← Home

Overrides Over Time in Massachusetts

Every Prop 2½ operating override ballot question across all 305 municipalities, FY1990 to FY2027. Source: MA DOR DLS Municipal Databank.

• Operating overrides only. Debt exclusions (capital projects, school buildings) and underrides are not included. These are the permanent levy increases that fund day-to-day town and school budgets
• 4,640 override ballot questions across 305 towns since 1990; 2,268 passed (48.9%)
• The early '90s were a flood: 400-600 questions per year, pass rates as low as 25%
• Pass rates climbed through the 2000s-2020s, peaking above 80% in FY2019 and FY2022-23
• FY2025-26 marks a new surge in override volume as cost pressures outpace the 2.5% levy cap statewide
• Median override amount per election: $443K in FY1990, $1.7M in FY2026. Many early overrides were split into 10-30 ballot questions; the per-question and per-election lines converge as towns shifted to bundled votes

Override ballot questions per year

Each bar is one fiscal year. Dark = passed, light = did not pass.

---

Statewide pass rate

Percentage of override ballot questions that passed in each fiscal year.

---

Median override amount per year

Many towns split one override into multiple ballot questions (Chatham FY1990 had 31). The per-election line combines all questions from one town in one year; the per-question line shows each ballot item individually. The two converge in recent years as towns shift toward bundled questions.

Notes and sources

Override vote data from the Massachusetts Department of Revenue, Division of Local Services Municipal Databank, Proposition 2½ Override/Underride Database. Dataset includes 4,640 operating override ballot questions (underrides excluded) across 305 municipalities from FY1990 to FY2027.

Each ballot question counts as one vote, regardless of dollar size. A single town election can have multiple ballot questions (e.g. Needham FY1997 had ten separate override questions).

FY2027 data is partial as of April 2026. One Marblehead row (FY2025, "Fix Clerk Error", $0) is excluded as a clerical correction.

"Pass rate" is the number of questions that passed divided by total questions in that fiscal year. It does not weight by dollar amount or population.

"Per ballot question" median is the median of all individual ballot question amounts in a given fiscal year. "Per election" median aggregates all questions from the same town in the same fiscal year into a single combined amount before computing the median. The gap between the two lines reflects how many towns split overrides into multiple line-item questions rather than one bundled vote. In nominal (not inflation-adjusted) dollars.