Fiscal evidence brief · U.S. 1920–2026
Temporary, emergency, and one-time U.S. taxes rarely revert to pre-crisis baselines. In the documented sample, most are extended, made permanent, or replaced.
Current documented sample
Provisional. Drawn from n ≈ 6–8 well-documented federal and prominent state/local cases. Not a census, not a confidence-interval estimate, and sensitive to case selection and definitions.
Case ledger
Each levy was labeled temporary or emergency and carried a stated horizon. Align the promised path with what actually happened after the sunset day.
Federal Telephone Excise Tax
Trigger
War financing: Spanish-American War, WWI, Depression, WWII, Vietnam; later deficit reduction
Stated sunset
Multiple explicit expiration dates tied to each conflict’s end
Actual outcome
Series of repeal-and-reenactment episodes; not one continuous tax. Final repeal in 2006 driven mainly by court rulings and administrative decisions, not strict sunset enforcement
Evidence
CRS and secondary histories
Federal Gasoline Tax
Trigger
Great Depression emergency measure — temporary 1-cent-per-gallon levy
Stated sunset
Explicitly intended to last one year; stated sunset after one year
Actual outcome
Repeatedly extended by Congress; converted into permanent financing for the Federal Highway Trust Fund, where it remains in effect
Evidence
Federal tax histories
Pennsylvania Johnstown Flood Tax
Trigger
Devastating flooding recovery — temporary 10% tax on liquor sold in state stores
Stated sunset
Explicitly tied to the disaster-recovery trigger; implied post-recovery sunset
Actual outcome
Never repealed once recovery ended; revenue redirected to the General Fund; rate later increased to 18%; persists as of 2026
Evidence
State fiscal histories
Somerset County Disaster Recovery Fund
Trigger
Severe flooding; framed as a targeted response to that specific event
Stated purpose
One-time or event-driven recovery funding
Actual outcome
In early 2026, county commissioners voted to make the fund permanent to support ongoing emergency readiness
Evidence
Local reporting
Florida Disaster-Preparedness Sales-Tax Holiday
Trigger
Disaster was preparedness — temporary two-week sales-tax holiday on emergency supplies, renewed annually
Stated form
Explicit short-term windows, renewable
Actual outcome
Per 2025–2026 state budget reports, the temporary holiday was replaced by a permanent sales-tax exemption for the same supplies — embedding the change in baseline law. Produces a permanent tax reduction rather than an increase, but the same pattern of temporary emergency provisions becoming enduring fiscal policy
Evidence
State budget coverage
Enacted in 1932 as a temporary one-year emergency levy. Drag the path beyond the marked sunset — it continues into permanent Highway Trust Fund financing.
At handle
1932 — temporary 1¢/gal, one-year emergency measure
Relationship to sunset
Still inside the promised one-year window
Quantified overview
Genuine expiration exactly as promised is rare across the documented instances since 1920. Hover a bar for the category definition.
Sample. Includes the telephone-tax extension cycle, the gasoline tax, the Johnstown Flood Tax, and the Somerset fund among others. Directional signal only.
n ≈ 6–8 well-documented federal and prominent state/local cases. No confidence intervals.
3 cases with usable detail: Johnstown, Somerset, Florida.
2–3 cases with explicit rider/reconciliation references; primary records are sparse.
Mechanism
The ratchet effect does not require the tax text itself to survive. Two permanence channels operate — often together.
The tax never sunsets, or sunset clauses are repealed or extended through subsequent legislation.
Even a technical repeal can leave the spending, agencies, or infrastructure in place.
These mechanisms compound: baseline budgeting raises the repeal threshold, reconciliation and riders provide low-friction extension pathways, and habituation plus bureaucratic incentives make reversion politically costly. The ratchet operates as a flywheel rather than a one-way valve.
Bounds on the signal
Historians and fiscal analysts broadly concur on the directional ratchet but differ on precise quantification. The ~20% true-sunset figure is a provisional best estimate, not a settled statistic.
n ≈ 6–8 well-documented cases. Thousands of narrower local bond fees and disaster surtaxes are poorly tracked and may alter the ratios if fully catalogued.
“One-time” or “emergency” is self-reported at enactment. No standardized coding across jurisdictions.
Coverage favors salient federal excise taxes. Confidence is higher for CRS-documented federal episodes and lower for local variation.
The gasoline tax dwarfs many local levies; the unweighted case count does not reflect fiscal mass. No confidence intervals can be calculated from the available sample.
Provenance
This brief was assembled by becoming a fiscal-history analyst focused on sunset clauses and the ratchet literature — the specialist best matched to documenting temporary-tax outcomes with explicit hedges on sample size. Other framings were weighed and set aside.
Method: chain-of-verification on discrete claims (trigger, stated sunset, actual outcome) for each named levy; separate legislative from economic permanence; refuse point estimates without the hedges the evidence supports. Self-reviewed across multiple passes. Four sources cited; zero additional-only consults beyond the cited set.
The “temporary” label functions, in this record, more as political cover at enactment than as a binding commitment to expire.
Return the sunset marker to what the evidence supports: a directional signal, not a binding commitment. In the documented sample since 1920, roughly four in five temporary taxes do not reset to the pre-crisis baseline.