Bibliographic record
Abstract
Federal regulatory activity surges during the final quarter of presidential election years that result in a change in administrations. Scholars such as Jay Cochran, Antony Davies and Veronique de Rugy, and Anne Joseph O’Connell have well-documented the phenomenon of “midnight regulations” promulgated between Election Day and Inauguration Day. One common criticism of midnight regulations is that the quality of analysis accompanying these regulations is likely to be lower than those accompanying earlier or later regulations, possibly because more regulatory activity tends to stretch the Office of Information and Regulatory Affairs’s limited resources for reviewing regulations — and perhaps also taxes agencies’ own resources. But is there statistical support for this criticism of midnight regulations? In a recent study, we tested this claim by using score data from the Mercatus Center’s Regulatory Report Card. The Report Card consists of expert assessments of the regulatory impact analyses (RIAs) that must accompany economically significant regulations. The experts rank RIAs on a Likert scale of zero to five using criteria derived from Executive Order 12866 and OMB Circular A-4. As we detailed in an earlier The Regulatory Review essay, the Report Card assesses the quality of RIAs and the extent to which issuing agencies claim to have used the analysis to make decisions. Because we started using the Report Card to evaluate the quality of RIAs for regulations proposed in 2008, the score data include the Bush administration’s midnight regulations. The accompanying chart illustrates graphically what we found econometrically. On average, midnight regulations have slightly lower Report Card scores than other regulations. However, the difference in mean scores between the two groups is not statistically significant. Instead, our research did reveal two related groups of regulations with statistically significant lower average scores than all rules overall: midnight regulations proposed after June 1, 2008 (“rushed midnight regulations”), as well as regulations proposed after June 1, 2008, but left for the Obama administration to finalize (“rushed leftovers”). The Bush administration tried to limit the presence of midnight regulations by finalizing regulations before Election Day. More specifically, the Bush administration instructed agencies that all regulations they planned to finish before the end of the Bush presidency should be proposed by June 1, 2008, and finalized by November 1, 2008. We found that these rushed midnight regulations indeed had lower-quality analysis, and the difference was statistically significant. Rushed midnight regulations also had significantly lower scores for use of analysis. We also found that the rushed leftovers had lower scores for use of analysis, and this difference was also statistically significant. Our findings lead us to conclude that for rushed leftovers decisionmakers were less likely to explain how the analysis affected their decisions or how they planned to evaluate the regulations’ performance in the future. We do not know if this occurred because these were supposed to be midnight regulations that did not quite beat the clock, or because outgoing officials knew that the final decisions would be made by the next administration so they felt less need to justify the regulations based on analysis. While it may not be possible yet to discern the precise motivations or reasons for our empirical results, the key finding is important. The rushed nature of regulations proposed after June 1, 2008, appears to have been responsible both for the lower quality and diminished use of regulatory impact analysis.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.008 | 0.025 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.004 | 0.002 |
| Scholarly communication | 0.006 | 0.003 |
| Open science | 0.001 | 0.003 |
| Research integrity | 0.002 | 0.004 |
| Insufficient payload (model declined to judge) | 0.057 | 0.014 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".