# GovParti Civic Intelligence Platform
## Comprehensive Intelligence Report

**Generated:** March 28, 2026  
**Coverage:** 3,219 U.S. Counties | 16 Active Data Sources | 109+ Confirmed Findings  
**Rule:** All data sourced from authoritative U.S. government agencies. No synthetic or imputed values.

---

## Executive Summary

GovParti's county intelligence layer covers the full American county landscape — 3,205 counties, ~329M residents — across 13 government data sources. The data reveals stark structural inequalities:

- A **21.5-year spread** in life expectancy at birth (67.99 yrs → 89.50 yrs)
- National average unemployment: **2.77%** — but counties range from 0.06% to 14.75%
- **918 counties** (29%) have no recorded grocery stores
- **$6.6 billion** in campaign finance concentrated in a handful of wealthy counties
- **2,791 counties** (86.7%) have at least one primary care shortage area
- National average civic health score: **42.7/100** — ranging from 0.0 to 79.9

---

## Part I: Data Source Inventory

| Source | Agency | Counties Covered | Status |
|---|---|---|---|
| Demographics (40+ ACS vars) | U.S. Census Bureau | 3,205 | ✅ |
| Healthcare Access (HPSA) | HRSA | 3,205 | ✅ |
| Environmental Justice (EJScreen) | EPA | 3,205 | ✅ |
| Disaster History (FEMA, 1953–) | FEMA | 3,204 | ✅ |
| Community Health (27 metrics) | CDC PLACES | 3,126 | ✅ |
| Election History 2000–2024 | MIT Election Lab | 3,102 | ✅ |
| Traffic Safety (FARS, 5yr avg) | NHTSA | 3,100 | ✅ |
| Income & Tax Returns (IRS SOI) | IRS | 3,125 | ✅ |
| Food Environment (18 metrics) | USDA | 3,127 | ✅ |
| Life Expectancy | CDC / NCHS | 3,093 | ✅ |
| Employment & Unemployment (ACS) | Census / BLS | 3,222 | ✅ |
| Campaign Finance 2022 (FEC bulk) | FEC | 3,064 | ✅ |
| Fixed Broadband Coverage | FCC Form 477 Area Table 2021 | 3,213 | ✅ |
| Medicare Spending & Utilization | CMS Geographic Variation PUF 2022 | 3,146 | ✅ |
| VA Facility Distance | VA Lighthouse API + Census CenPop 2020 | 3,213 | ✅ |
| Federal Spending FY2025 | USASpending.gov | 226+ | 🔄 Expanding |
| Labor Organizing (NLRB) | NLRB | 3,205 | ⚠️ Baselines only |
| Substance Use Facilities | SAMHSA | 3,205 | ❌ CAPTCHA blocked |
| Drug Overdose Death Rates | CDC WONDER | — | 🔄 Pending |

---

## Part II: Life Expectancy & Health Geography

### National Range

| Metric | Value |
|---|---|
| Average life expectancy | **77.8 years** |
| Highest county | 89.50 years |
| Lowest county | 67.99 years |
| Spread (max − min) | **21.5 years** |

The 21.5-year spread is equivalent to eliminating all cardiovascular disease mortality — driven by the compounding effects of income, geography, healthcare access, food environment, and social determinants.

### Longest-Lived Counties

| County | State | Life Exp. | Population | Avg Income |
| --- | --- | --- | --- | --- |
| Cheyenne County, Colorado | CO | 89.50 yrs | 1,732 | $57,000 |
| Wayne County, Utah | UT | 89.30 yrs | 2,557 | $67,000 |
| Haskell County, Kansas | KS | 88.60 yrs | 3,695 | $71,000 |
| Stanton County, Kansas | KS | 87.90 yrs | 2,018 | $67,000 |
| Custer County, Idaho | ID | 87.70 yrs | 4,411 | $60,000 |
| Hampshire County, Massachusetts | MA | 87.64 yrs | 156,595 | $88,000 |
| Crook County, Wyoming | WY | 87.55 yrs | 7,339 | $79,000 |
| Sherman County, Texas | TX | 87.30 yrs | 2,434 | $57,000 |

### Shortest-Lived Counties

| County | State | Life Exp. | Population | Avg Income |
| --- | --- | --- | --- | --- |
| Bibb County, Georgia | GA | 67.99 yrs | 156,543 | $62,000 |
| Wayne County, North Carolina | NC | 68.21 yrs | 117,606 | $54,000 |
| Breathitt County, Kentucky | KY | 68.22 yrs | 13,438 | $41,000 |
| Logan County, West Virginia | WV | 68.26 yrs | 31,826 | $49,000 |
| Mobile County, Alabama | AL | 69.15 yrs | 413,162 | $64,000 |
| Kenton County, Kentucky | KY | 69.56 yrs | 169,817 | $94,000 |
| Baltimore city, Maryland | MD | 69.61 yrs | 577,193 | $65,000 |
| Trumbull County, Ohio | OH | 69.64 yrs | 201,367 | $54,000 |

### Healthcare Access Shortages (HRSA HPSA Designations)

| Shortage Type | Counties Affected |
|---|---|
| Primary care physicians | **2,791** / 3,219 |
| Mental health providers | **2,911** / 3,219 |
| Dental providers | **2,549** / 3,219 |

---

## Part III: Economic Geography (IRS SOI 2021)

IRS Statistics of Income provides 2021 tax year data for 3,125 counties. Average AGI per return serves as a reliable proxy for household income.

- **National average AGI per return:** $66,573
- Counties in the bottom income quintile average **75.7 years** life expectancy
- Counties in the top income quintile average **79.7 years** life expectancy
- Income-to-longevity gradient: **+1.0 year of life per income quintile step upward**

---

## Part IV: Employment & Unemployment (ACS 2023, 3,222 Counties)

Labor market conditions are a foundational civic health indicator — unemployment correlates with poverty, health outcomes, and civic disengagement.

### National Labor Market

| Metric | Value |
|---|---|
| Counties with employment data | **3,222** |
| National average unemployment rate | **2.77%** |
| Lowest county rate (10k+ pop) | **0.31%** (Trousdale County, Tennessee, TN) |
| Highest county rate (10k+ pop) | **11.73%** (St. Helena Parish, Louisiana, LA) |

### Unemployment as a Health & Wealth Predictor

| Unemployment Tier | Counties | Avg Life Expectancy | Avg Income |
| --- | --- | --- | --- |
| Low (<3%) | 2024 | 78.0 yrs | $67,750 |
| Medium (3-6%) | 1019 | 77.2 yrs | $65,407 |
| Very High (9%+) | 9 | 77.2 yrs | $45,667 |
| High (6-9%) | 33 | 77.1 yrs | $45,848 |

Low-unemployment counties have **0.8 extra years** of life expectancy versus medium-unemployment counties, and higher unemployment is associated with dramatically lower household incomes. The income gap between low- and high-unemployment counties exceeds $22,000 per return.

### Highest Unemployment Counties (U.S. only, 10k+ pop)

| County | State | Unemployment Rate | Population |
| --- | --- | --- | --- |
| St. Helena Parish, Louisiana | LA | 11.73% | 10,849 |
| Bethel Census Area, Alaska | AK | 8.52% | 18,487 |
| Big Horn County, Montana | MT | 8.06% | 12,963 |
| Phillips County, Arkansas | AR | 8.01% | 15,910 |
| Jasper County, Texas | TX | 7.80% | 32,807 |
| Petersburg city, Virginia | VA | 6.91% | 33,365 |
| San Juan County, Utah | UT | 6.61% | 14,466 |
| Imperial County, California | CA | 6.59% | 179,319 |

### Lowest Unemployment Counties (10k+ pop)

| County | State | Unemployment Rate | Population |
| --- | --- | --- | --- |
| Trousdale County, Tennessee | TN | 0.31% | 11,805 |
| Hughes County, South Dakota | SD | 0.38% | 17,732 |
| Pope County, Minnesota | MN | 0.56% | 11,363 |
| Davison County, South Dakota | SD | 0.58% | 19,936 |
| Nemaha County, Kansas | KS | 0.58% | 10,213 |
| Mitchell County, Iowa | IA | 0.64% | 10,548 |
| Pecos County, Texas | TX | 0.64% | 14,983 |
| Cass County, Iowa | IA | 0.72% | 13,115 |

---

## Part V: Food Environment — The Food Infrastructure Paradox (USDA 2021)

*Source: Working Paper No. 3*

### Infrastructure Crisis

| Metric | Value |
|---|---|
| Counties with zero grocery stores | **918 (29.2%)** |
| Counties with zero farmers markets | **872 (27.7%)** |
| National avg food insecurity rate | **12.5%** |
| National fast food : grocery ratio | **4.1:1** |

### The Fast Food Paradox: Ratio Increases With Wealth

| Income Quintile | Avg Income | Fast Food/1k | Grocery/1k | FF:Grocery | Life Exp. |
| --- | --- | --- | --- | --- | --- |
| Q1 (Poorest) | $45,144 | 0.61 | 0.260 | 3.3:1 | 75.7 yrs |
| Q2 () | $54,213 | 0.64 | 0.237 | 3.8:1 | 76.9 yrs |
| Q3 () | $61,043 | 0.64 | 0.215 | 4.0:1 | 77.8 yrs |
| Q4 () | $70,005 | 0.69 | 0.225 | 4.4:1 | 78.7 yrs |
| Q5 (Wealthiest) | $102,515 | 0.75 | 0.198 | 4.8:1 | 79.7 yrs |

### The Farmers Market Effect

| Farmers Market Density | Counties | Avg Life Exp. | Food Insecurity |
| --- | --- | --- | --- |
| High (>0.15/1k) | 281 | 79.0 yrs | 11.2% |
| Medium | 835 | 77.9 yrs | 12.1% |
| Low (<0.05/1k) | 1148 | 77.6 yrs | 12.5% |
| Zero markets | 844 | 77.6 yrs | 13.3% |

**1.4 extra years of life** are associated with high farmers market density versus zero markets. Farmers market investment is a high-return, community-controlled public health lever at a fraction of hospital construction costs.

---

## Part VI: FEC Campaign Finance (2022 Election Cycle)

### National Picture

Full FEC 2022 individual contributions bulk file (4.9GB compressed, 63.9M records processed):

| Metric | Value |
|---|---|
| Total contribution transactions | **59,731,954** |
| Total dollars raised (itemized) | **$6.63 billion** |
| Small-dollar transactions | 54,539,919 (91.3%) |
| Counties with FEC activity | **3064** / 3,219 |

*Note: Counts FEC transaction records, not unique individuals.*

### Top 10 Counties by Transaction Volume

| County | State | Transactions | Raised | Rate/1k |
| --- | --- | --- | --- | --- |
| Los Angeles County, California | CA | 2,098,771 | $290.4M | 213.1 |
| New York County, New York | NY | 996,780 | $311.5M | 612.4 |
| Maricopa County, Arizona | AZ | 976,966 | $94.1M | 217.5 |
| Cook County, Illinois | IL | 960,575 | $162.1M | 185.2 |
| San Diego County, California | CA | 851,147 | $71.4M | 259.3 |
| King County, Washington | WA | 833,881 | $115.5M | 368.5 |
| Orange County, California | CA | 759,624 | $82.4M | 240.1 |
| Harris County, Texas | TX | 729,872 | $101.9M | 153.4 |
| Alameda County, California | CA | 595,436 | $67.1M | 360.4 |
| Santa Clara County, California | CA | 580,091 | $93.7M | 304.8 |

### Top 12 States by Dollars Raised

| State | Counties | Transactions | Total Raised |
| --- | --- | --- | --- |
| CA | 58 | 9,180,808 | $1,052M |
| NY | 62 | 3,595,046 | $602M |
| FL | 67 | 4,702,227 | $514M |
| TX | 254 | 4,787,741 | $498M |
| IL | 102 | 2,053,203 | $248M |
| MA | 14 | 1,553,025 | $241M |
| PA | 67 | 2,422,881 | $235M |
| VA | 133 | 1,959,339 | $233M |
| WA | 39 | 2,154,446 | $210M |
| DC | 1 | 528,042 | $187M |
| OH | 88 | 1,740,571 | $175M |
| GA | 159 | 1,675,866 | $172M |

### Campaign Finance as a Wealth Proxy

| Donation Activity Tier | Counties | Avg Life Exp. | Avg Income |
| --- | --- | --- | --- |
| Very High (500+/1k) | 134 | 79.1 yrs | $91,739 |
| High (200-500/1k) | 566 | 78.7 yrs | $79,306 |
| Medium (50-200/1k) | 1536 | 77.8 yrs | $65,033 |
| Low (<50/1k) | 746 | 76.9 yrs | $57,181 |

Higher-donation counties are wealthier and longer-lived. The very-high-donation tier shows a **3.4-year life expectancy premium** over low-donation counties — a wealth gradient, not a civic virtue effect.

---

## Part VII: Civic Health Scores

GovParti's 7-component Civic Health Score (0–100) uses percent-rank scoring across economic, health, infrastructure, education, environment, transparency, and civic participation dimensions:

| Metric | Value |
|---|---|
| Counties scored | **3219** |
| National average | **42.7 / 100** |
| Highest score | 79.9 / 100 |
| Lowest score | 0.0 / 100 |

### Top 10 — Highest Civic Health

| County | State | Score |
| --- | --- | --- |
| San Francisco County, California | CA | 79.9/100 |
| King County, Washington | WA | 79.5/100 |
| Hennepin County, Minnesota | MN | 79.4/100 |
| Marin County, California | CA | 79.3/100 |
| Los Alamos County, New Mexico | NM | 78.1/100 |
| Dane County, Wisconsin | WI | 78.0/100 |
| Fairfax County, Virginia | VA | 77.0/100 |
| San Diego County, California | CA | 74.4/100 |
| Denver County, Colorado | CO | 74.4/100 |
| Montgomery County, Maryland | MD | 74.1/100 |

### Bottom 10 — Lowest Civic Health

| County | State | Score |
| --- | --- | --- |
| Río Lajas barrio, Dorado Municipio, Puerto Rico | PR | 12.1/100 |
| Río Grande barrio, Rincón Municipio, Puerto Rico | PR | 12.8/100 |
| Pueblo barrio, Corozal Municipio, Puerto Rico | PR | 14.3/100 |
| Yaurel barrio, Arroyo Municipio, Puerto Rico | PR | 14.4/100 |
| Terranova barrio, Quebradillas Municipio, Puerto Rico | PR | 14.6/100 |
| Torre barrio, Sabana Grande Municipio, Puerto Rico | PR | 15.2/100 |
| Toa Alta barrio-pueblo, Toa Alta Municipio, Puerto Rico | PR | 15.3/100 |
| Sunflower County, Mississippi | MS | 15.3/100 |
| Palmas barrio, Cataño Municipio, Puerto Rico | PR | 15.5/100 |
| Macon County, Georgia | GA | 15.8/100 |

---

## Part VIII: Traffic Safety (NHTSA FARS 2019–2023)

5-year average fatality data from NHTSA's Fatality Analysis Reporting System:

| Metric | Value |
|---|---|
| Counties with data | 3,109 |
| Avg annual fatalities | 71,013 |
| Average rate | 42.8 per 100,000 residents |
| Highest county rate | 1223.7 per 100,000 |

Rural counties bear a disproportionate traffic fatality burden: higher speeds, longer emergency response times, and limited trauma center access combine to make rural roads dramatically more deadly per mile traveled than urban roads.

---

## Part IX: Federal Spending (USASpending.gov FY2025)

*Phase 1 ingestion running — coverage expanding toward all 3,219 counties.*

| Metric | Value |
|---|---|
| Counties covered (live) | **226+** |
| Total federal obligations tracked | **$0.51 trillion** |
| Avg federal spending per capita | **$12,391** |

Federal spending distribution reveals structural patterns: rural counties often receive more federal dollars per capita through direct payments (Social Security, Medicare, farm subsidies), while urban metros generate more federal tax revenue than they receive.

---

## Part X: Disaster History (FEMA 1953–Present)

- **3,213 counties** with FEMA disaster declaration records
- **67,529 total disaster declarations** since program inception
- Hurricane corridors (FL, LA, TX), tornado alley (OK, KS, NE), and wildfire zones (CA, OR, WA) record the highest declaration rates
- COVID-19 (2020) generated the largest single-year declaration surge in program history

---

## Part XI: CMS Medicare Data — The Healthcare Spending Paradox (Sessions 14–15)

*Source: CMS Geographic Variation Public Use File 2022 (3,143–3,146 counties); VA Lighthouse API + Census CenPop 2020 centroids (3,213 counties); USDA Food Environment Atlas 2024 (3,090 counties)*

### The Medicare Paradox: More Spending, Worse Outcomes

The most counterintuitive finding in the entire GovParti dataset: **America spends the most Medicare dollars per person in the counties where people live the shortest lives.**

| Cohort | Life Expectancy | Medicare Spend/Beneficiary | ER Visits/1,000 | Readmission Rate |
|---|---|---|---|---|
| Lowest (<75yr, n=1,234) | <75 years | **$11,588** | **646** | **17.42%** |
| Highest (≥80yr, n=281) | ≥80 years | $9,990 | 508 | 15.58% |
| **Gap** | **>5 years** | **+$1,598 more** | **+27% higher** | **+1.84pp higher** |

This is the county-level **Dartmouth Atlas effect**: higher Medicare spending tracks sicker populations and emergency-based care, not better health. The sickest communities cost Medicare the most — not because they receive better care, but because untreated chronic disease managed through emergency departments costs dramatically more than prevented disease managed through primary care.

*National averages: $10,856/beneficiary | 596.7 ER visits/1,000 | 16.3% readmission rate*

### VA Geographic Abandonment — Distance Confirms the Gap

VA facility distance data (3,213 counties, avg **21.8 miles**) reveals a care displacement pattern:

| Distance Zone | Medicare ER Utilization | Medicare Spend/Beneficiary | Interpretation |
|---|---|---|---|
| <25 miles (VA nearby) | 573/1,000 | $10,634 | VA absorbs veteran primary care |
| 26–50 miles (displacement zone) | **606/1,000** | **$11,122** | Veterans substitute Medicare ER for VA |
| >75 miles (hyperremote) | 435/1,000 | $9,596 | Low utilization = care avoidance, not health |

**115 counties** are effectively beyond VA reach (>60 miles). The 26–50 mile "displacement zone" is the most expensive: veterans who lose convenient VA access substitute costly Medicare emergency visits rather than forgoing care, generating the highest per-beneficiary spending in the dataset.

### Medicare Advantage Adoption Paradox

MA plans concentrate in counties with the highest chronic disease burden — the exact opposite of what market selection would predict:

| MA Adoption | Spend/Beneficiary | ER/1,000 | Readmission | Life Expectancy |
|---|---|---|---|---|
| Highest (≥60%, n=335) | $10,981 | 625 | 18.03% | 75.30yr |
| Lowest (<30%, n=782) | $10,676 | 567 | 15.11% | 76.75yr |

MA extra benefits (dental, vision) attract sicker enrollees. MA is being adopted most where it has the least ability to overcome structural health deficits.

### SNAP: One of the Few Programs That Correctly Targets Distress

Per-capita SNAP benefits ($30.92 national average, 3,090 counties) follow a clean gradient toward higher need — a rare exception to the inverse-care-law pattern seen throughout this dataset. Importantly: high food-insecurity counties (16–20% insecure) receive $32.39/capita vs. $33.54 in moderate food-insecurity counties — a $1.15 **SNAP undercoverage gap** that likely reflects structural barriers (fewer authorized stores, lower take-up rates, work requirement cliffs) rather than policy intent.

### Key Findings (F97–F101)

- **F97 — Medicare ER Utilization Gap: Rural Primary Care Deficit**: Nonmetro counties show 607.4 ER visits/1,000 vs 579.5 in Metro (+4.8%). Mid-range counties 26–50 miles from VA facilities show highest ER utilization (606.5/1,000) and highest spending ($11,122/beneficiary) — geographic fragmentation of care delivery.
- **F98 — "Spend More, Live Less" Medicare Paradox**: Counties with lowest life expectancy spend $1,598 MORE per Medicare beneficiary, have 27% higher ER utilization, and 1.84pp higher readmission rates than highest life expectancy counties. Poverty explains most of the gradient: 18.9% poverty rate in <75yr counties vs 9.5% in ≥80yr.
- **F99 — VA Distance and Healthcare Utilization Displacement**: The 26–50mi displacement zone generates the highest Medicare ER utilization (606.5/1,000) and spending ($11,122). Hyperremote counties >75mi show lowest utilization (434.8/1,000) — not better health, but care avoidance. 115 counties confirmed beyond practical VA reach.
- **F100 — Medicare Advantage Adoption Paradox**: Highest MA adoption counties show WORSE outcomes on every metric — a selection effect where extra benefits attract the sickest enrollees. MA is being adopted most where it can help least.
- **F101 — SNAP Coverage Gap: High-Need Counties Get Less**: Per-capita benefits peak in moderate food insecurity counties ($33.54) then fall in high food insecurity counties ($32.39) — a $1.15 gap that widens as need increases. Structural barriers drive undercoverage, not policy intent.

---

## Part XII: Working Papers

### Working Paper No. 2
**Compounding Deprivation: Income, Longevity, and the Geography of Civic Disadvantage**  
📄 `papers/working_paper_02_compounding_deprivation.md`

Key findings: 21.5-year life expectancy spread; r≈0.65 income-longevity correlation; ~12M Americans in compounding deprivation (bottom quintile both income AND life expectancy); Mississippi, West Virginia, Alabama, Kentucky dominate low-end cluster.

### Working Paper No. 3
**The Food Infrastructure Paradox: Commercial Density, Community Markets, and Population Health Across 3,144 American Counties**  
📄 `papers/working_paper_03_food_infrastructure_paradox.md`

Key findings: Fast food-to-grocery ratio INCREASES with income (3.3:1 poorest → 4.8:1 wealthiest); 918 counties have no grocery stores; 1.4-year life expectancy premium for high farmers market counties; Arkansas Delta worst case: Lonoke County, 18.9% food insecurity, 72.6 years life expectancy.

### Working Paper No. 4
**Money, Power, and the Illusion of Civic Finance: Campaign Contributions as a Wealth Proxy Across 3,064 American Counties**  
📄 `papers/working_paper_04_campaign_finance_wealth_proxy.md`

Key findings: Campaign finance participation rate is nearly perfectly correlated with income decile (r≈0.96–0.98 across deciles) — making FEC data a powerful wealth thermometer rather than a civic virtue indicator. Small-dollar transactions constitute 91.3% of transaction volume but only 25.7% of total dollars, revealing structural dominance by large donors. The life expectancy premium from lowest to highest donation tier is 3.4 years — identical to the income effect. The gradient plateaus at Tier 3→4 (high → very high), consistent with marginal political giving by the already-wealthy rather than civic engagement.

---

## Part XIII: Open Gaps & Next Steps

### Recently Resolved

| Gap | Resolution | Session |
|---|---|---|
| FCC broadband coverage (25/100 Mbps) | FCC Form 477 Area Table (Socrata `xvwq-qtaj`, June 2021) — 3,213 counties, all speed tiers | Session 23 |
| CMS Medicare spending & utilization | CMS Geographic Variation PUF 2022 — 3,143–3,146 counties | Session 14–15 |
| VA facility distances | VA Lighthouse API + Census CenPop 2020 haversine — 3,213 counties | Session 15 |

### Currently Filling Automatically

| Gap | Action | Status |
|---|---|---|
| Federal spending (all 3,219 counties) | USASpending Phase 1 running | 🔄 226+ counties done |
| Drug overdose death rates | Testing CDC accessible endpoints | 🔄 In progress |

### Structurally Blocked — Confirmed Dead Ends

| Gap | Blocker | What's Needed |
|---|---|---|
| SAMHSA substance use facilities | reCAPTCHA v3 on facility locator | N-SSATS public-use file from icpsr.umich.edu (free account) |
| NLRB labor organizing cases | Public API returns no data | Full export from nlrb.gov/resources/data-downloads |
| CRDC 2020–21 Special Education data | Angular SPA at ocrdata.ed.gov; 403 on data.ed.gov — all paths blocked | CRDC bulk download via direct OCR agency access |
| Drug overdose death rates | CDC WONDER form-only access | NCHS compressed mortality bulk file at nchs.gov |
| FEC donations by party | Bulk file lacks party field | FEC committees22.zip join + re-process of 4.9GB contributions file |

### How You Can Help
Two files that would immediately unlock new analysis:
1. **N-SSATS** — icpsr.umich.edu → search "N-SSATS" → download county-level public use file (substance use treatment facilities)
2. **NLRB** — nlrb.gov/resources/data-downloads → election case export

---

*GovParti Civic Intelligence Platform*  
*All data sourced from U.S. government agencies and public research institutions*  
*Last updated: March 28, 2026*