Notice & Comment

Quantifying the Effects of Humane Society v. Department of Agriculture, by Mark Febrizio

Withdrawn Public Inspection Documents Include Significant Rules

This article is a cross-post from the GW Regulatory Studies Center. The original post can be found here.

Download this Insight (PDF)

Access the code and data (GitHub repository)


The majority decision in Humane Society of the United States v. Department of Agriculture (USDA) has made it more difficult for newly inaugurated presidents to undo the regulatory agenda of the previous administration. In July, the DC Circuit Court found that federal agencies unlawfully withdrew a final rule that appeared for public inspection in the Federal Register because the agency did not use the notice-and-comment process. In a prior article, I summarized the court case that prompted this change and discussed its implications, such as extending the midnight period at the end of a president’s tenure. But assessing the scope of the change, in addition to its nature, is relevant to understanding the policy impacts.

The Office of the Federal Register publishes government documents for public inspection each business day. Public inspection offers a preview of documents—including rules, proposed rules, agency notices, and presidential memoranda—that will be officially published in a subsequent issue of the Federal Register.

I used data from the past two presidential transitions to quantify the number and types of public inspection documents that newly inaugurated presidents withdrew. This analysis produced several findings: 1) a relatively small number of documents are affected; 2) a few important documents were withdrawn before publication during each transition; 3) President Trump was more active than President Biden in withdrawing posted but unpublished rules. Even if small in number, the withdrawals have had a meaningful effect by preventing significant regulations of the outgoing administration from being published.

Data Sources

Using the Federal Register’s Application Programming Interface (API), I tracked the number and type of documents that were filed for public inspection during a presidential transition and withdrawn before being published. Public inspection documents that are withdrawn have an editorial note accompanying their listing, helping to identify which agency actions were reversed before publication.

Because the earliest public inspection data I could find were from October 2011, this analysis focuses on the past two presidential transitions—Obama to Trump in 2017 and Trump to Biden in 2021. In addition, I searched and, which contain complementary information on regulations, to gather more context about final rules that were withdrawn.


I retrieved data on all documents appearing for public inspection from January 15 to January 31 in 2017 and 2021 (totaling 1,219 and 1,026 actions, respectively). In total, 60 documents were withdrawn from public inspection during the 2017 and 2021 presidential transitions. In January 2017, the incoming Trump administration withdrew 42 public inspection documents, while the Biden administration withdrew 18 documents in 2021 that were posted for inspection but not published. During both transitions, the majority of these documents were notices, some were final rules, and relatively few were proposed rules. Neither administration withdrew any presidential documents posted for inspection, which makes sense because the president has full discretion to repeal executive orders from prior administrations. Figure 1 shows the number of documents withdrawn by type for each transition year.

The data indicate that withdrawals are concentrated among a handful of agencies, with most experiencing fewer than three. Examining withdrawals by agency is also useful because it may show relative differences in the policy priorities both within and between administrations. For instance, President Trump prevented 16 documents from being published by the Department of Transportation (DOT), while President Biden only withdrew one. This difference could reflect several possibilities—such as a major push on transportation issues at the end of President Obama’s term, the Trump administration gearing up for its own transportation policy agenda, or both. Other agencies—including the Department of Health & Human Services (HHS), the Department of Agriculture (USDA), and the Department of the Interior (DOI)—experienced a similar number of withdrawals in 2017 and 2021 (between 4 and 7 each). Figure 2 illustrates the number of documents withdrawn by agency for each transition year.

To an extent, the number of withdrawals made by a president could be influenced by the amount of activity conducted by their predecessor. One way to account for this effect is to look at withdrawals as a proportion of all public inspection documents (i.e., the withdrawal rate).

In a transition year, attributing documents to the issuing president may be difficult because a document submitted by the outgoing administration might not publicly appear until after inauguration day. This reality influenced the choice to search public inspection documents appearing from January 15–31 to capture all withdrawals. Table 1 calculates the proportion of withdrawals to public inspection documents appearing in that timeframe.

Table 1: Withdrawal Rate of Public Inspection Documents, Jan. 15–31[1]

Transition YearDocument TypeWithdrawn DocumentsAll DocumentsWithdrawal Rate
2017Presidential Document0380
2017Proposed Rule3750.04
2021Presidential Document0560
2021Proposed Rule3490.061

Regardless of president, Table 1 suggests that incoming presidents withdraw a very small proportion of the documents their predecessors post for public inspection. Notably, the withdrawal rates for all documents and final rules were higher in 2017 than 2021. In other words, both the quantity and the rate of withdrawals were higher during the Obama-Trump transition relative to the Trump-Biden transition. This finding is consistent with prior research on regulatory suspensions—another common feature of the midnight period—that indicated Republican presidents delay final rules of the previous administration more frequently than Democratic presidents do. Differences in the withdrawal rate at the agency-year level are also evident. Although I do not dive into those results here for the sake of brevity, the data are available in a public GitHub repository.

The prior discussion treats each document of the same type as equivalent, but in reality, rules may have starkly different impacts and importance. A common way to illustrate the relative importance of rules is to look at their priority status in the Unified Agenda, including whether they are “significant” under Executive Order 12866.[2] Unfortunately, since the withdrawn documents were removed from the Federal Register, this check cannot be done by looking at the original public inspection documents.

Based on a manual search of the Federal Register,, and for the proposed rules predating the rules appearing for public inspection, I conclude that seven documents were significant, four were not significant, and two were exempt from review under Executive Order 12866. I was unable to determine the priority status for four rules.[3] For this analysis, I only looked at withdrawals of final rules, not notices or proposals. Table 2 shows the priority status for each withdrawn rule, based on its associated notice of proposed rulemaking (NPRM).

Table 2: Rules Withdrawn from Public Inspection during Two Presidential Transitions[4]

DateDocument NumberAgencyNPRMRINPriority
1/18/20172016-31493DOT2016-205802137-AF18Substantive, Nonsignificant
1/19/20172017-00854USDA2016-176480579-AE19Other Significant
1/19/20172017-00949EPA2014-243472040-AF26Other Significant
1/23/20172017-00522DOC2016-200180648-BF95Routine and Frequent
1/23/20172017-00242DOE2014-241511904-AB96Economically Significant
1/23/20172017-00803OPM2016-014763206-AM76Other Significant
1/23/20172017-01336DOTN/A2120-AA64Routine and Frequent
1/23/20172017-00686TREASN/AN/AExempt (foreign affairs)
1/24/2017C1-2016-29882EPA2016-29882N/AExempt (pesticide tolerances)
1/24/20172017-01554DOT2016-263342120-AA64Routine and Frequent
1/19/20212021-01212HHS2019-203150910-AH44Other Significant
1/19/20212021-01214HHS2019-057870910-AH89Other Significant
1/19/20212021-00789DOI2020-156501024-AE57Other Significant

In general, the Trump administration withdrew rules more indiscriminately than the Biden administration. The 2017 withdrawals included a mixture of significant rules, routine actions, and those exempt from centralized regulatory review (e.g., rules related to foreign affairs). Biden’s withdrawals were relatively targeted, with all three rules withdrawn in 2021 being significant.

Other factors that likely affect the scope and rate of withdrawals are the short window of time a new administration has to withdraw documents from public inspection and the associated logistical challenges of this task.[5] As a result, the scope of withdrawals might be influenced by how well-equipped the transition team is to navigate these logistics and assess which public inspection documents align with the incoming administration’s priorities.

Policy Implications

Previously, I discussed how the court’s ruling in Humane Society v. USDA effectively extends the midnight period when outgoing presidents rush to finalize their regulatory priorities. Such an extension would, among other consequences, invite more poorly justified rules because midnight rules are often supported by lower-quality regulatory analysis. But the quantity and significance of the specific actions affected matters too.

These data indicate that a relatively small number of public inspection documents are affected by the court case—and even fewer rules. By comparison, hundreds of rules and dozens of significant rules are published each midnight period (i.e., the period between Election Day and inauguration day for an outgoing administration). Specifically, President Obama issued 864 midnight rules (40 economically significant) and President Trump finalized 739 midnight rules (48 economically significant) at the end of their presidencies.[6]

Although the results do not reveal a very large number of reversals, they may still have a meaningful impact in terms of regulatory effects. Several consequential rules were withdrawn before publication during the last two presidential transitions. Trump withdrew four significant rules in 2017 (including one economically significant rule), and Biden withdrew three significant rules in 2021.

What happens to these withdrawn public inspection rules, especially those that were significant? If the court’s decision applies retroactively, then some could “rise from the regulatory graveyard” as Circuit Judge Neomi Rao cautioned in her dissent. In light of that possibility, agencies may want to pay attention to any consequential actions withdrawn while on public inspection that might attract litigation from interested parties.

Mark Febrizio is a senior policy analyst at the George Washington University Regulatory Studies Center.

This insight reflects the views of the author and does not represent an official position of the GW Regulatory Studies Center or the George Washington University. The Center’s policy on research integrity is available at

[1]    Sources for Table 1: Federal Register and author’s calculations. The data are available in the GitHub repository.

[2]    A regulation’s significance is generally indicated in the text of its Federal Register publication and/or in the “priority” field of the Unified Agenda.

[3]    I could not locate a notice of proposed rulemaking for some rules. One potential explanation could be that some of the rules were not preceded by a proposal (e.g., an interim final rule). These values are marked as N/A in the table.

[4]    Sources for Table 2: Federal Register for public inspection editions and NPRMs; Python Memento Client for archive links; for RINs and Unified Agenda entries; for confirming information. Additional context for identifying NPRMs can be found in the files “rules_withdrawn_priority.csv” and “nprm_data.csv” in the GitHub repository.

[5]    For instance, the agency official tasked with submitting a withdrawal letter to the Office of the Federal Register may only have several hours to do so, in addition to many other inauguration day duties.

[6]    These data were retrieved from the Federal Register API are contained in the file “midnight_rules_by_president_year_month.csv” in the GitHub repository. I excluded corrections to final rules from these counts. The data on economically significant rules include the last 3 months of data from the “Cumulative Economically Significant Final Rules by Administration” chart on Reg Stats (accessed October 4, 2022):

Revised 10/10/2022: A prior version incorrectly implied that proposed rules were withdrawn at a higher rate during the 2017 transition, rather than just final rules and notices.

Print Friendly, PDF & Email