vigorous head nodding as I read along—so I’ll just respond briefly on two counts.

First, Professor Walker notes that arguments based on agency incentives have been controversial in the law literature relating to Auer (he cites Sunstein and Vermeule colorfully likening the incentives argument a “phantasmal terror”).  Here, I’ll just point out that my argument is not about Auer per se (although Walker is right to draw a connection to it).  Instead, the case I make about judicial review is much broader.  As I see it, agencies factor in many different types of legal issues when writing a rule, including Auer and Chevron, but also the body of case law in the area they are regulating, and potentially also prior litigation relating to that specific rule. Simultaneously, they are factoring in the likelihood that the rule might be brought to court in the first place; most rules are not challenged and that frees agencies up considerably in those instances.  In other words, I don’t think my results necessarily challenge those that find that Auer does not change how agencies draft rules, because Auer is just one part of a larger puzzle.

Second, does procedural politicking, as I have articulated it, call for a rescission of Auer? This is a critical point, because as Walker writes, changes may be afoot.  Here, I’ll reiterate a related argument that I make in the book (and that Walker highlights): revoking Auer would change how agencies approach rulemaking.  It might encourage greater reliance on adjudication, which as Professor Bremer notes, comes with its own set of issues. And, as Walker emphasizes, there are tradeoffs to dispensing with Auer (tradeoffs that Aaron Nielson has clearly laid out). These are not trivial, and we need to be taking incentives seriously when we consider tinkering with institutional norms like deference.

IX. Bernard Bell’s Procedural Politicking or Bureaucratic Perfidy, Part II

Professor Bell’s second response in this symposium takes up two bigger picture issues about the book.    Bell first takes me to task for not engaging more deeply with the normative implications of procedural politicking.  Part of this owes to differences in disciplinary norms and training.  Much to the chagrin of many legal scholars, we political scientists tend to adopt a “just the facts, ma’am” kind of approach.

These disciplinary distinctions aside, Bell is right that I do not stay awake nights worrying about how procedural politicking is undermining the administrative state.  This can be attributed to how I perceive procedural politicking.  Where Bell sees “perfidy” and “chicanery,” I see a much more benign and natural type of human behavior.

Humor me for a brief interlude.  Imagine that you have been laboring over a project for months.  You have a sense that your boss will not approve of your approach, even though you, as the expert, feel it is the best possible way to tackle the problem.  You can hand the project to your boss for signoff late on a Friday afternoon (when she will have limited time to review it and will be itching to be head home) or you could save it and give it to her on Monday morning (when she has fresh eyes and perhaps more time to engage with it).   In this scenario, I think most of us would choose to get the approval on Friday.  Note that nothing nefarious has happened here; the rules for the signoff process are still being followed, but you have manipulated the process in a way that advantages the outcome you prefer (i.e., obtaining sign-off) and that you also believe is in the organization’s best interest.

This is how I view procedural politicking.  It would be truly worrisome behavior if the bureaucrats engaged in this were somehow out to undermine our political system or obtain personal gain.  But all the best evidence we have (e.g., see here, here, and here) suggests that bureaucrats—both careerists and appointees—tend to have pretty moderate policy preferences and are very often public-service motivated.  Now we may still want “responsive competence” of the sort that Bell argues is crucial for administrative legitimacy, but my (perhaps naïve) hope is that the broader political controls keep the ship steering in the right direction when it really matters to those in political seats.  That is, procedures are only intended to matter at the margins.

This relates to Bell’s second point, that procedural politicking isn’t always effective. I agree.  Bell offers some great examples of when it has backfired for agencies, as does Stuart Shapiro in his comments on the Trump administration.   They’re both right, but I do not see this as in tension with the broader argument.  Bureaucrats do not have perfect foresight.  They forecast how others might respond and make a strategic gambit.  Sometimes they will make bad forecasts about what others will tolerate. Sometimes circumstances change and the best-laid plans fail. And, as I note in my response to Shapiro, bending the rules can be effective, but breaking the rules is rarely a winning strategy for agencies.

X. Wrap-up

Whew!  That was a lot.  These thoughtful reviews have given me much to reflect on.  If you’re still reading along this far into this post (and into the symposium), thanks. And thanks again to all of the participants for making this such a fun and engaging enterprise.

This post concludes a symposium reviewing Bending the Rules: Procedural Politicking in the Bureaucracy, a new book by Dr. Rachel A. Potter, Assistant Professor of Politics at the University of Virginia. All of the posts can be read here.