*This is the fifteenth and final post in a series on Andrew Rudalevige’s new book, By Executive Order: Bureaucratic Management and the Limits of Presidential Power. For other posts in the series, click here.

This post allows me to check off one of my New Year’s resolutions (one more than in 2021, already!): to express my deep gratitude to the distinguished scholars and practitioners who took the time to engage with my recent book By Executive Order: Bureaucratic Management and the Limits of Presidential Power. Their generosity of time and intellect is hugely appreciated, as are their incisive comments on the book and its implications. Special thanks go to Bridget Dooling, both for organizing this symposium and for her own contribution to the conversation, touting the virtues of a “rowdy bureaucracy.”

Bureaucratic rowdiness is of course one of the central themes of the book. It seeks to highlight the role of the broader executive branch in the issuance of executive orders — to show the pluralism in unilateralism. Pundits, but presidents too, tend to describe executive action in singular terms. “Congress wouldn’t act, so I signed an executive order,” as George W. Bush said in 2004. “That means I did it on my own.”

But as the book stresses, “on my own” is more rhetoric than reality. Would-be unitary executives need to deal with the fact that the executive branch is (to borrow from poet Louis MacNeice) “incorrigibly plural.” Nikesh Jindal, a former associate general counsel at OMB — to whom I apologize for the EEOB flashbacks — noted that “almost every action taken that is described as Presidential in nature is, in reality, the byproduct of extensive back-and-forth dealings with stakeholders throughout the Executive Branch.” As several contributors observed, that pluralism is even within the White House staff itself, broadly defined, adding another degree of difficulty to presidents’ desire for control. After all, as Ken Mayer puts it, “Internalizing transaction costs by central coordination does not reduce those costs to zero” — and creating a counter-bureaucracy to manage the regular bureaucracy may add brand new ones.

In some ways, as Kristin Hickman gently suggested, the book is a triumph of the obvious: “no one should be surprised to learn” about presidents’ reliance on policy experts and interagency review. In a different scholarly world, indeed, such might be the case. But the study of executive action faded out of favor for a time in political science, only to return with books like Mayer’s pioneering With the Stroke of a Pen (my copy blooms with Post-its, abounds in corner creasing and scrawled notes-to-self). Still, as George Krause points out, in much of this work a “core premise [was] that unilateral executive powers can be equated with presidential authority.” Institutional constraints were crucial — the other branches of government had a key role in affecting executive order (EO) issuance, and public opinion was added to the mix more recently — but the wider executive branch was mostly tacit. By Executive Order seeks to deal the bureaucracy back in — to establish, as Krause has written elsewhere, that presidential-agency relations are a “two-way street.”

That conclusion was central to an earlier body of scholarship (Graham Allison’s Essence of Decisionis one early ‘70s poster child) centered on the complexity of executive branch decision-making processes and the factors aiding bureaucratic autonomy or presidential control. We are en route, I hope, back to that future, with the present project very much part of an “emerging scholarly convergence” (as Noah Rosenblum) emphasizing the importance of executive pluralism. Indeed, for me the symposium illustrates the value of conversations across disciplinary silos — between those in political science, constitutional and administrative law, and public administration (not to mention in presidential administration!)

For instance, Rosenblum wonders how we should grapple with change over time even within the bounds of a relatively constant governance structure. Does it matter that Progressives’ justification for executive control differs from that of more recent adherents to the unitary executive? In the language of American political development, how much of the change represents intercurrence, how much supersession? With what impact on the ground? (Or, I suppose, on the Oval Office rugs.) A very different question comes from Christopher Walker in noting the potential interaction of the formulation process with legal interpretation, as with Tara Leigh Grove’s work on interpretive theory (I drew on Grove’s invaluable interviews of recent EOP staff in the book). Does a pluralistic formulation process imply a more purposivist theory? For what it’s worth, my immediate instinct is to stick with the EO’s text, given the president’s ultimate responsibility to put her name to it — as the book shows, she doesn’t always do so. And it is worth keeping in mind, too, Adam White’s point that an EO may be only the starting line for the actual execution of a given policy: a plan to make plans, regulatory or otherwise. The broader point is the productive wider discussions that can grow from these cross-disciplinary seeds.

Contributors agreed that executive action was a “pluralistic process” (cf. Walker) but varied in how they evaluated the substantive and normative value of bureaucratic politics. Emily Bremer sees the presidency as empowered by bureaucracy, which is “full to the brim with information that no single person could ever individually possess” and allows for delegation even as presidents retain veto power over its recommendations. She suggests bureaucratic management enhances presidential power rather than limiting it, contra the book’s subtitle.

That management, for Jindal, lies in leveraging bureaucratic expertise, both to extract information and to generate support for later implementation. For Dooling, too, it is a plus for presidents to have to make the case for their preferences. And Boris Bershteyn, OMB general counsel and acting OIRA administrator in the Obama administration, holds that “successful presidents do not wage war on the executive branch they supervise.” He concedes the EOP is often tempted to write off agency commentary as “nitpicking, pedantry, and provincialism” but urges presidential staff to realize that nearly all agency feedback on draft EOs helps prompt helpful revisions, improving legality and administrability. He thinks the book may over-emphasize agency self-serving — if so, this is probably the result of its array of what Dooling calls “passive-aggressive memos” drawn from the archives and frankly too fun not to quote throughout the text. (One personal favorite relates to the relationship of coyotes and sheep on public lands and can’t be reproduced in a G-rated blog post. But see p. 212. The related one about “eco-freaks” is good too.)

By contrast, Shane Pennington worries about what bureaus are not telling presidents — or us. Presidents may need to keep things under wraps, but agency secrecy “is an abomination,” impeding accountability and retarding needed policymaking. From his position in private practice dealing with regulatory issues, he argues that agencies have too much sway, not too little, and that “agency control of presidential action” is far more troubling than the reverse. Meanwhile William Yeatman wonders whether the transaction costs facing presidents in this relationship are evenly distributed. Given the evidence of the Trump “resistance,” are liberal agencies too inclined and empowered to impede conservative presidents? A full answer requires scholars to carry out the rather robust “is that the presidency, or just Trump?” research agenda already underway. But it would be useful, too, to follow up on Jennifer Selin’s urgings to explore the relationship from the agency side: how and when do agencies bring information to the president’s attention? How do they “strateg[ize] to influence the use of political principals’ tools of control?” This was something I had hoped to explore in the book, in fact, before running up against a lack of consistent documentation (or simply not being able to find it — always a hazard of archival work). There are hints of agency tactics in the penumbras of my statistical analysis, consistent for instance with agencies more ideologically distant from the president proposing fewer draft EOs. Here more data are devoutly to be wished.

Yeatman’s broadest concern, shared by other contributors, concerns Congress’s abdication from its former role in administrative policymaking. If executive action makes up the bulk of the policy process, it’s problematic that a leg has dropped off the iron triangle. At the very least, as White argues, Congress needs to decide in advance what it wants presidents to do: broad statutory delegations of power require presidents and agencies to actually define a given policy before the executive branch can begin to execute. “Broader delegations produce slower administration and feebler administration,” and since presidents then change, “unsteadier administration.” Hickman, for her part, returns to Pennington’s theme of transparency, though her prescription differs: if EOs are going to be so central to the policymaking process, she argues that the process that applies to regulation — e.g., more public notification and participation and clearer cost-benefit analysis — apply to EO formulation as well. (That development might affect my answer to Walker’s question on interpretive theory.)

The symposium also teed up plenty of questions for future work. In this vein, Sharece Thrower suggests research paths that explore “the intersection of executive order formation and policy durability.” Or: how should we consider the common practice of amending existing orders — does that count as “new” policy? Or: how much does it matter that OMB, and not the White House counsel’s office (as in the Trump years), runs the clearance process? Or: does the formulation process matter for the quality of implementation? For the durability of a given EO? Arguably, EOs that get a lot of agency input should be less likely to be quickly revoked. (Though there are clearly other variables: Jindal describes his hard work gaining departmental buy-in on a 2007 order that did meet that fate in 2009.) 

Pluralism inside the EOP — easy to see, but harder to measure — is yet another topic of interest. And more broadly: how do presidents select between multiple strategies of policy change in the first place? Long ago Dave Lewis and I created a clunky flow chart as we tried to think about the relationship of politicization versus centralization. But it’s another thing altogether to systematize the relative transaction costs of acting in different policy arenas.

Thrower gets the last word, then: By Executive Order, she says, “should be only the first step in contemplating how the bureaucracy is crucial for unilateral policymaking.” I couldn’t agree more. I thank all the symposium participants for their productive prodding — and I thank them as well for all the productive interdisciplinary conversations to come. It’s great to know that none of us are, well, on our own.

Andrew Rudalevige is the Thomas Brackett Reed Professor and Chair of the Department of Government at Bowdoin College.