Law Review
Northwestern Law
Northwestern University Law Review : Colloquy : 2009 : Beermann

Combating Midnight Regulation

Jack M. Beermann[*]

[download pdf]

The flurry of regulatory activity by the outgoing administration of President George W. Bush has raised, once again, the specter of midnight regulation.[1]  In contrast to the late-term action of the Clinton Administration, much of the Bush Administration's late-term action seems to have been more deregulatory than regulatory, but from a political and legal standpoint, that distinction may not make much, if any, difference.  While midnight regulation provokes an instinctively negative reaction, it is not completely clear what is wrong with it.  This uncertainty arises in part because of the different reasons for midnight regulation.  In my earlier work on this subject, I identified four possible reasons for late-term action, and in this Essay I add a fifth, although I confess a lack of knowledge on whether this fifth reason is actually a significant factor in midnight regulation.  The original four are: (1) the natural human tendency to work to deadline, which has been referred to in the literature as the "Cinderella constraint;"[2] (2) hurrying to take as much action as possible near the end of the term to project the administration's agenda into the future; (3) waiting to take potentially controversial action until the end of the term when the political consequences are likely to be muted; and (4) delay by some external force that prevented the administration from taking desired action until late in the term.  The fifth, and new, possible reason for late-term action I call "timing."  Timing is a form of waiting, not based on potential negative consequences, but rather, based on the desire to achieve something positive before the presidential election in order to help either one's own reelection bid or the election prospects of the incumbent party.  One can imagine, for example, the President delivering an October surprise of favorable regulatory action for the automobile industry if Michigan looks like a swing state in the upcoming election.

Whatever the reason for midnight regulation, there seems to be a general perception that something has gone wrong when an outgoing administration takes important action while the incoming administration is waiting to take over.  Most late-term action is subject to the obvious question of why, if the regulation was deemed so important, the administration failed to act during the previous three or seven and three-quarters years.  The lines of normative critique of midnight regulation are fairly evinced by each of the factors posited above.  Thus, even though the Constitution leaves the incumbent in office for approximately eleven weeks after election day, the public feels uncomfortable when an outgoing administration waits until late in the term to take politically controversial action or loads up on late-term actions to project its policy preferences in the future.

There is one consequence of midnight regulation that may not be completely obvious and should be highlighted.  Especially as our collective experience with midnight regulation has grown, the outgoing President knows that the incoming administration is likely to look carefully at late-term actions by the outgoing administration.  President George W. Bush certainly had plenty of experience with the time and energy it took for his administration to freeze and then review dozens of late-term actions taken by the Clinton Administration.[3]  Some late-term action is so likely to be overturned by the incoming administration that the outgoing administration may have acted merely to embarrass the new President or force the new President to expend political capital on the matter.  Before they act, outgoing administrations should take into account the distraction and energy necessarily associated with reviewing late-term actions.  The President takes an oath to "faithfully execute the office of President of the United States,"[4] and the outgoing President arguably violates that oath if he overloads the incoming administration with midnight rules and other late-term actions that impede the incoming President's ability to "take care that that the laws be faithfully executed" immediately upon taking office.[5]

Taking into consideration the factors discussed above, this Essay examines possible ways to combat midnight regulation.  The discussion begins with a recent proposal in Congress to restrict rulemaking activity during the last ninety days of an outgoing administration by giving the incoming administration the power to "disapprove" of regulations adopted during those final ninety days.  Part I of the Essay explains the bill and identifies potential problems with it.  Part II offers two alternative approaches: one involves a simple reform to administrative law, and the other outlines statutory proposals that differ from the bill proposed by Representative Nadler.

I.  Representative Nadler's Proposal

On January 6, 2009, Representative Jerrold Nadler of New York introduced H.R. 34.[6]  The fundamental provision of the bill is that no rule adopted in the final ninety days of an outgoing administration (a "midnight rule") can go into effect until ninety days after the appointment of a new agency head by the new President.[7]  The newly appointed agency head may, during his or her first ninety days in office, disapprove of a midnight rule by publishing a notice of disapproval in the Federal Register and providing a notice of disapproval to "the congressional committees of jurisdiction."[8]  However, if the new agency head takes no action within ninety days, the midnight rule goes into effect.

Additionally, the bill contains provisions that would enable the outgoing President to put a midnight rule directly into effect.  The outgoing President can do so by declaring in an Executive Order, with separate written notice to Congress, that the rule is "necessary because of an imminent threat to health or safety or other emergency; necessary for the enforcement of criminal laws; necessary for national security; or issued pursuant to a statute implementing an international trade agreement."[9]  The bill further provides that it applies retroactively to all rules issued after October 22, 2008, which is ninety days before President Barack Obama's inauguration.  As these broad provisions might suggest, even a cursory reading of the bill reveals that it has some relatively serious drafting problems.  Although an exhaustive enumeration is not possible in this Essay, the following discussion focuses on a few operational issues that demonstrate why the bill is a less than ideal solution to the midnight regulation problem.

A.  Blanket Delay

The provision, imposing a blanket delay of the effective date of all rules adopted in the last ninety days of an outgoing administration until ninety days after the appointment of a new agency head, creates several problems.  The bill's language does not provide exceptions for instances in which the incoming administration would rather have the midnight rules go into effect, for example if the incoming administration is of the same party or in which midnight rules were the product of cooperation between the incoming and outgoing administrations.  There is no provision for allowing an incoming President to allow all or some midnight rules to go into effect immediately.[10]  The bill grants only the outgoing President the authority to place midnight rules directly into effect and only if certain conditions are met.[11]  If those conditions are not met, or if the outgoing administration would prefer to leave the issue to the incoming administration, the bill's lack of flexibility would be problematic.  In some circumstances, and perhaps especially in times of crisis when quick and decisive action is necessary, incoming and outgoing administrations may work together without regard to the election and inauguration cycle.  In such cases, it would likely be beneficial to allow the incoming administration to allow midnight rules to go into effect immediately. 

The blanket delay of midnight rules may not be as politically beneficial to the incoming President as one might think.  For example, the phenomenon of "waiting" until after election day may enable an outgoing President to take actions that might benefit the incoming administration with less regard for political consequences.  An outgoing President acting responsibly may take difficult action at the end of the term to pave the way for a smooth transition.  By automatically delaying the effective date of rules in this category unless the outgoing President uses the authority discussed above to advance the effective date, the bill would increase the political costs to the outgoing administration and thus discourage this sort of cooperative action.

B.  New Agency Head

Another problem related to the blanket delay, perhaps unlikely to occur in most transitions, is that the bill could place midnight rules into limbo for extended periods of time.  The effective date for midnight rules is ninety days after the appointment of a new agency head, during which time the agency head has the authority to disapprove the rule.  There is no provision for a holdover agency head, that is, if the incoming President chooses to keep the outgoing administration's agency head in place, the bill could be read to preclude rules issued in the last ninety days of the outgoing administration from going into effect for many months or years.  This may seem to be a hyper-technical reading of the bill, but a subject of a midnight rule may have a legal argument that the rule has not gone into effect if there is no new agency head.  This is obviously an unintended consequence—no one wants to discourage a new President from keeping agency heads from the prior administration in office.  However, a lengthy delay could also result if the appointment of a new agency head encounters a confirmation problem or if the incoming administration falls behind in naming new agency heads.  Of course, if one views midnight rules as virtually always a scourge, then the potential for indefinite delay will not be viewed as much of a problem.  Nevertheless, it does not seem likely that the author of the bill intended to create the possibility of indefinite delays in the effective dates of rules.

Automatically delaying all midnight rules until the appointment of a new agency head could also cause political problems for an incoming administration.  The incoming administration would find itself in the potentially uncomfortable position of appearing responsible for every rule issued in the last ninety days of the outgoing administration.  Newly appointed agency heads may then have to spend their first ninety days reviewing midnight rules rather than beginning to work on the new President's agenda.  This aspect, however, may be seen by some as a virtue.  With the automatic delay of all midnight rules, the incoming administration can at least delay the day of reckoning until the ninetieth day after the appointment of a new agency head.  If the incoming administration was granted discretion to pick and choose among midnight rules to delay, intense pressure may be brought to bear at the very outset of the administration, and the new agency heads may be forced to act with great haste in a tense environment.  Thus, it is a judgment call whether discretion is better or worse than an automatic delay, and for reasons discussed in Part II B, in my judgment the potential benefits of discretion arguably outweigh the potential costs.

C.  Rules That Should Be Exempt

Another general problem with the bill is that it fails to account for rules for which a delay may be legally questionable or unnecessary.  First, there is no indication that rules required by statutory deadlines or court orders are exempt.  In their transition-related regulatory review instructions to agencies, Presidents George W. Bush and Barack Obama both exempted rules required by statutory or judicially imposed deadlines.[12]  Representative Nadler's proposal would create a potential conflict with regard to such rules, and the bill should either exempt them or at a minimum clarify whether the bill's intent is to delay rules' effective dates beyond deadlines established by statute or court order.  Second, there is no mention of rules not subject to the notice and comment procedures of the Administrative Procedure Act (APA).  Is the intent to include personnel rules and rules relating to government contracts, grants, and benefits, which are completely exempt from APA § 553?  What about interpretative rules, policy statements and guidance documents, which are exempt from § 553's notice and comment provisions?  There are good reasons, discussed in Part II below, for not including rules that the new administration could easily alter without notice and comment.  The bill should at least address this issue. 

D.  Disapproval Only

A related procedural problem with the bill, equally if not more significant, is that the incoming administration's only option is to disapprove the midnight rule or allow it to go into effect as written.[13]  There is no option to revise a midnight rule.  This means that if the new agency head concludes that some sort of rule is necessary, even one that is very close but not exactly the one promulgated by the prior administration, the agency must either accept the imperfect rule or engage in expensive and time-consuming rulemaking proceedings to promulgate what might be an only slightly different rule.  This can be a real waste if the rulemaking record produced by the prior administration is likely to support the rule preferred by the new administration, given that the record will still be fresh.  Thus, the bill could lead either to waste or to a new administration allowing a suboptimal rule to go into effect.

E.  Independent Agency Rules

Another significant problem is that the bill is not very specific about which agencies it covers, and there are good reasons to not extend coverage of the bill to independent agencies.  Because the bill would be inserted into the APA as 5 U.S.C. § 555a, presumably the APA's definition of "agency" would apply.  This broad definition includes independent agencies such as multi-member commissions and the National Labor Relations Board.[14]  The bill does not explicitly exempt independent agencies from its coverage, but the bill is not well-designed for application to them.  First, the bill does not identify an "agency head" for a multi-member agency.  Is it the agency chair, or a majority of the agency?  Since new Presidents upon taking office do not normally appoint new agency heads for independent agencies, if the bill is construed to apply to such agencies, it could force independent agencies to forego issuing rules for the last ninety days of each presidential term.  Second, the midnight rule problem is not as serious with regard to independent agencies because they are bipartisan and because their members serve for terms of years that do not coincide with the presidential election cycle.  Out of loyalty to the President, independent agencies dominated by the incumbent's party might time certain actions with the political consequences to the President in mind or to avoid rejection by the next Congress.  However, because independent agencies are not subject to direct supervision by any executive branch official, they are less likely than executive branch agencies to be acting for the reasons that contribute to the general disfavoring of midnight rules.  Midnight rule reform is simply not needed for independent agencies, and the bill should address this.[15]

F.  Other Procedural Problems

There are a few other potential procedural problems with the bill which, though less likely to cause problems, are still worth addressing.  The first concerns the use of the word "adopted" in the definition of "midnight rule."  A midnight rule is "a rule adopted by an agency within the final 90 days a President serves in office."[16]  The problem is that the bill does not contain a definition of the word "adopted."  There is no indication as to what precise event demonstrates that an agency has adopted a rule.  There are several possibilities, including publication in the Federal Register, submission to the Federal Register for publication, signing a paper indicating that the agency has adopted a rule, and so on.  Any bill on this subject should make it crystal clear what it means for a rule to be "adopted."

Further, the definition of "midnight rule" creates two situations in which it will not be immediately known upon adoption whether a rule is a midnight rule.  One situation is where the President is running for reelection, and it will not be known until after election day whether a rule adopted before election day and less than ninety days before January 20 is a midnight rule.  This is not a serious problem, because it throws into uncertainty only rules adopted in the last two weeks before election day, and none of these rules will have yet gone into effect since no rule's effective date can be less than thirty or sixty days after adoption, depending on whether it is a major rule subject to the Congressional Review Act.[17]  Nonetheless, it seems unlikely that the drafters anticipated the situation in which an agency can adopt a rule without knowing whether it will go into effect as stated in the rule (in as little as thirty days) or not until ninety days after inauguration day, which could amount to a delay of 180 days.

The second situation of uncertainty involves the potential for an unexpected application of the bill when the President does not complete the term, because of either death in office or resignation.  In such cases, the bill would seem to convert to midnight rules all rules issued in the ninety days before the President left office, delaying their effective date until ninety days after the new President (normally the incumbent Vice-President) appoints new agency heads.  Perhaps midnight regulation concerns exist if a President is in a political fight that appears headed for impeachment or resignation (think, for example, of Illinois Governor Rod Blagojevich's appointment of Roland Burris to fill Barack Obama's Senate seat while under indictment and subject to impeachment proceedings) but occurrences like this are exceedingly rare.  Given that the new President, being the incumbent Vice President, is almost certainly of the same political party as the prior President and may retain, at least for a time, many of the prior President's agency heads, midnight regulation should not be much of a concern, and application of the bill would lead to the problem of prolonged delay discussed above when no new agency head is appointed.  Thus, the bill as written could require 180 days of uncertainty even when no one thinks that the prior President had acted for any of the reasons we normally attribute to the midnight regulation problem.

For these reasons and more, the bill proposed by Representative Nadler is not an attractive vehicle for midnight rule reform.  However, this does not mean that some form of midnight regulation reform is not desirable or possible.  The next Part proposes some possibilities, some of which build upon Representative Nadler's proposal.

II.  Midnight Regulation Reform Possibilities

This Part raises two different tacks for dealing with midnight regulation.  The first section addresses a particular feature of administrative law that makes it difficult for an incoming administration to repudiate late-term action by the outgoing administration.  Basically, under current doctrine, once a rule goes into effect, any change must be justified by good reasons even if the original rulemaking record would have justified a different rule.  In the context of midnight regulation, this is an unfortunate requirement.  The Supreme Court could loosen up on arbitrary and capricious review in cases involving changes to rules in transition periods, or even more generally in cases where the change is made soon after the issuance of the original rule, or Congress could legislatively provide for more deferential review.  The second section outlines a statutory model that would grant the incoming administration the power to review and reject or modify late-term action by the outgoing administration, similar to the model that the administrations of Presidents George W. Bush and Barack Obama followed when they confronted the mass of midnight regulation left behind by their predecessors.  This model is not different in principle from the bill proposed by Representative Nadler, but it employs a more nuanced approach that is more sensitive to the political and legal realities of midnight regulation.  Finally, a simpler statutory approach is raised, which would allow agencies at all times to revise or rescind recently issued rules when unexpected negative feedback erupts after a rule is issued.

This discussion assumes that some sort of reform to make it easier to combat midnight regulation is desirable.  However, it should be noted that even if it is agreed that midnight rules tend to be problematic, it is not completely obvious that reform is necessary.  Presidents may already have sufficient tools to deal with midnight regulation, as demonstrated by action taken by the administrations of Presidents Ronald Reagan, Bill Clinton, George W. Bush, and Barack Obama to combat the midnight regulatory activity of their respective predecessors.[18]  For example, in a memorandum very similar to the one issued at the outset of the prior administration, on his first day in office President Obama directed his administration not to issue any new rules until his appointees had a chance to review them, to withdraw from publication any proposed or final rules that had been sent to the Federal Register but not yet published, and to consider extending the effective date of published rules that had not yet gone into effect so that a new appointee could review them.[19]  With regard to published rules, President Obama's order directed that if the agency decided to delay the effective date, it should immediately reopen the comment period for thirty days to allow public comment on whether changes should be made before the rule is allowed to go into effect.

Although incoming Presidents already have tools to combat midnight regulation, those tools may not always be up to the task.  While the short duration of a temporary suspension may often preclude resolution of a legal challenge while the controversy remains live and justiciable, in a decision involving the suspension of a midnight rule by the Bush Administration, the Second Circuit held that once the rule was finalized and published in the Federal Register, in this case on January 22, 2001—two days after President George W. Bush took office—the rule had taken effect and could not be suspended or altered without notice and comment.[20]  Thus, it appears that there is at least a need for reform in the possibly rare situation in which action against midnight rules is subject to legal challenge.

A.  Administrative Law Reform

The outgoing administration of G.W. Bush was careful to attempt to shield its late-term output from revision by the incoming Obama Administration.  Josh Bolten, President Bush's Chief of Staff, issued a memorandum on May 9, 2008, instructing agencies not to propose any new rules after June 1, 2008 and to finalize all rules by November 1, 2008.[21]  The Bush Administration portrayed this as taking the high road against midnight regulation by avoiding the unseemly specter of rules being published in the Federal Register up to and even past inauguration day, as had occurred at the end of the Clinton Administration.  In truth, it was part of an effort to shield its midnight rules from the types of actions that the Bush Administration had taken against President Clinton's midnight rules.  Rules completed more than sixty days before inauguration day would be final and thus not subject to a freeze or easy process of revision by the new administration.  Thus, Bolten was actually providing agencies a roadmap for avoiding what the Bush Administration was able to do to some of the Clinton Administration's midnight rules.

The Supreme Court's application of the arbitrary and capricious standard to rescission and revision of rules has created some of the difficulties that incoming administrations encounter when trying to undo midnight rules.  Under Motor Vehicle Manufacturers Ass'n v. State Farm Mutual Automobile Insurance Co.,[22] once a rule becomes final, rescission or revision must be justified by reasons for the rescission or revision even if the original record would have supported a different rule or a decision not to adopt any rule.[23]  This understanding gives an administration a powerful tool, for better or for worse, to project its agenda beyond the end of its term.

In State Farm, the Carter Administration's rule requiring passive restraints in passenger cars was promulgated in 1977, the first year of the Carter Administration, but it did not require any car to be equipped with passive restraints until 1982, the second year of the next presidential term.[24]  This put the Reagan Administration in the uncomfortable position of being the first administration to enforce this regulation even though it did not agree with the regulation and had actually campaigned on a deregulatory platform.  When the Reagan Administration's new agency head attempted to rescind the requirement, the Court did not ask whether the original rulemaking record would have supported a decision to make no rule at all.  Rather, it asked whether there was sufficient new evidence or policy reasons for rescinding the rule that had been adopted.[25]  As soon as the initial regulation was adopted, it became the baseline for evaluating future agency action even though it had not actually gone into effect.[26]

This application of the arbitrary and capricious standard should be reexamined.[27]  At least with regard to rules that can fairly be characterized as midnight regulation, when a new administration takes office, it should have the freedom to revise or rescind rules that were adopted by the prior administration if the original rulemaking record would have supported the new administration's decision.[28]  The baseline should be the statute and regulatory situation before the outgoing administration acted.  The possible counterargument is that this idea pays insufficient heed to the role of agency expertise and makes regulation look overly political.  This objection, however, presumes that midnight regulation is less political than an incoming administration's potential reactions to the regulation.  To the contrary, there are good reasons to believe that a great deal of late-term administrative action is political. Moreover, it is also likely, that in the rush to deadline, even action motivated largely by expertise will contain more errors than action taken under less time pressure.  Thus, there are certainly practical reasons for affording a new administration the freedom to revise or rescind midnight rules when the record would support such a change. 

Additionally, two developments in administrative law lend support to the idea that a different application of the arbitrary and capricious standard to midnight regulation is possible within the constraints of the law.  The first is the recent recognition, under the Chevron[29] rubric, that agencies are free to reinterpret statutes even when a prior agency interpretation has been upheld on judicial review.[30]  Under step two of the Chevron doctrine, which is reached when an agency interprets an ambiguous statute, a court defers to any reasonable statutory interpretation.  The Court's decision in National Cable & Telecommunications Ass'n v. Brand X Internet Services made clear that because Chevron's step two is a very deferential standard of judicial review, under which various interpretations might have been upheld, agencies are free to alter their interpretations as long as the new interpretation is also reasonable.  The baseline against which the new interpretation is judged is not the initial interpretation but the statute being interpreted.

The analogy between an agency's ability to reinterpret and a new administration's ability to revise or rescind is strengthened by the fact that the Court understands that interpretation under Chevron is not simply a quest for the best understanding of the words used by Congress, but instead requires choosing among plausible interpretations based at least in part on considerations of policy.  The Supreme Court recognized this in Chevron when it stated:

In these cases, the Administrator's interpretation represents a reasonable accommodation of manifestly competing interests, and is entitled to deference: the regulatory scheme is technical and complex, the agency considered the matter in a detailed and reasoned fashion, and the decision involves reconciling conflicting policies. . . .  While agencies are not directly accountable to the people, the Chief Executive is, and it is entirely appropriate for this political branch of the Government to make such policy choices—resolving the competing interests which Congress itself either inadvertently did not resolve, or intentionally left to be resolved by the agency charged with the administration of the statute in light of everyday realities.[31]

In its Brand X decision, the Court reiterated this understanding, stating that filling statutory gaps "involves difficult policy choices that agencies are better equipped to make than courts."[32]  Thus, the Chevron doctrine—as applied in Brand X to allow an agency to adopt a new, reasonable, interpretation—represents an instance of policy alteration in which the baseline is the original statute and situation, rather than the intervening policy that is being repudiated.  There is no apparent reason why an incoming administration should not enjoy the same freedom to repudiate the late-term actions of its predecessor as long as the new administration pursues its policies within the limits imposed by the statutory framework.

Chevron, together with the additional instances detailed below, also stands for the proposition that the application of the arbitrary and capricious standard can vary with the circumstances.  The APA standard of judicial review that applies to agency rules is the arbitrary and capricious standard.[33]  The Chevron opinion paraphrased the statutory standard but then actually constructed what appeared to be a new standard applicable to agency statutory interpretation.  The Chevron Court stated that "legislative regulations are given controlling weight unless they are arbitrary, capricious, or manifestly contrary to the statute."[34]  This is a paraphrase of the APA's "arbitrary, capricious, an abuse of discretion or otherwise contrary to law."[35]  The Court inserted the word "manifestly" apparently to indicate a degree of deference to agency legal interpretations that does not appear to be provided for in the statute as written.[36]

In at least two subsequent situations, the Court has explicitly endorsed a more deferential application of the arbitrary and capricious standard based on the context of the agency action involved.  In Massachusetts v. EPA,[37] which rejected the EPA's refusal to regulate global warming gases, the Court explained that agency decisions rejecting rulemaking petitions are subject to judicial review under the arbitrary and capricious standard, and quoted with apparent approval the D.C. Circuit's rule that such decisions are reviewed on a very deferential standard: "Refusals to promulgate rules are thus susceptible to judicial review, though such review is 'extremely limited' and 'highly deferential.'"[38]

This highly deferential version of the arbitrary and capricious review is similar to the standard that the Court has applied when reviewing agency refusals to bring enforcement actions (when such refusals are reviewable because governing law contains clear criteria for when enforcement is required).  In such cases, the Court has stated that review should be very narrow, that the reviewing court should ordinarily look only at the statement of reasons the agency has provided for not bringing an enforcement action, and that the reviewing court should not even consider the factual basis for the decision or facts offered in opposition to the decision.[39]  This is a far cry from the usual practice in cases applying the arbitrary and capricious standard in which the reviewing court looks carefully at the facts and reasoning underlying the decision in what has been denominated "hard look" review.[40]

The lesson to be learned from these examples is that the arbitrary and capricious standard is not a static principle that applies uniformly in all contexts.  Rather, the Court has found it appropriate to adjust the scope of review when it finds a different standard more appropriate.  While the Court in State Farm rejected the argument that review should be relaxed when an agency deregulates, that decision does not necessarily reject the desirability of an adjustment when an incoming administration acts to deal with midnight regulation by its predecessor.  Efforts that enable an incoming administration to easily alter a previous administration's late-term action may discourage midnight regulation in the first place.

Although the focus here is on midnight regulation, this proposal for reform should probably not be confined to revisions or rescissions by a new administration, but might be extended to increase agency flexibility whenever an agency changes its position shortly after issuing a rule.  If an agency issues a rule that causes an outcry, either right after issuance or perhaps as a more distant effective date approaches, there are good reasons for allowing the agency to make revisions without having to use the initial rule as a baseline.  At this early stage, the agency ought to be able to reshape the rule in any way that would have been supportable based on the original rulemaking record.  In addition to preserving a measure of flexibility, this would also have a desirable procedural effect.  The Supreme Court adverted to the "inherent advantages of informal rulemaking" when it rejected the argument that courts have the power to require additional procedures in complicated or important rulemaking proceedings.[41]  Along these lines, flexibility in the period immediately following notice and comment would enable an agency to issue a rule after a single notice and comment period, knowing that if the rule provokes an unanticipated outcry, revisions supportable under the original record could still be made.  Without this flexibility, agencies might be more reluctant to make desirable revisions if such revisions would require not only a second round of notice and comment, but also greater justification.  They might be more cautious during the initial rulemaking proceeding, perhaps by adding a second notice and comment period if there is any perceived likelihood that a rule will provoke a strong negative reaction.  While some people might think more notice and comment would be a good thing, the Supreme Court's preference is to follow Congress's model of a short and sweet notice and comment process.[42]

B.  Alternative Statutory Possibilities

If the Supreme Court declines the invitation to build a solution for the midnight regulation problem into administrative law, another possibility is for Congress to create a statutory solution along the lines of Representative Nadler's proposal, but without the defects of that particular bill.  The key would be to design a proposal that discouraged the outgoing administration from engaging in midnight rulemaking, that gave the incoming administration effective tools to deal with it, or both.  Despite its limitations, Representative Nadler's bill could serve as the basis for a better proposal if the problems identified in the first Part of this Essay were remedied.  Most of the problems with the bill could be solved with a few simple changes.  Below are five changes that directly address the deficiencies highlighted in the Nadler bill.

First, the bill should define midnight rules as "final rules adopted by an agency in the 90 days before inauguration day, i.e. October 22 through January 20 in presidential election years."  "Adopted" should be defined as something like "sent to the Federal Register for publication as required by law."  These changes would clarify the timing issues—that the bill should apply only to presidential transitions in election years and that a rule is "adopted" when it is put in the queue for publication in the Federal Register. 

Second, the bill should give the incoming administration the option to suspend midnight rules for a certain period after inauguration day, perhaps sixty days, rather than suspend all midnight rules automatically.  If the incoming administration takes no action within the sixty day period, the rules should go into effect as adopted.  This avoids the uncertainty of tying the time period to the appointment of a new agency head, which may not happen at all if the incoming administration is of the same party or if there are holdovers from the prior administration, or may not happen for a long time if there is a delay in naming or confirming an incoming agency head.  Making action optional with the new administration is preferable to a blanket automatic suspension because it allows the incoming administration to focus its attention on issues it finds important.  This could have the negative effect of forcing the incoming administration to examine all midnight rules quickly, but this seems to be a better alternative than a blanket suspension.  This preference is based partly on the perhaps naïve hope that future administrations will seek to avoid midnight rulemaking as consciousness of the problem increases.

Third, the incoming administration should have the option to alter as well as disapprove of midnight rules, assuming that the alteration would be supported by the original rulemaking record.  If the agency finds it advisable to make changes beyond those supportable under the original rulemaking record, the agency should have the power to suspend a midnight rule to allow the time to conduct a new rulemaking proceeding.  Under Representative Nadler's proposal, review is an all-or-nothing proposition.  This ignores the possibility that the incoming administration will find only some of the features of a midnight rule objectionable.  It is easy to imagine the incoming administration objecting only to a limited number of features of a large midnight rule, and there is no reason for not allowing amendments rather than wholesale rejection.  Because of the fear that midnight rules may be undesirable due to hasty drafting and excessive interest group involvement, the incoming administration should also have the time to conduct a new rulemaking proceeding, if it decides one is necessary, without the midnight rule going into effect.  Administrations have done this without specific authorization.  For example, the Clinton Administration suspended the abortion gag rule in its first week in office and did not promulgate a substitute until its last year in office.[43]  It would be better if the incoming administration acted with statutory authority rather than created a situation of doubt and potential costly litigation. 

Fourth, "agency" should be defined in the bill to include only Executive Branch agencies and not independent agencies that are less subject to presidential control.  The worst dangers of midnight regulation relate virtually exclusively to the politics of the presidential transition.  As explained above, because independent agencies are not subject to much if any presidential control, midnight rules issued by them are less likely to be designed to embarrass or hinder the incoming administration or to project an outgoing administration's policies into the future.  Further, "midnight" at an independent agency may come later, when expiring terms or impending resignations shift control of the agency to members of the new President's party.  It would be very difficult and probably unnecessary to write a statute that identified the proper time during which an independent agency's rulemaking activity should be curtailed due to the possibility of midnight regulation.

Fifth, rules not subject to notice and comment should not be included in any midnight rule reform, mainly because the incoming administration could very easily alter any such midnight rule under current law.[44]  If an outgoing agency issues a policy statement or interpretive rule in its waning moments, the incoming agency can simply issue revisions without delay.  While this would have some of the negative effects of midnight regulation discussed above, the lack of substantive effect and the ease of rescission or amendment make midnight interpretive rules and policy statements a much less serious problem than midnight legislative rules.

In addition to the above proposals, which are designed to cure the defects in Representative Nadler's approach, an alternative statutory possibility could consist of a more general reform dealing with the administrative law issues identified in Part II.A.  A provision could be crafted that would give all agencies the power to revise or rescind rules shortly after adoption, thereby providing for revisions based on midnight regulation problems, or simply to allow second thoughts after promulgation even in the absence of a transition in administrations.[45]  This provision could, for example, give all agencies the power to revise or rescind rules within ninety days of adoption without a new rulemaking proceeding if the new rule (or no rule at all) would have been supported adequately by the original rulemaking record.  This provision has two virtues—it is simpler than one focused on the midnight rulemaking problem, and it increases regulatory flexibility in all periods, transition or not.

Finally, a far simpler proposal would be to ban outright the adoption of final rules by executive agencies for a specified period of time prior to inauguration day, for example 120 days, with exceptions for rules required by emergency, statutory deadline, or court order.  This possibility has several virtues.  For one, it would reduce the load that incoming administrations currently bear—there would be few or no midnight regulations to sift through since all regulations would have been finalized longer before election day than is currently the case.  While this reform might be viewed as an empty gesture since it would merely move up the deadline, in effect, changing the clock and making midnight come a few months early in election years has the additional virtue of placing all late-term rules in the sunshine of the public record long enough before election day for the public to be made aware of them.  This should discourage many midnight rules that appear to be farewell gifts to interest groups.  It would, however, not allow cooperation between incoming and outgoing administrations in my "waiting category," within which a cooperative outgoing administration clears away difficult issues late in the term as an aid to the incoming administration.  Further, the ban would apply even when it would not be necessary, for example, when an incumbent is reelected or when the new President is of the same party as the prior President.  It also assumes the worst—that all outgoing administrations are behaving badly rather than carrying on the business of government to the end, perhaps even in consultation with the incoming administration.  In such cases, the work of career agency employees could needlessly be slowed to a crawl every four years around election time.


Recent research confirms that midnight regulation has long been a feature of the transition between administrations.[46]  The general view is that midnight regulation is an illegitimate vehicle for projecting an outgoing administration's policy agenda beyond the end of its term.  For this and additional reasons disfavoring midnight regulation, Representative Jerrold Nadler's proposed Midnight Rule Act is a step in the right direction.  With the refinements discussed above, it might prevent some of the most egregious examples of midnight regulation.  Reforms to administrative law or simpler proposals that simply suspend rulemaking activity at the end of each presidential term might accomplish the same goal.  This Essay has outlined and analyzed these various possibilities with the hope of enhancing the understanding of ways of combating midnight regulation.


*.  Professor of Law and Harry Elwood Warren Scholar, Boston University School of Law.  © 2009, Jack M. Beermann, all rights reserved except that a license is hereby granted for non-profit reproduction and distribution for educational purposes.  Thanks to Michael Harper, Gary Lawson, William P. Marshall, David Mason, and Nina Mendelson for helpful comments on an earlier draft of this Essay.

1.  "Midnight regulation" is loosely defined as late-term action by an outgoing administration.  There are many types of midnight regulation and many reasons why the volume of administrative action tends to increase near the end of an administration.  For a detailed look at the phenomenon, see Jack M. Beermann, Presidential Power in Transitions, 83 B.U. L. Rev. 947 (2003).

2.  See, e.g., Jay Cochran III, The Cinderella Constraint: Why Regulations Increase Significantly during Post-Election Quarters (Mar. 8, 2001) (unpublished manuscript, on file with the Northwestern University Law Review), available at (link).

3.  See Beermann, supra note 1, at 949 n.6 (discussing Memorandum for the Heads and Acting Heads of Executive Departments and Agencies, 66 Fed Reg. 7702 (Jan. 20, 2001), issued by Andrew Card, White House Chief of Staff for the first five years of the administration of President George W. Bush, which suspended the effectiveness of some of the midnight rules of the Clinton administration).

4.  U.S. Const. art. II, § 1.

5.  See id. art. II, § 3; Jack M. Beermann & William P. Marshall, The Constitutional Law of Presidential Transitions, 84 N.C. L. Rev. 1253, 1273–76 (2006).

6.  H.R. 34, 111th Cong. (2009), available at (link).

7.  Id. § 2(a).

8.  Id. § 2(c)(2).

9.  Id. § 2(b)(2).  This provision is copied verbatim from the Congressional Review Act, 5 U.S.C. §§ 801–808 (2006), except for the addition of the requirement that the President act by Executive Order.  The bill also provides that the outgoing President's determination that a midnight rule should go into effect does not deprive Congress of its authority to reject a rule under the Congressional Review Act.

10.  The new administration could avoid the provisions of the bill by re-promulgating midnight rules immediately upon taking office, but that would take some work to make sure the process of re-promulgation was done properly.  Nor, in most cases, would it save much time because of the minimum thirty or sixty days required for rules to go into effect, which would presumably start over again on re-promulgation.

11.  See H.R. 34 111th Cong. § 2(b) (2009).

12.  See Executive Office of the President, Memorandum for the Heads and Acting Heads of Executive Departments and Agencies, 66 Fed. Reg. 7702 (Jan. 20, 2001) (from Andrew Card, then White House Chief of Staff) (link); Executive Office of the President, Memorandum for the Heads of Executive Departments and Agencies, 74 Fed. Reg. 4435 (Jan. 26, 2009) [hereinafter Emanuel Memorandum] (from Rahm Emanuel, current White House Chief of Staff) (link).

13.  The procedure for disapproving a rule contains a quirk that should be abandoned. It provides that the agency head disapproving a rule does so by "publishing a statement of disapproval in the Federal Register and sending a notice of disapproval to the congressional committees of jurisdiction."  The main problem is the requirement that notice go to the "congressional committees of jurisdiction."  The agency head should be required to send notice to Congress and allow Congress itself to decide which committees should be informed.  Committee jurisdiction is a matter of legislative rule and practice, which agency heads cannot interpret authoritatively.  The agency head could, as a courtesy, send notice to any committee that is known to engage in oversight of the particular administrative function, but the effectiveness of the disapproval should not depend on the agency making an accurate determination of which committees have jurisdiction over the matter in the disapproved rule.

14.  See 5 U.S.C. § 551(1) (2006) (defining "agency" to include "each authority of the Government of the United States") (link).  It has never been suggested that this definition does not include the independent agencies.

15.  It should also be noted that incoming administrations have not included independent agencies in their actions aimed that the midnight regulatory activity of their predecessors.  Both the Card and Emanuel memoranda were directed to the heads of "Executive Departments and Agencies."  This is probably the result of the lack of midnight activity from independent agencies and doubts about the President's authority to manage their activities.

16.  See H.R. 34 § 2(a).

17.  5 U.S.C. § 801 (2006) (link).

18.  For an extended discussion of the tools incoming Presidents have to combat midnight regulation, see Beermann, supra note 1, at 982–94.

19.  Emanuel Memorandum, supra note 12.

20.  Natural Res. Def. Council v. Abraham, 355 F.3d 179 (2d Cir. 2004).  This case involved a rule regarding the energy efficiency of central air conditioning and heat pumps.  The statutory structure underlying this rule caused it to be a particularly tricky midnight rule to deal with because under the statute, in what the court called an "anti-backsliding" provision, the agency (Department of Energy) was prohibited from ever lowering efficiency standards—new rules could only increase efficiency.  Thus, once the court held that the rule was final and effective when published in the Federal Register, the agency was stuck with it.  See id. at 197.  Even if the agency had conducted a new notice and comment period, as required for published rules by President Obama's directive, the anti-backsliding provision may still have prevented the incoming administration from changing this midnight rule.

21.  Memorandum from Josh E. Bolten to Heads of Executive Departments and Agencies (May 9, 2008) (on file with the Northwestern University Law Review) available at (link).

22.  463 U.S. 29 (1983) (link).

23.  See id. at 44–46.

24.  See Modified Standard 208, 42 Fed. Reg. 34289 (July 5, 1977).  There is nothing necessarily nefarious about time lines like this.  It makes perfect sense in many situations for new rules to be phased, strengthened, or even weakened over a long period of time, and in this case, the automobile manufacturers needed a substantial amount of time to prepare for compliance with the rule.  Long lag times can also facilitate the exploration of other options.  For example, one feature of the passive restraint rule would have rescinded the requirement entirely had two-thirds of states adopted laws requiring auto occupants to wear seatbelts within a certain time.  Ultimately, Congress acted to require airbags in all cars, which is more satisfactory than agency action from a democratic standpoint.

25.  See Motor Vehicle Mfrs. Ass'n, 463 U.S. at 40–44.

26.  See also Natural Res. Def. Council v. Abraham, 355 F.3d 179, 197 (2d Cir. 2004) (holding that once energy efficiency rules were published in the Federal Register, they could not be altered in way that violated the statute's anti-backsliding provision).

27.  Beermann, supra note 1, at 1009–15.

28.  See id. at 1014.  This would be procedurally similar to what often occurs when a court overturns an agency rule on judicial review.  When the problem with the rule is lack of a reasoned explanation, the court can remand the rule to the agency for a better explanation without requiring the agency to engage in a new round of notice and comment.  See, e.g., Chamber of Commerce of U.S. v. S.E.C., 443 F.3d 890, 900 (D.C. Cir. 2006) ("Where the court does not require additional fact gathering on remand . . . the agency is typically authorized to determine, in its discretion, whether such fact gathering is needed. . . .   If the agency determines that additional fact gathering is necessary, then notice and comment are typically required."); Shays v. Federal Election Comm'n, 414 F.3d 76, 112 (D.C. Cir. 2005) ("Nothing in our holding necessarily precludes the FEC from remedying deficiencies in its explanation and repromulgating this rule on remand.").

29.  Chevron U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837 (1984) (link).

30.  Nat'l Cable & Telecomms. Ass'n v. Brand X Internet Servs., 545 U.S. 967, 981–82 (2005) (link).

31.  Chevron, 467 U.S. at 865–66 (citations omitted).

32.  Brand X, 545 U.S. at 980.

33.  5 U.S.C. § 706(2)(A) (2006) (link).

34.  See Chevron, 467 U.S. at 843.

35.  5 U.S.C. § 706(2)(A).

36.  See Chevron, 467 U.S. at 843.

37.  549 U.S. 497, 527 (2007) (link).

38.  Id. (quoting Nat'l Customs Brokers & Forwarders Ass'n of Am., Inc. v. United States, 883 F.2d 93, 96 (D.C. Cir. 1989)).

39.  In Dunlop v. Bachowski, 421 U.S. 560, 573 (1975) (link), the Court stated that review of a refusal to initiate enforcement action should be very narrow and should not ordinarily look at the record beyond the bare reasons offered by the agency for not enforcing.  "The necessity that the reviewing court refrain from substitution of its judgment for that of the Secretary thus helps define the permissible scope of review.  Except in what must be the rare case, the court's review should be confined to examination of the 'reasons' statement, and the determination whether the statement, without more, evinces that the Secretary's decision is so irrational as to constitute the decision arbitrary and capricious. Thus, review may not extend to cognizance or trial of a complaining member's challenges to the factual bases for the Secretary's conclusion . . . ."  Id.

40.  See Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402 (1971) (link); Motor Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29 (1983) (link).

41.  See Vt. Yankee Nuclear Power Corp. v. Natural Res. Def. Council, Inc., 435 U.S. 519, 547 (1978) (link).

42.  See id.; see generally Jack Beermann & Gary Lawson, Reprocessing Vermont Yankee, 75 Gw. L. Rev. 856 (2007) (link).

43.  See Beermann, supra note 1, at 963–64.

44.  There are two sets of exceptions to the APA's notice and comment requirements.  First, the APA rulemaking provision does not apply to rules involving military and foreign affairs functions and rules relating to "agency management or personnel or to public property, loans, grants, benefits, or contracts."  5 U.S.C. § 553(a) (2006) (link).  The notice and comment requirements also do not apply to "interpretative rules, general statements of policy, or rules of agency organization, procedure, or practice."  5 U.S.C. § 553(b).  Publication requirements may still apply.  See 5 U.S.C. § 552(a)(1)(D) (2006) (link).

45.  A statute to this effect might have led the Second Circuit to allow President George W. Bush's Department of Energy to act against the last-minute rule issued by the Clinton Administration concerning the energy efficiency of central air condition and heat pumps that the Second Circuit held the Bush Administration could not revise in Natural Resources Defense Council v. Abraham, 355 F.3d 179 (2d Cir. 2004).  It is unclear whether this statutory reform would override the anti-backsliding provision of the statute at issue in Abraham, but in most situations it would allow agencies to adjust their rules for a brief period after issuance.

46.  See Antony Davies & Veronique de Rugy, Midnight Regulations: An Update (George Mason Univ. Mercatus Center Working Paper No. 08-06, 2008) (copy on file with the Nortwestern University Law Review), available at
WP0806_RSP_Midnight%20Regulations.pdf (link).


Copyright 2009 Northwestern University      

Cite as: 103 Nw. U. L. Rev. Colloquy 352 (2009),

Persistent URL:


(back to top)