F04Annual (end of fiscal year) · US$ billions / % of GF revenues
Rainy-day fund (Budget Stabilization Account): balance and share of GF revenues
Rainy-day balance and its share of General Fund revenue. A constitutional cap is context, not a finding that the fund is adequate.
Kind of fucked
Inside its 2017–19 range · getting worseJudged on “BSA as % of GF revenues”. A smaller rainy-day fund is bad. Scored reading: Rainy-day fund 7.89 % of GF revenue.
How verdicts work →BSA as % of GF revenues
6.65% · 2027 · enacted_estimateDashed / hollow = preliminary or estimate
Download CSVSource & method ↓
Where this sits in the reading
Position versus 2017–19 is descriptive. “Within range” doesn’t mean acceptable — 2017–19 may itself have been strained.
What this is
Past-year actual from each January Governor's Budget; 2025-26 revised and 2026-27 enacted estimates from LAO Figure 3 (shown hollow). % of revenues uses DOF Chart A revenues.
- Unit
- US$ billions / % of GF revenues
- Frequency
- Annual (end of fiscal year)
- Geography
- BSA balance ($B); BSA as % of GF revenues
- Coverage
- From 30 Jun 2016 · latest observation 30 Jun 2027
- Release
- Jan (Governor's Budget), Jun/Jul (enacted)
- Typical lag
- Actuals ~6-18 months
- Source
- BSA / DOF Chart A · lao.ca.gov ↗
- Vintage
- Governor's Budget 2017-18, Fund 9658 past-year column · Governor's Budget 2018-19, Fund 9658 past-year column · Governor's Budget 2019-20, Fund 9658 past-year column · Governor's Budget 2020-21, Fund 9658 past-year column · Governor's Budget 2021-22, Fund 9658 past-year column · Governor's Budget 2022-23, Fund 9658 past-year column · Governor's Budget 2023-24, Fund 9658 past-year column · Governor's Budget 2024-25, Fund 9658 past-year column · Governor's Budget 2025-26, Fund 9658 past-year column · Governor's Budget 2026-27, Fund 9658 past-year column · LAO 2026-27 Spending Plan
- Pipeline
- semi-automated (PDF parsed; LAO estimates hard-coded in build_metrics.bsa)
Every observation in this file
Loading…
Think this chart is wrong?
Errors in a figure, an axis, a calculation or a citation don’t need a competing dataset — just the evidence. Accepted fixes are published on the corrections page.