Routine County Public Health Budgets Lower Premature Deaths After Separating Outbreak Spending
We compare pandemic emergency surge spending with annual and multi-year routine budget models across California counties.
County boards review health budgets across calm years and outbreak years. From 2020 to 2023, state and federal relief grants rose fast in counties with high death rates. Splitting outbreak aid from base care shows how steady county health funds save lives.
Tracking routine pre-pandemic budget shifts across 943 county-years (2003 to 2019) shows that each $10 per resident in county health spending prevents 2.38 deaths per 100,000 annually ($32 in health value per $1 spent) and 4.32 deaths per 100,000 over 13-year cumulative horizons ($59 per $1 spent).[1][2] During the 2020 to 2023 pandemic window, emergency grants jumped 64% and show +2.47 deaths per 100,000 per $10 because surge funds followed pandemic death spikes.[1][2] Multi-year routine-budget models confirm 5.31 fewer deaths per 100,000 across the single-regime 2003 to 2016 horizon.[3]
Table 1. From Pandemic Surge Feedback to Annual and Multi-Year Routine-Budget Checks
We compare four models of local health spending ($10 per resident) and death rates per 100,000.
| Model Measure |
Model 1: Pandemic Surge
COVID-19 Surge Window
|
Model 2: Annual Panel
Pre-COVID Annual Baseline
|
Model 3: 13-Yr Span
13-Year Cumulative Change
|
Model 4: Routine-Budget IV
Supply-Shift IV Check
|
|---|---|---|---|---|
| Source of Budget Shifts | Emergency COVID-19 relief+64.2% | Within-county routine budgets | 13-year sustained budget shifts | Variance-based routine-budget check |
| Deaths per $10 per Resident | +2.47pandemic surge | 2.38 fewer deathsannual | 4.32 fewer deaths13-yr | 5.31 fewer deathsroutine IV |
| Effect Range & Precision | +1.00 to +3.94 range | 0.85 to 3.91 fewer range | 1.30 to 7.34 fewer range | 2.04 to 8.58 fewer range |
| Subgroup & Regime Checks | Pooled 2003-23 attenuated-0.25 | Urban (-2.59) & Rural (-2.42) | 16-yr 2003-19 span: 5.91 fewer | Strong first-stage & valid instruments |
| Cost per Life Saved | Not defined+2.47 sign | $420,168 per life saved | $231,481 per life saved | $188,324 per life saved |
| Return per $1 Spent | Looks like 0:1reverse feedback | $32 per $1$13.6M life value | $59 per $1$13.6M life value | $72 per $1$13.6M life value |
Table 1 Notes & * Modeled Estimates
Expand notes
Table 1 Notes (Mode B Empirical Estimates): We use county-clustered standard errors across 57 California counties. The 2020-2023 pandemic window (N = 220) has β = +0.247 per $1 (+2.47 per $10, SE = 0.75, p = 0.0010), while pooled 2003-2023 TWFE (N = 1,163) is attenuated to -0.25 per $10 (p = 0.710). The 2003-2019 pre-COVID TWFE OLS baseline (N = 943) has β = -0.238 per $1 (-2.38 per $10, SE = 0.78, p = 0.0022, wild cluster bootstrap p = 0.007, 95% CI -3.91 to -0.85). The 13-year cumulative long-difference OLS (2003-2016, N = 57) has -4.32 per $10 (SE = 1.54, p = 0.0052; 16-year 2003-2019 excl. 5 LGRS outliers is -5.91, p = 0.0330). The 13-year Long-Difference Lewbel (2012) 2SLS model (2003-2016, N = 57) has -5.31 per $10 (SE = 1.67, p = 0.0015, first-stage F = 111.20, Hansen J p = 0.2132). Cost per life saved is $10 / (|β_10| / 100,000).
Column Scope. COVID-19 Surge Window. 57 Counties (2020-2023, N=220).. Pre-COVID Annual Baseline. 57 Counties (2003-2019, N=943).. 13-Year Cumulative Change. 57 Counties (2003-2016, N=57).. Supply-Shift IV Check. 13-Yr Routine-Budget Model (2003-2016, N=57)..
Model Crosswalk. Plain table labels map to standard study models. Raw Yearly Average / Simple Avg is an unadjusted OLS model where routine and outbreak budgets mix. County & Year FE is a Two-Way Fixed Effects (TWFE) panel model. Routine-Budget Analysis / 21-Year County Budget Model is a Two-Stage Least Squares (2SLS / Lewbel 2012) model that isolates routine base budget shifts. County Board Budget Rules uses Brown (2014) county board rules.
Two-Ruler Benchmark Summary. Mode B Empirical Estimates (N = 943 Pre-COVID & N = 220 COVID-19 County-Years, 2003 to 2023). Ruler 1: Pandemic Emergency Window (2020-2023 Surge Grants): +2.47 Deaths / 100K (4.85-Point Sign Flip). During 2020 to 2023 (N = 220), county health spending jumped +64.2% ($99.65 to $168.70/capita) as federal COVID-19 relief followed mortality waves, flipping the slope to +2.47 deaths/100K per $10 and attenuating pooled 2003-2023 TWFE to -0.25. Ruler 2: Routine-Budget Analysis (Annual TWFE & Cumulative Horizons): 2.38 Yr | 4.32 13-Yr | 5.31 2SLS Fewer / 100K. Tracking routine pre-pandemic budgets (2003 to 2019, N = 943) gives 2.38 fewer deaths/100K annually. Over 13-year cumulative horizons (2003 to 2016), protection builds to 4.32 fewer deaths/100K in OLS and 5.31 fewer in long-difference 2SLS.
Method & Data Audit (TooEarlyToSay.com): Public Health Spending Regime Analysis →
Table 2. Why Separating Pandemic Surges and Exposure Horizons Builds Budget Trust
We show how the annual county panel, 13-year cumulative differences, and routine-budget IV models reinforce each other.
| Choice Factor | COVID-19 Surge Window | Pre-COVID Annual Baseline | 13-Year Cumulative Change | Supply-Shift IV Check |
|---|---|---|---|---|
| Data & Time Scope | Emergency pandemic window across 57 counties from 2020 to 2023. | Routine pre-pandemic panel across 57 counties from 2003 to 2019. | 13-year single-regime cumulative change from 2003 to 2016. | 13-year routine-budget IV model from 2003 to 2016. |
| How Outbreak Funds Are Handled | Captures +64.2% COVID-19 grant surges following pandemic death spikes (+2.47). | Separates routine prevention budgets ($99.65/cap) from pandemic relief ($168.70/cap). | Accumulates multi-year chronic disease and heart gains (48.4% heart). | Uses variance-based instruments with strong first-stage checks. |
| Budget Risk if Misread | Pooling 2020-2023 with 2003-2019 attenuates the slope to -0.25. | Judging budgets only on 1-year returns omits 45% of cumulative 13-year gains. | Ignoring the FY 2016-17 state reporting schedule shift creates artificial jump outliers. | Single-year IV has weak first stages, whereas 13-year cumulative IV is strong. |
| Best Policy & Budget Action | Separate Emergency Relief Windows. Never pool emergency pandemic grants with baseline prevention budgets. | Use -2.38 ($32:$1) as Yr 1 Floor. Use 2.38 fewer deaths for conservative annual budgeting. | Use -4.32 ($59:$1) for Multi-Year ROI. Show boards that sustained 13-year investments save lives at $231,481 each. | Cite 13-Yr Routine-Budget IV (-5.31). Show that outbreak spending mix-corrected cumulative returns reach $72 per $1. |
Table 2 Notes: Regime Separation, Exposure Horizons & Lewbel 2SLS Diagnostics
Expand notes
• Pandemic Regime Separation: During 2020 to 2023 (N = 220), intergovernmental COVID-19 grants pushed mean county public health spending up 64.2% (from $99.65 to $168.70 per capita in 2023 USD) while mortality rose from 703.4 to 795.4 per 100,000. That reverse-causality surge yields β = +0.247 per $1 (+2.47 per $10, p = 0.0010), a 4.85-point shift above the pre-pandemic baseline (-2.38 per $10, p = 0.0022).
• Subgroup & Cause-Specific Channels: Within the 2003 to 2019 pre-COVID baseline (N = 943), protective effects hold in both large urban counties (>500,000 residents: -2.59 per $10, p < 0.0001) and smaller/rural counties (≤500,000 residents: -2.42 per $10, p = 0.0136), as well as the single-regime 2003-2016 SCO schedule (-2.62 per $10, p = 0.0128). Avoidable chronic disease mortality falls by -2.84 per $10 (p = 0.0022), led by heart disease (-1.15 per $10, p = 0.0066, accounting for 48.4% of the all-cause reduction).
• Data Provenance: Exact slopes and diagnostics come from src/frontend/js/lha-calculator.js, public/tools/lha-calculator/method.html, and Value_of_Public_Health/paper/ajph_submission.tex (Tables 1-3 and Online Appendix Table A3).
• Extended Cell Context. Data & Time Scope (COVID-19 Surge Window). Emergency pandemic window across 57 counties from 2020 to 2023 (N = 220). · Data & Time Scope (Pre-COVID Annual Baseline). Routine pre-pandemic panel across 57 counties from 2003 to 2019 (N = 943). · Data & Time Scope (13-Year Cumulative Change). 13-year single-regime cumulative change from 2003 to 2016 (N = 57). · Data & Time Scope (Supply-Shift IV Check). 13-year routine-budget IV model from 2003 to 2016 (N = 57).
Why Regime Separation and Multi-Year Horizons Build Budget Trust
County boards and state lawmakers set health budgets in real time. During the 2020 to 2023 COVID-19 emergency, federal and state relief grants boosted county health budgets by 64.2% in response to pandemic death waves. Because emergency aid chased high mortality (+2.47 deaths per 100,000 per $10, p = 0.001), pooling pandemic emergency years with routine budgets makes steady prevention programs look flat (-0.25, p = 0.710).
Separating routine pre-pandemic prevention budgets (2003 to 2019, N = 943) from pandemic surge grants reveals clear, compounding mortality reductions across California's 57 county health departments. Each $10 per resident averts 2.38 deaths per 100,000 annually ($420,168 per life saved, $32 return per $1), building to 4.32 fewer deaths over 13-year cumulative horizons ($231,481 per life, $59 return per $1), 5.91 over 16 years ($80 return per $1), and 5.31 in 13-year routine-budget models ($72 return per $1). Budget leaders can back base prevention with verified annual and multi-year proof.
Data Sources, Model Inputs, Notes & Endnotes
(Numbered endnotes [1]–[5], state data sources & math rules)
Expand details
Case Study Focus: Study Credibility & Measurement · Routine-Budget Identification (Mode B 57-County Panel)
Numbered Endnotes & Main Sources
- [1]California State Controller's Office: County Financial Transactions Reports & LGRS Schedule Harmonization (57-County Public Health Spending Panel, 2003–2023, N = 943 pre-COVID and N = 220 COVID-19 era county-years).
- [2]CDPH Vital Statistics: County All-Cause & Cause-Specific Mortality Files (2003–2023) & CAPHE LHA Methodology.
- [3]Brown, T. T. (2014): "How Effective Are Public Health Departments at Preventing Mortality?", Economics & Human Biology, 13, 34–45; and Lewbel, A. (2012), JBES, 30(1), 67–80.
- [4]Brown, T. T. (2016), American Journal of Public Health, 106(8), 1477-1482
- [5]Lewbel, A. (2012), JBES, 30(1), 67-80
Mode B Panel & Published Studies: All slopes (+2.47 pandemic surge, -0.25 pooled 2003-2023, -2.38 pre-COVID annual TWFE, -2.62 single-regime 2003-2016, -2.59 urban, -2.42 smaller/rural, -4.32 13-year long-difference OLS, -5.91 16-year long-difference OLS, and -5.31 / -11.49 Long-Difference Lewbel 2SLS), standard errors, and diagnostics (F = 111.20, Hansen J p = 0.2132) are Mode B estimates from the 57-county working-paper panel (ajph_submission.tex).
Main Sources: We cite Brown, T. T. (2014), "How Effective Are Public Health Departments at Preventing Mortality?", Economics & Human Biology, 13, 34-45 and Brown, T. T. (2016), American Journal of Public Health, 106(8), 1477-1482. We also use Lewbel, A. (2012), JBES, 30(1), 67-80, California State Controller County Financial Transactions, and CDPH Vital Statistics Mortality Data.
Companion Method & Code Guides (TooEarlyToSay.com)
- Public Health Spending: When Emergency Aid Flips the Sign. Across 57 California counties from 2003 to 2023 (1,163 county-years), each dollar per capita corresponds to -0.238 fewer deaths per 100,000 in 2003 to 2019 versus +0.247 in 2020 to 2023.
Common Questions & Method Notes
- Why does the 2020-2023 pandemic window (+2.47) have the opposite sign of the pre-pandemic routine budget result (2.38 fewer deaths)?
- During 2020 to 2023, emergency COVID-19 relief grants pushed county public health spending up 64.2% ($99.65 to $168.70 per capita) in response to pandemic mortality waves. That reactive surge (+2.47 deaths/100K per $10) reverses the sign when pooled with routine pre-pandemic prevention (2.38 fewer deaths/100K).
- Why do 13-year and 16-year cumulative long-difference models find larger effects (-4.32 to -5.91) than annual TWFE (-2.38)?
- Public health prevention works primarily through chronic disease control (where heart disease accounts for 48.4% of the mortality drop). Sustained multi-year investments compound over 4 to 16 years and filter out single-year budget noise, nearly doubling the protective effect from 2.38 to 4.32-5.91 fewer deaths per 100,000.
- Does public health spending reduce mortality in both urban and rural California counties?
- Yes. In the 2003 to 2019 pre-COVID panel (N = 943), a $10 per capita increase reduces mortality by 2.59 deaths per 100,000 in large urban counties (>500,000 residents, p < 0.0001) and by 2.42 deaths per 100,000 in smaller and rural counties (≤500,000 residents, p = 0.0136).