Federal rulemaking pace varies enormously across agencies (Gini 0.80) and has declined modestly but unevenly nationwide since 2010

Summary

Federal rulemaking pace varies enormously between agencies and has drifted down modestly and unevenly nationwide since 2010. The Department of Transportation alone issued a 2010-2025 average of 749 final rules a year — nearly seven times the Department of Commerce's 337 and the Interior Department's 90 — and this gap has persisted at roughly the same order of magnitude for 16 straight years. Concentration across the ~64 agencies that issue rules is extreme (Gini coefficient 0.80 for 2025), meaning a handful of agencies (DOT, DHS, EPA, Commerce) account for the great majority of all final rules published. Nationally, final-rule output fell from 3,502 in 2010 to 2,442 in 2025 (-30%, linear trend of about -61 rules/year, r=-0.80), but the year-to-year path is not a smooth decline — it dips sharply after each presidential transition (2017, 2021, 2025) and partially recovers in between. Agencies also differ sharply in how volatile their own year-to-year pace is: Defense's rule count swings the most relative to its average (coefficient of variation 0.37) while Commerce is the steadiest (CV 0.10).

How the pace varies across agencies

Ranking all agencies (a joint rule naming two agencies counts toward both) by total final rules issued 2010-2025 in fedregister.fr_documents:

The spread is about 8-to-1 between the busiest agency (DOT) and the tenth-busiest shown here, and roughly 25-to-1 against the smallest rule-issuing agencies once the full ~64-agency roster is included — reflected in a Gini coefficient of 0.80 for 2025 final-rule volume, where 0 would mean every agency issues an equal share and 1 would mean one agency issues everything. About 90% of agencies by count account for well under a third of all rules (see Lorenz curve computed via the gini_coefficient tool). This concentration is structural, not a one-year artifact: DOT, DHS, EPA and Commerce have occupied the top four spots throughout 2010-2025.

Volatility (coefficient of variation of each agency's own annual count) also differs meaningfully: Defense (0.37) and Agriculture (0.29) swing much more from year to year, relative to their own average pace, than Commerce (0.10) or Transportation (0.11) — the biggest issuers are also, on this measure, the steadiest.

How the pace has changed over time

Nationally, total final rules fell from 3,502 (2010) to a peak of 3,855 (2016, the final full year of the Obama administration) then declined through most of the following decade to 2,442 in 2025 — the lowest full-year total in this 16-year window. A simple linear trend across all complete years (2010-2025) gives a slope of about -61 final rules per year (r = -0.80), though the actual path is stepwise rather than smooth: each change of administration (2017, 2021, 2025) is followed by a one-to-two-year trough as the incoming administration pauses the prior pipeline and launches its own priorities, a pattern independently documented by the Competitive Enterprise Institute's long-running Federal Register tracking ("a drop in page counts between administrations is typical, as incoming presidents freeze the pipeline temporarily"). The 2024-to-2025 drop is the sharpest single-year move in the window: final rules fell about 25% (3,248 -> 2,442), which matches CEI's independently reported ~26% decline in total final rules over the same period, and this warehouse's 'significant' (OIRA EO 12866-flagged) final rules fell about 45% (345 -> 189) in the same year — directionally consistent with, though not numerically identical to, CEI's narrower "economically significant" rule count (down 66%, 129 -> 44); the two figures use different significance criteria and are not the same measure.

The trend is not uniform across agencies. DHS (slope ~ -18/year) and EPA (slope ~ -16/year) show the steepest declines among the top agencies; Interior is essentially flat (slope ~ +0.15/year) over the full window, and Transportation and Commerce decline only mildly (~-5 and -4/year respectively) despite being by far the largest issuers.

Data, validation, and caveats

Source: fedregister.fr_documents (govinfo.gov bulk XML archive), doc_type='RULE' (final rules), 2010-2025 — data_coverage confirmed all 16 years fully loaded with no interior gaps (2026 is partial, through September, and excluded from full-year comparisons). Agency attribution uses the denormalized agency_names field, split on commas; a document naming multiple agencies is counted once per named agency, so per-agency totals sum to more than the national total — this is expected and not a data error.

Independent validation: this warehouse's national final-rule counts were checked directly against the Competitive Enterprise Institute's own Federal Register tabulations (10,000 Commandments series) and match exactly for the three years CEI cites by name: 3,248 final rules in 2024 (CEI: 3,248), 3,018 in 2023 (CEI: 3,018), and 2,964 in 2019 (CEI: 2,964). The 2016 figure is very close (3,855 here vs. CEI's 3,853, immaterial). This cross-check gives high confidence in the base counts.

Caveat on the 'significant rules' figure: fedregister.fr_significance's `significant` flag is the Federal Register API's own EO 12866 significance field, but it is broader than the OIRA "economically significant" ($100M+ effect) category CEI reports separately — the two are directionally consistent (both show a sharp 2024-2025 drop) but not numerically comparable, and this is flagged rather than reconciled to a single number.

What this does not establish: the year-over-year and agency-over-agency comparisons here are descriptive, not causal — the post-transition troughs are a well-documented institutional pattern (regulatory freezes at the start of a new administration) rather than something tested here with a controlled design, and CEI's own analysis notes a possible confound (a shift toward fewer, larger/costlier rules and more sub-regulatory guidance not counted in a Federal Register rule tally) that this count-based measure cannot detect.

Every query behind this report

11 warehouse calls ran in this session, in order. Each is reproducible against the same snapshot.

WITH exploded AS (
  SELECT document_number, "year", doc_type, trim(CAST(a AS VARCHAR)) AS agency
  FROM fedregister.fr_documents, UNNEST(string_split(agency_names, ',')) AS t(a)
  WHERE doc_type = 'RULE' AND "year" BETWEEN '2010' AND '2025'
)
SELECT agency, count(*) AS total_rules
FROM exploded
GROUP BY agency
ORDER BY total_rules DESC
FETCH FIRST 20 ROWS ONLY

WITH exploded AS (
  SELECT document_number, "year", doc_type, trim(CAST(a AS VARCHAR)) AS agency
  FROM fedregister.fr_documents, UNNEST(string_split(agency_names, ',')) AS t(a)
  WHERE doc_type = 'RULE' AND "year" BETWEEN '2010' AND '2025'
)
SELECT agency, "year", count(*) AS n_rules
FROM exploded
WHERE agency IN ('Department Of Transportation','Department Of Homeland Security','Environmental Protection Agency','Department Of Commerce','Department Of Health And Human Services','Department Of Agriculture','Department Of The Treasury','Department Of Defense','Federal Communications Commission','Department Of The Interior')
GROUP BY agency, "year"
ORDER BY agency, "year"

SELECT "year", doc_type, count(*) AS n
FROM fedregister.fr_documents
WHERE "year" BETWEEN '2010' AND '2025' AND doc_type IN ('RULE','PRORULE')
GROUP BY "year", doc_type
ORDER BY "year", doc_type

SELECT "year", max("month") AS max_month, count(*) AS n
FROM fedregister.fr_documents
WHERE "year" IN ('2025','2026')
GROUP BY "year"

SELECT "month", count(*) AS n FROM fedregister.fr_documents WHERE "year"='2025' GROUP BY "month" ORDER BY CAST("month" AS INT)

WITH exploded AS (
  SELECT document_number, trim(CAST(a AS VARCHAR)) AS agency
  FROM fedregister.fr_documents, UNNEST(string_split(agency_names, ',')) AS t(a)
  WHERE doc_type = 'RULE' AND "year" = '2025'
)
SELECT agency, count(*) AS n_rules FROM exploded GROUP BY agency

SELECT regr_slope(CAST(n AS DOUBLE), CAST("year" AS DOUBLE)) AS slope_per_year,
       regr_intercept(CAST(n AS DOUBLE), CAST("year" AS DOUBLE)) AS intercept,
       corr(CAST(n AS DOUBLE), CAST("year" AS DOUBLE)) AS r
FROM (
  SELECT "year", count(*) AS n
  FROM fedregister.fr_documents
  WHERE doc_type='RULE' AND "year" BETWEEN '2010' AND '2025'
  GROUP BY "year"
) t

SELECT "year", count(*) FILTER (WHERE significant = true) AS econ_significant
FROM fedregister.fr_significance
WHERE "year" BETWEEN '2016' AND '2025'
GROUP BY "year"
ORDER BY "year"

SELECT s."year", count(*) FILTER (WHERE s.significant = true) AS n
FROM fedregister.fr_significance s
WHERE s.document_number IN (SELECT document_number FROM fedregister.fr_documents WHERE doc_type='RULE')
  AND s."year" BETWEEN '2016' AND '2025'
GROUP BY s."year"
ORDER BY s."year"

WITH exploded AS (
  SELECT trim(CAST(a AS VARCHAR)) AS agency, "year", count(*) AS n
  FROM fedregister.fr_documents, UNNEST(string_split(agency_names, ',')) AS t(a)
  WHERE doc_type = 'RULE' AND "year" BETWEEN '2010' AND '2025'
  GROUP BY trim(CAST(a AS VARCHAR)), "year"
)
SELECT agency,
       avg(n) AS avg_annual_rules,
       stddev_samp(n) AS sd_annual_rules,
       stddev_samp(n)/avg(n) AS cv,
       regr_slope(CAST(n AS DOUBLE), CAST("year" AS DOUBLE)) AS slope_per_year,
       min(n) AS min_n, max(n) AS max_n, count(*) AS n_years
FROM exploded
WHERE agency IN ('Department Of Transportation','Department Of Homeland Security','Environmental Protection Agency','Department Of Commerce','Department Of Health And Human Services','Department Of Agriculture','Department Of The Treasury','Department Of Defense','Federal Communications Commission','Department Of The Interior')
GROUP BY agency
ORDER BY avg_annual_rules DESC

WITH exploded AS (
  SELECT trim(CAST(a AS VARCHAR)) AS agency, "year", count(*) AS n
  FROM fedregister.fr_documents, UNNEST(string_split(agency_names, ',')) AS t(a)
  WHERE doc_type = 'RULE' AND "year" BETWEEN '2010' AND '2025'
  GROUP BY trim(CAST(a AS VARCHAR)), "year"
)
SELECT agency,
       avg(CAST(n AS DOUBLE)) AS avg_annual_rules,
       stddev_samp(CAST(n AS DOUBLE)) AS sd_annual_rules,
       CAST(stddev_samp(CAST(n AS DOUBLE)) AS DOUBLE)/CAST(avg(CAST(n AS DOUBLE)) AS DOUBLE) AS cv
FROM exploded
WHERE agency IN ('Department Of Transportation','Department Of Homeland Security','Environmental Protection Agency','Department Of Commerce','Department Of Health And Human Services','Department Of Agriculture','Department Of The Treasury','Department Of Defense','Federal Communications Commission','Department Of The Interior')
GROUP BY agency
ORDER BY cv DESC