A quantitative study of how executive, legislative, and judicial power have expanded, contracted, and shifted across 235 years of American government — from Washington's first term through the 119th Congress and October Term 2025.
Executive Power Index (EPI) · Legislative Power Index (LPI) · Judicial Power Index (JPI) · Computed March 2026
The American constitutional experiment rests on a deceptively simple premise: separate the power to make law, enforce law, and interpret law — then force those branches into perpetual negotiation. Two and a half centuries later, the data tells a more complicated story. Power doesn't stay balanced; it migrates. And the tools each branch wields to assert dominance have multiplied dramatically since the founding.
This study constructs three parallel indices — the Executive Power Index (EPI), Legislative Power Index (LPI), and Judicial Power Index (JPI) — each built from six raw metrics, normalized and weighted to create a single composite score per term, Congress, or Court session. The goal is not to declare winners and losers but to map how institutional capacity evolves and how branches respond to one another.
All three branches operated with minimal institutional infrastructure. Presidents issued a handful of executive orders (Washington averaged 1 per year), Congress passed modest bodies of law with no committee oversight apparatus, and the Supreme Court — deciding under 50 cases annually — was still establishing its basic authority. The JPI for this era is driven almost entirely by caseload, with judicial review appearing only sporadically (Marbury v. Madison in 1803 marks the first law struck down). The EPI and LPI are both near historical lows, and the "balance" is less a tension between strong forces than a shared condition of institutional modesty.
Lincoln's presidency marks the first sustained spike in executive unilateralism — his EO rate was the highest of any president to that point, and the Civil War validated broad claims of emergency authority. Congress responded with aggressive Reconstruction legislation, veto overrides of Andrew Johnson (who issued 29 vetoes, most famously on civil rights), and the first serious use of impeachment. The judiciary scaled up rapidly: the Supreme Court decided over 240 cases in some terms, and began striking down laws at a pace it would not match again until the 20th century. This is the first era where all three indices rise simultaneously — a signal of institutional expansion across the board, driven by national crisis.
This period sees the sharpest divergence between branches. The JPI peaks in the late 19th century when the Court routinely decided 250–300+ cases per term and struck down laws at unprecedented rates (the Lochner era). The EPI spikes dramatically with the Progressive presidents — Theodore Roosevelt, Wilson, and above all FDR, who issued executive orders at a rate of 307 per year and used vetoes at an historic pace. Legislative output swelled too: Congresses in this period regularly enacted 500–700 laws per session, reaching a sustained plateau. The New Deal era represents the most significant power shift in the dataset: executive authority expanded into economic regulation, emergency powers, and agency creation, while the Court initially resisted (striking 8 laws in OT 1935 alone) before capitulating after the "switch in time" of 1937.
The Cold War locked in elevated executive power (EOs, emergencies, recess appointments) while Congress built out its modern institutional machinery. The LPI peaks in the mid-1970s: the 94th Congress (1975–77) scored the highest LPI of any Congress in history, driven by the post-Watergate expansion of oversight, the creation of the budget resolution process (1974 Budget Act), and a record 8 veto overrides of Gerald Ford. Meanwhile, the Warren and Burger Courts sustained historically high caseloads (200–370 cases per term) with aggressive constitutional activism, creating the highest sustained JPI readings of the modern era. This is the only period where all three branches simultaneously register high composite scores — genuine tripartite engagement.
The modern era is defined by divergence. Executive power has intensified through presidential memoranda, emergency declarations, and regulation — the EPI for the last four administrations (Obama, Trump I, Biden, Trump II) averages well above the historical mean. Legislative output has contracted: modern Congresses enact fewer than 400 laws per session (down from 800+ during the Cold War), though new tools like the Congressional Review Act have opened novel channels of influence — the 119th Congress has already enacted a record 22 CRA resolutions. Judicial throughput has collapsed: the Roberts Court decides 60–80 cases per term (vs. 250+ a century ago), but strikes laws and overrules precedent at higher per-case rates than any prior era, producing outsized constitutional impact from a shrinking docket.
Comparability. Each index uses the same fundamental architecture: six raw metrics → per-unit normalization (where applicable) → min-max scaling to [0, 1] → weighted sum. This makes the indices internally consistent but not directly comparable in absolute terms. An EPI of 0.400 does not mean "more powerful" than an LPI of 0.350 — the scales, metric counts, and historical baselines differ. The overlay chart above re-normalizes each index to its own [0, 1] range for visual comparison only.
Partial terms. The 119th Congress, Trump II's presidency, and OT 2025 are all partial. Their scores reflect activity through approximately March 2026 and will change as more data accrues. Per-year rates (used in the EPI) partially compensate for short durations but can overstate annualized activity during intensive early-term bursts.
Metric availability. Many metrics are era-gated: National Emergency Declarations begin with Carter (1978 NEA), CRA resolutions begin with the 105th Congress (1997), and detailed SCOTUS split/cert data begins with OT 1946. Pre-availability values are zeroed, meaning earlier terms structurally cannot score as high as later ones. This is an acknowledged limitation — the indices are most reliable for within-era comparisons.
Weighting. All three indices assign 0.35 weight to the primary volume metric (EOs, Laws Enacted, Cases Decided), 0.20 to the secondary check-on-power metric (Vetoes, Veto Overrides, Laws Struck), and distribute the remaining 0.45 across four additional metrics. The choice of weights reflects editorial judgment about which tools represent the most consequential exercise of branch power. Sensitivity analysis shows that the top-5 and bottom-5 rankings are robust across reasonable weight perturbations (±0.05 on any single metric).
All 47 Presidential Terms · Washington (#1) through Trump II (#47) · Computed March 2026
Executive Orders (published in the Federal Register), Vetoes (regular + pocket), Clemency Acts (pardons + commutations), National Emergency Declarations (new declarations under the 1976 NEA), Recess Appointments, and Economically Significant Rules (>$100M impact). Each metric is converted to a per-year rate (to normalize for varying term lengths), then min-max normalized across all 47 terms to [0, 1]. The EPI is then: 0.35 × Norm_EO + 0.20 × Norm_Veto + 0.15 × Norm_Clemency + 0.15 × Norm_Emergency + 0.10 × Norm_Recess + 0.05 × Norm_Reg. Metrics not tracked in earlier eras are set to 0 (Norm = 0).
| President | Years | EO / yr | Veto / yr | Clem / yr | Emerg / yr | Recess / yr | Reg / yr | EPI Score |
|---|
Executive Orders
American Presidency Project — Executive Orders TableVetoes
U.S. Senate — Summary of Bills Vetoed, 1789–PresentClemency (Pardons & Commutations)
DOJ Office of the Pardon Attorney — Clemency StatisticsNational Emergency Declarations
Brennan Center for Justice — Guide to Emergency PowersRecess Appointments
Congressional Research Service — Recess Appointments FAQ (RL33310)Economically Significant Regulatory Rules
GW Regulatory Studies Center — Reg Stats DatabaseAdditional References
American Presidency Project (UCSB) — Primary Data TablesAll 119 U.S. Congresses · 1st Congress (1789) through 119th Congress (2026 partial) · Computed March 2026
Laws Enacted (public laws signed or enacted via override), Veto Overrides (presidential vetoes overridden by both chambers), Budget Resolutions (concurrent budget resolutions adopted, post-1974), Oversight Hearings (committee/subcommittee meetings as proxy), Impeachment Proceedings (articles introduced in the House), and CRA Resolutions (Congressional Review Act disapproval resolutions enacted, post-1996). Each metric is min-max normalized across all 119 Congresses to [0, 1]. The LPI is then: 0.35 × Norm_Laws + 0.20 × Norm_Overrides + 0.15 × Norm_Budget + 0.10 × Norm_Oversight + 0.10 × Norm_Impeach + 0.10 × Norm_CRA. Metrics that did not exist in a given era are set to 0 (Norm = 0).
| Congress | Laws Enacted | Veto Overrides | Budget Res. | Oversight | Impeach. | CRA | LPI Score |
|---|
Public Laws Enacted
GovTrack.us: Historical Statistics about Legislation in the U.S. CongressVeto Overrides
U.S. Senate: Presidential Vetoes, 1789 to PresentBudget Resolutions
CRS Report RL30297: Congressional Budget Resolutions — Historical InformationOversight Hearings
Brookings Vital Statistics: Tables 6-1 & 6-2 (committee meetings, 80th–113th)Impeachment Proceedings
U.S. House: Complete List of Impeachment ActionsCongressional Review Act (CRA) Resolutions
CRS Report R43992: The CRA — Frequently Asked QuestionsAll Supreme Court October Terms · OT 1791–OT 2025 (partial) · Computed March 2026
Cases Decided (merits opinions), Laws Struck Down (federal & state statutes invalidated), Precedents Overruled (prior rulings expressly reversed), 5-4 Splits (ideological split decisions at minimum winning margin), Constitutional Cases (subset involving constitutional questions), and Cert Granted (certiorari petitions granted). Each metric is min-max normalized across all ~235 terms to [0, 1]. The JPI is then: 0.35 × Norm_Cases + 0.20 × Norm_LawsStruck + 0.15 × Norm_Precedent + 0.15 × Norm_Splits + 0.10 × Norm_ConstCases + 0.05 × Norm_Cert. Metrics not tracked in early eras are set to 0 (Norm = 0).
| Term | Cases Decided | Laws Struck | Precedents Overruled | 5-4 Splits | Const. Cases | Cert Granted | JPI Score |
|---|
Primary Data — Cases Decided
Wikipedia: Number of U.S. Supreme Court Cases Decided by YearLaws Struck Down
Library of Congress: Table of Laws Held UnconstitutionalPrecedents Overruled
Library of Congress: Table of Decisions OverruledSCOTUSblog Stat Packs
SCOTUSblog: Stat Pack Archive (OT 1995–OT 2024)Additional References
Justia: U.S. Supreme Court Opinions