The article's aggregate California series falls during the recession, rises during the ARRA period, and falls again as support was winding down in 2010-11. That sequence describes timing; it does not identify what ARRA prevented or caused. [1]
The article's aggregate series rises from $50.2 billion to $53.8 billion, a $3.6 billion or 7.2% increase, and then declines by 8.7%. These descriptive comparisons are not publicly reproduced. The table combines federal, state, and local funding, so it does not isolate SFSF or broader ARRA categories.
The Numbers That Don't Add Up
The article's aggregate series rises from $50.2 billion to $53.8 billion, a $3.6 billion or 7.2% increase. The table combines federal, state, and local funding. It cannot be treated as an estimate of State Fiscal Stabilization Fund or total ARRA education support, and no public output currently reproduces it.
ARRA is one plausible contributor to the observed funding series. The State Fiscal Stabilization Fund was one education component of the broader 2009 law, which also included programs such as Title I support. [2] This article does not reproduce a California total for either SFSF or all ARRA education categories.
| Period | CA K-12 Education Funding | Change |
|---|---|---|
| 2007-08 (Pre-recession) | $56.4 billion | Not applicable |
| 2008-09 (Recession) | $50.2 billion | -11.0% |
| 2009-10 (ARRA) | $53.8 billion | +$3.6 billion (+7.2%) |
| 2010-11 (Final ARRA-funded year) | $49.1 billion | -8.7% |
Source: California Department of Education budget summaries; includes federal, state, and local funds. CDE reported $271.4 million in SFSF available in 2010-11, and the Legislative Analyst's Office described ARRA funds as available through that year.
The aggregate funding series rises during the ARRA period. It does not show how many layoffs or program cuts were prevented.
What This Shows
Article-reported timing: The aggregate series remains positive during 2009-2010. That temporal pattern alone does not identify ARRA's causal effect.
The article labels the 2010-11 decline a fiscal cliff. CDE reported $271.4 million in SFSF available that year, and the Legislative Analyst's Office described ARRA funding as available through 2010-11. The aggregate change occurred while support was still available and winding down. The table does not identify a counterfactual.
Then the Cliff
ARRA was explicitly temporary. In 2010-11, support was winding down and some funds remained available.
IDEA's maintenance-of-effort rules constrained some reductions in State and local special-education spending, subject to defined exceptions and an adjustment in certain years. [3] General education does not have the same IDEA Part B funding condition, but districts face other legal, contractual, enrollment, and service constraints.
The budget incidence is therefore an empirical question. Maintenance of effort may narrow some options, but this aggregate series cannot show which categories absorbed a reduction or how exceptions, staffing, enrollment, and district choices affected the result.
Why This Matters
Temporary funding can support programs that later face a different revenue environment. Whether that creates a structural mismatch depends on the program commitments, spending rules, exceptions, and replacement revenue available to each district.
This isn't an argument against stimulus spending. It's an argument for thinking through what happens when temporary money meets permanent mandates.
Limitations
Poverty classification instability. District poverty rates changed substantially between 2011 and 2018 (correlation = 0.28). Using 2018 poverty data to classify districts during the recession period introduces measurement error.
Sign sensitivity. Our main finding reverses depending on which year's poverty data we use to classify districts. With 2011-12 FRPM (contemporaneous to treatment), the coefficient is -0.074 (high-poverty districts cut more). With 2018-19 FRPM, it's +0.047 (high-poverty districts cut less). We report results using 2011-12 data because it measures poverty during the actual recession period, but this specification sensitivity means our estimates should be interpreted with considerable caution. The low correlation between 2011 and 2018 poverty rates (r = 0.28) reflects substantial changes in district composition over time, making retrospective poverty classification inherently unreliable.
Causal identification. The event study shows no significant pre-trends, but this may reflect statistical power limitations. We cannot rule out that high-poverty and low-poverty districts were on different trajectories before the recession.
Data and Methods
Data sources: California SACS financial reports (2006-2021); Free/Reduced Price Meal eligibility rates (CDE); District characteristics (CDE).
Sample: The article describes California district-level SACS records, but its previously stated unified-district count is withdrawn. A public analytic file is needed to identify the included districts and district types.
Critical specification note: Results are sensitive to poverty classification. The main finding changes sign depending on which year's FRPM data is used. See Limitations section for details.
Statistical approach: Difference-in-differences comparing high-poverty (Q5) vs. low-poverty (Q1) districts; event study specification with year fixed effects; standard errors clustered at county level (58 clusters).
Notes
[1] Evans, W. N., Schwab, R. M., & Wagner, K. L. (2019). The Great Recession and public education. Education Finance and Policy, 14(2), 298-326. ↩
[2] U.S. Department of Education. (2009). American Recovery and Reinvestment Act of 2009: State Fiscal Stabilization Fund. ↩
[3] 34 C.F.R. § 300.203 defines the IDEA Part B maintenance-of-effort standards; § 300.204 lists exceptions, and § 300.205 provides a limited adjustment in certain years. ↩
Frequently Asked Questions
What does the aggregate series show during the ARRA period?
The article's aggregate series rises from $50.2 billion to $53.8 billion, a $3.6 billion or 7.2% increase. The table combines federal, state, and local funding, does not isolate SFSF or broader ARRA categories, and is not publicly reproduced.
What happened during the final ARRA-funded year?
The article reports an 8.7% aggregate decline from $53.8 billion to $49.1 billion in 2010-11. CDE reported $271.4 million in SFSF available that year, and the Legislative Analyst's Office described ARRA funds as available through 2010-11. The combined series does not isolate those funds or establish what funding would have been under another support schedule.
What does the 2010-11 decline establish?
The article proposes maintenance-of-effort rules as one explanation for different spending patterns. It does not reproduce a decomposition showing that the rules caused or concentrated all reductions.
What sample and methods were used?
The article describes district-level SACS data and difference-in-differences and event-study specifications, but the previously stated unified-district count is withdrawn because it does not match the statewide universe. No matching public output identifies the analytic sample, and the poverty result changes sign across classifications.
Suggested Citation
Cholette, V. (2025, December 10). The fiscal cliff schools faced after the stimulus. Too Early To Say. https://tooearlytosay.com/research/education-policy/stimulus-saved-schools-then-worse/Copy citation