By Christopher H. Marraro, a partner with BakerHostetler in the firm’s Washington, DC office, and Gary C. Marfin, who was Associate Dean of the School of Engineering at Rice University and Manager of Government Relations with Conoco.


In 2018, ProPublica and The Washington Post published a research paper on Congress in which they observed:

For more than 200 years, Congress operated largely as the country’s founders envisioned — forging compromises on the biggest issues of the day while asserting its authority to declare war, spend taxpayer money and keep the presidency in check.

Today, on the eve of a closely fought election (2018) that will determine who runs Capitol Hill, that model is effectively dead.1

It was a portentous finding. The demise of the Congressional “model” has had ramifications for the other branches, for politics appears to abhor a vacuum, no less than nature.  Into the void, the Executive Branch has increasingly relied upon its own Executive Orders to set policy and actualize changes in the structure of government itself, altogether dispensing with the messy process of bicameralism.

Our aim in this Legal Backgrounder is to show 1) how Executive Orders and the administrative state have been, and are being, weaponized; 2) how such orders are usurping the policy-making role of the legislative branch; and, 3) how federal courts might play a role in returning policy matters to the legislative branch. 

We note at the outset, we are not interested in what might be termed Executive Orders issued for internal administrative purposes. Some of these may ultimately be important. President Joe Biden’s order establishing a Commission to examine the Supreme Court comes to mind.  But our interest is in those Executive Orders issued with the aim of creating U.S. policy, or those Executive Orders aimed at revoking Executive Orders issued by previous administrations.


One of President Donald Trump’s primary objectives both as a presidential candidate and early in his administration was to diminish the administrative state.  If Trump had in mind a specific agency, set of regulations, or policy area to deconstruct, he did not disclose it. The president’s aspiration seemed simply to reduce the administrative state’s size and scope and the cost associated with it, both its operating cost and the cost to the regulated community. Steve Bannon, President Trump’s chief strategist during the early days of the Administration, used the word “deconstruction,” but either way he and the President were aligned on the policy objective if not on the eschatological rhetoric. The willingness to deconstruct affected the administration’s selection of Cabinet appointees.  As Bannon explained:

Every business leader we’ve had in is saying not just taxes, but it is – it is also the regulation. … Cabinet appointees, they were selected for a reason and that is the deconstruction, the way the progressive left runs, is if they can’t get it passed, they’re just gonna put it in some sort of regulation in – in an agency. That’s all gonna be deconstructed and I think that that’s why this regulatory thing is so important.2

What does it mean to say ”that’s all gonna be deconstructed?” We would come to learn that, at minimum, deconstruction referred to Executive Orders, issued in the early days of Trump’s administration, which were designed to reduce the size, scope, and cost of the administrative state.  Emblematic of the deconstruction Orders was 13771.  This Order, published in March of 2017, gained more prominence than any other targeting the administrative state.3  The Order reads in part:

Sec. 2. Regulatory Cap for Fiscal Year 2017. (a) Unless prohibited by law, whenever an executive department or agency (agency) publicly proposes for notice and comment or otherwise promulgates a new regulation, it shall identify at least two existing regulations to be repealed.4

Four years later, after the rhetorical smoke cleared, the administrative state, far from being deconstructed, was, at best, selectively diminished. When the Office of Management and Budget5 reported on the results of Trump’s deregulatory efforts, it found a total of $198.6 billion in savings government-wide from FY2017 to 2020.  Agencies issued 538 deregulatory actions and 97 regulatory actions for a ratio of 5.5 to 1.0 over the period. Comparing significant deregulatory to significant regulatory actions in 2020 yielded a ratio of 1.3 to 1, short of the 2 for 1 objective, but a slight decline in regulations just the same.

Although Trump referenced the administrative state as a whole, the “deconstructions” were rather confined. Remarkably, the vast bulk of the savings, as illustrated in the chart below, came from two agencies: the Department of Transportation and the Environmental Protection Agency. Together these Agencies accounted for $189.7 billion, or 95.5% of the total $198.6 billion, in savings over the period.

Taking a more granular look at deregulatory activity under Trump, Cary Coglianese and others from the University of Pennsylvania Law School, examined data from the semiannual edition of the regulatory agenda from Fall 2017 to Spring 2020.  They were interested not in regulations that were withdrawn at the proposal stage, such as regulations having imposed no prior compliance costs, but on actions completed by promulgation. In addition, they focused on regulations that were categorized as economically significant under the long-standing designation for such regulations. They found that “those actions designated as ‘deregulatory’ make up only one quarter of the overall number of completed actions within the regulatory agency,” leading them to conclude:

This means that, rather than there being more deregulatory actions than other actions, as the Trump Administration’s claims have implied, there was, in fact, just the opposite. Overall about three completed actions in the regulatory agenda appear for every action designated as deregulatory.6

Trump’s deconstruction efforts may have put a dent in the administrative state, but little more. As Clyde Wayne Crews Jr., a policy director at the Competitive Enterprise Institute, observed: “when all is said and done the administrative state cannot be said to have fundamentally changed under Trump.”7  Similarly, Stuart Shapiro, Professor and Associate Dean of Faculty at the Edward J. Bloustein School of Planning and Public Policy at Rutgers University, maintains that, “all in all, the impact of the deregulatory actions of this Administration is quite likely to be negligible on balance.”8

More ink may have been spent trying to determine if Trump satisfied his two-for-one target than in whether the target was worth attaining in the first place.  Consider first what E.O. 13771 and Trump’s overall deconstruction efforts neglected: specifically the existence of benefits.  Trump’s sole focus on cost in the deconstruction calculus might well have been the result of the Administration’s discussions with the business community, the members of which are likely much more aware of the costs imposed by regulations than any benefits gained from them. 

A second problem confronted Trump’s deconstruction efforts: like other administrations before him, Trump’s administration needed the administrative state. The war on the administrative state was partly real, but partly rhetorical.  Presidents invariably rely on the administrative state. Trump certainly did. As Crews puts it: Trump “sported regulatory impulses of his own” in immigration, antitrust and elsewhere.9  In fact, Trump, rather than taking aim at the totality of the administrative state, could have been much more selective in identifying targets for deconstruction, as we’ll see below. 

This brings us to the final obstacle on the road to deconstruction. While the effort to deconstruct the administrative state was no doubt applauded in some quarters, it was never embraced as a nationwide campaign; there was no national mandate. This is partly because not all parts of the administrative state are viewed as equally problematic by all members of any given party. Exactly which parts of the administrative state are problematic can vary along partisan lines and within partisan groups. As Jacobs, King, and Milkis observed: 

Consequently, while liberals seek to build administrative capacity to design and implement social welfare policies, conservatives have sought to redeploy and extend that power in pursuit of their own partisan goals: enhancing national defense, homeland security, border-protection, and local policing; and establishing more market-oriented policies in education, climate change, and government service.10   

As we have seen, a wide range of different groups may have, at some level, supported deconstruction, but is a majority of the electorate willing to dismantle the whole edifice?  While deconstruction may have garnered its share of applause at political rallies, it’s quite likely that different people were applauding the fall of different elements of the administrative state, even as they appeared to applaud in support of the state’s overall demise. 


An Executive Order can be issued quickly and, apart from internal reviews, virtually unilaterally. The unilateral attribute is especially advantageous during periods of significant partisan divisions.  The dilemma, of course, is that Executive Orders issued by one President can be revoked by subsequent President(s) and, for all practical purposes, revoked just as easily as they were issued.  In his first 100 days, for example, President Biden signed “more than 60 executive actions, 24 of which are direct reversals of Trump’s policies.”11

One of the principal policy divisions between Biden and Trump concerned the role of the administrative state itself. Trump viewed the administrative state as little more than a cost center, a burden on the business community and the economy. Hence, E.O. 13771 and the accompanying orders aimed at deconstruction.  Far from President Trump’s declaration that the administrative state should be deconstructed, if not outright destroyed, President Biden claimed that the administrative state was essential in addressing the challenges facing the nation:

To tackle these challenges effectively, executive departments and agencies must be equipped with the flexibility to use robust regulatory action to address national priorities.  This order revokes harmful policies and directives that threaten to frustrate the Federal Government’s ability to confront these problems and empowers agencies to use appropriate regulatory tools to achieve these goals.12

These differing perceptions of the administrative state cannot be resolved through Executive Orders.  Admittedly, it’s possible that reconciling differing perceptions of the administrative state is a fool’s errand anyway; the administrative state is so large that its deconstruction, while it may resonate with some voters, is not a practical national goal. Nevertheless, assuming these differing perceptions merit examination, such an examination requires a broad-based debate, dialogue, and discussion, and involve a far broader range of stakeholders than can be found in a single presidential administration.  Executive Orders, in contrast, are “temporarily definitive” statements.  They are orders from the President; as such, they are not in the marketplace of ideas.

In issuing his various Executive Orders, Biden apparently did not view himself as creating policy.  Cognizant of the Executive Orders he had issued that revoked prior Orders, Biden explained:

And I want to make it clear — there’s a lot of talk, with good reason, about the number of executive orders that I have signed — I’m not making new law; I’m eliminating bad policy.

The obvious question is: bad policy according to whom? Or, as the philosopher John Locke asked of the exercise of political authority in general, “Who shall be judge when this Power is made right use of?”

Judicial Moats Protecting the Agencies

One structural criticism of the administrative state is that it thwarts the Constitution by replacing legislative action with agency rulemaking. 

Fortunately, the Administrative Procedure Act (APA)13 and its attendant “arbitrary and capricious” standard of judicial review has propelled the development of a body of law that constrains “free for all” and irrational agency decision making by imposing a duty on agencies to “examine relevant data” and to “articulate” a “rational connection between the facts found” and the agency’s policy choice.14 But nearly thirty years ago the U.S. Supreme Court held that a President’s Executive Orders are not subject to the protections of the APA.15

That does not mean that Executive Orders are insulated from judicial review. They are not. The Supreme Court has long held that Executive Orders are subject to review under the Due Process Clause of the Fifth Amendment. Although Executive Orders are subject to judicial review, a crucial question given that they lie beyond the reach of the APA is what should be the appropriate standard under which a court reviews the reasonableness of Executive Orders. Should it be mere “rationality” as when a court reviews the reasonableness of legislation and upholds its reasonableness if the court can conceive of a rationale, or should review entail the more exacting “arbitrary and capricious” standard. At a time of proliferating Executive Orders resulting from a legislative stalemate in Congress, this critically important issue surprisingly has received only scant debate and sparse scholarly review.16 Judicial review of Executive Orders even has been described by some commentators as “disordered.”17

We contend that courts should subject Executive Orders to the more exacting “arbitrary and capricious” reasonableness standard. They should ensure that policy should be clearly explained in the Executive Order and the policy choice made there should be rationally related to facts and consistent with any operative statutes. We think that the recent preliminary injunction issued against President Biden’s Executive Order 1408 and the Department of Interior’s oil and gas lease “Pause” is exactly the type of prudent judicial review required. There, Judge Terry A. Doughty found that Executive Order 1408 failed to provide “any explanation for the Pause.”18 It seems to us that in this highly charged political environment where Congress is more often than not deadlocked on enacting social and economic legislation, the federal courts should stand as a strong arbiter of whether the rationale of Executive Order’s policy is based on sufficient explanation, adequate facts and is supported by the statute under which it is issued. Only then can we have confidence that such executive action in the absence of Congressional action is not an evasion of statutory policy or the product of irrational decision making. 


  3. The “deconstruction” Executive Orders also included: E.O. 13777, 13875, 13891, 13892, and 13893.
  4. Counting regulations is not universally appreciated.  Excellent reasons are provided by Ms. Short below: “First, it undermines the achievement of statutory goals. Second, it is costly and wasteful. Third, it crowds out meaningful dialogue and research about the real and difficult problems of regulation.“  See Jodi L. Short, The Trouble with Counting: Cutting through the Rhetoric of Red Tape Cutting, 103 Minn. L. Rev. 93 (2018).
  8. Professor Shapiro, it should be noted, is associated with  a publication affiliated with the Penn Program on Regulation. See Stuart Shapiro, Making Sense of the Trump Administration’s Regulatory Numbers, Jan. 14, 2020.
  9. Crews, supra n. 7.
  10. Nicholas F. Jacobs, Desmond King, and Sidney M. Milkis, Building a Conservative State: Partisan Polarization and the Redeployment of Administrative Power. Perspectives on Politics, 17(2) (May 13, 2019), 453-69, DOI.
  13. 5 U.S.C. §§ 551 et seq.
  14. See Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Ins. Co., 463 U.S. 29, 33-34 (1982).
  15. Franklin v. Massachusetts, 505 U.S. 788, 800-01 (1992).
  16. See David M. Driesen, Judicial Review of Executive Orders’ Rationality, Boston U. Law Rev., Vol 98:1013 – 66 (2018).
  17. See Erica Newland, Executive Orders in Court, 124 Yale L.J. 2026, 2045-46 Note 10 (2015) (observing that judicial review of executive orders is “disordered”).
  18. State of Louisiana et al. v. Joseph R. Biden, Jr. et al., 2:21-CV-00778 (June 15, 2021) (D. LA.).