Structural Correlates of Community Health Resilience: A Cross-Sectional Analysis of 53,889 U.S. Census Tracts

1. Introduction

1.1 The Concept of Community Health Resilience

The concept of community health resilience has gained attention as researchers and practitioners seek to understand why some communities achieve better health outcomes than their socioeconomic circumstances would predict. A community with high poverty and low educational attainment that nonetheless shows lower-than-expected chronic disease burden exhibits resilience—something is protecting residents despite structural disadvantage.

1.2 Who Benefits from Resilience?

But who benefits from this resilience? If the protective factors that enable resilience are inequitably distributed—concentrated in white, affluent communities while absent from communities of color—then resilience-based frameworks could inadvertently reinforce disparities. Celebrating "resilient communities" without examining who is excluded from resilience risks obscuring structural racism.

1.3 Research Questions

This paper examines the equity dimensions of community health resilience across 53,889 U.S. census tracts. We ask:

1. Is resilience equitably distributed by racial composition?
2. What structural factors correlate with higher resilience?
3. Does the relationship between race and resilience vary by state?
4. Who are the positive and negative outliers?

1.4 Structural Correlates Framework

We approach these questions through a structural correlates lens. Health disparities by race are not biological—they co-occur with centuries of residential segregation, disinvestment, environmental racism, and unequal access to resources. If resilience is lower in communities of color, this pattern is consistent with the presence of structural barriers, not community deficits.

The practical implication is that resilience can be cultivated. If some majority-minority communities achieve high resilience, the question becomes: what structural conditions are associated with this? These conditions—not race itself—are the targets for intervention.

2. Methods

2.1 Data Sources

Health Data: CDC PLACES 2020-2024 release, providing model-based small-area estimates for 36 health measures at the census tract level.

Demographic Data: American Community Survey (ACS) 5-Year Estimates 2022, providing racial composition, median household income, poverty rate, unemployment rate, educational attainment, and housing tenure.

2.2 Sample

After merging CDC PLACES with ACS data on census tract GEOID, our analytic sample comprised 53,889 tracts with complete data on health burden, resilience scores, and demographic characteristics.

2.3 Measures

Composite Health Burden Index (CHBI): Standardized composite of seven CDC PLACES measures: obesity, diabetes, coronary heart disease, hypertension, smoking, physical inactivity, and poor mental health days. Higher values indicate greater health burden.

Resilience Score: Residual from OLS regression predicting CHBI from four socioeconomic factors: median household income, poverty rate, unemployment rate, and percent with bachelor's degree or higher (R² = 0.42, indicating the model explains 42% of variance). Positive residuals indicate better-than-predicted outcomes (high resilience); negative residuals indicate worse-than-predicted outcomes (low resilience). Scores are standardized (mean=0, SD=1).

Majority-Minority Classification: Tracts where non-white population exceeds 50%. Sensitivity analysis with 40% and 60% thresholds confirmed robustness: the gap ranged from 0.26 SD to 0.54 SD, all p < 0.001.

2.4 Statistical Analysis

1. Descriptive Statistics: Resilience and burden by racial composition quintiles
2. Correlation Analysis: Pearson correlations between demographics and outcomes
3. Group Comparisons: Independent-samples t-tests comparing majority-white vs. majority-minority communities
4. State Stratification: Resilience gaps computed separately by state
5. Outlier Analysis: Characteristics of top 10% and bottom 10% resilience tracts

3. Results

3.1 Sample Characteristics

Of 53,889 tracts analyzed:

• 26.4% were majority-minority (non-white >50%)
• Mean % Black: 13.4%
• Mean % Hispanic: 17.1%
• Mean poverty rate: 13.5%
• Mean % with bachelor's degree: 32.2%

3.2 Finding 1: The National Resilience Gap

Majority-minority communities averaged significantly lower resilience than majority-white communities:

The raw mean difference is 0.38 SD. Using multilevel models to adjust for state clustering, the coefficient is 0.43 SD (Cohen's d ≈ 0.48)—a small-to-medium effect. It means majority-minority communities, on average, have worse health outcomes than predicted by their socioeconomic characteristics—while majority-white communities have better outcomes than predicted.

3.3 Finding 2: Correlations with Resilience

Educational attainment was the strongest positive correlate of resilience:

The moderate negative correlation between percent Black and resilience (r=-0.34, 95% CI [-0.35, -0.34]) is consistent with structural disadvantage, not community deficits. This was validated post-hoc against USALEEP life expectancy data (r=-0.44), confirming it is not a methodological artifact. Notably, percent Hispanic showed near-zero correlation with resilience (r=+0.01, 95% CI [-0.00, +0.02]), suggesting different structural dynamics—a finding that merits separate investigation.

3.4 Finding 3: Dramatic State-Level Variation

The national 0.43 SD gap masks state-level variation. Some states show much larger gaps; others show near-equity or even reversed patterns:

Interpretation: In Washington and California, majority-minority communities show HIGHER resilience than majority-white communities—a pattern that survives Bonferroni correction. This suggests that structural context, not racial composition per se, is associated with community resilience.

3.5 Finding 4: Who Are the Outliers?

Low-resilience tracts are disproportionately Black (39.7% vs. 13.4%) and majority-minority (56.2% vs. 26.4%). High-resilience tracts are slightly less Black than average but show similar Hispanic composition.

4. Discussion

4.1 Principal Findings

Our analysis reveals a 0.43 standard deviation gap in community health resilience between majority-white and majority-minority communities nationally (multilevel z=-41.83, p<0.001, 95% CI [-0.446, -0.406]). This gap is consistent with structural disadvantage—majority-minority communities appear to face barriers associated with worse health outcomes than their socioeconomic profile would predict.

The state-level variation (from +1.87 SD in DC to -0.42 SD in Washington) suggests that this pattern is not inevitable. Two states (Washington and California) show statistically robust reversed patterns after Bonferroni correction. This variation is consistent with the hypothesis that structural context—not racial composition per se—is associated with community resilience, though cross-sectional design precludes causal inference.

4.2 The Hispanic Paradox in Resilience

One of our most striking findings is the near-zero correlation between percent Hispanic and resilience (r=+0.01), validated at r=+0.002 with USALEEP life expectancy data. This stands in stark contrast to the moderate negative correlation for percent Black (r=-0.34).

This pattern is consistent with the well-documented "Hispanic paradox"—the observation that Hispanic Americans often show health outcomes similar to or better than non-Hispanic whites despite lower average socioeconomic status. Proposed mechanisms include healthy immigrant selection, protective social and family networks, and dietary patterns. This finding merits dedicated investigation.

4.3 The Education Gradient

Educational attainment showed the strongest positive correlation with resilience (r=+0.41, 95% CI [+0.40, +0.42]). This association may operate through multiple pathways:

1. Health literacy: Higher education enables better navigation of health systems
2. Employment quality: College graduates access jobs with better health benefits
3. Neighborhood selection: Education enables residential choice in healthier environments
4. Social capital: Educational networks provide health-promoting resources

The policy implication is that investment in educational equity may be associated with improved health equity, though causal inference requires longitudinal or experimental designs.

4.4 Why Some States Show Equity

Two states—Washington and California—show statistically robust reversed patterns after Bonferroni correction. Potential explanations include:

1. Immigration and selection: The healthy immigrant effect may boost resilience in communities with higher foreign-born populations
2. Policy environment: State-level health and social policies may buffer structural disadvantage
3. Community organization: Strong ethnic enclaves may provide protective social capital
4. Demographic composition: The specific mix of racial/ethnic groups and their geographic distribution may matter

Critical sensitivity finding: Majority-minority tracts in Washington and California have notably higher Asian composition (20.2% and 19.7%, respectively) compared to 10.0% nationally. When excluding Asian-plurality tracts from California, the "reversed" pattern disappears entirely (gap becomes +0.18 SD, favoring white-majority tracts). This suggests California's reversed pattern may be attributable primarily to tracts with high Asian composition rather than a general structural advantage for all minority groups. Washington's reversed pattern persists even after this sensitivity test, suggesting different dynamics.

Important caveats: (1) "Asian-American" aggregates heterogeneous communities with vastly different health patterns (e.g., Hmong vs. Indian American). (2) We interpret this as reflecting structural factors (selective immigration policies, metropolitan healthcare infrastructure), not inherent community characteristics. (3) This pattern should not invoke "model minority" narratives.

These explanations are speculative. Further research should investigate what structural conditions are associated with resilience equity in Washington and California, ideally using longitudinal or quasi-experimental designs that can better address causality.

4.5 Implications for Practice

5. Conclusions

Community health resilience is not equitably distributed. Majority-minority communities average 0.43 standard deviations lower resilience than majority-white communities nationally (multilevel z=-41.83, p<0.001), and the bottom 10% of tracts are 56.2% majority-minority.

State-level variation is substantial—from +1.87 SD (DC) to -0.42 SD (Washington)—and two states (Washington and California) show statistically robust reversed patterns after Bonferroni correction. This suggests that the relationship between racial composition and resilience is context-dependent, consistent with the role of structural factors.

Educational attainment emerged as the strongest correlate of resilience (r=+0.41). Investment in educational equity may be associated with improved health equity.

Practitioners should:

1. Audit resilience metrics for equity before use in resource allocation
2. Target structural determinants rather than labeling communities
3. Learn from states and communities that achieve resilience equity
4. Frame findings around structural barriers, not community deficits

The goal is not to celebrate resilience in some communities while ignoring its absence in others. The goal is to create structural conditions that enable all communities to thrive.

Acknowledgments

We acknowledge the communities represented in this analysis and the structural barriers they face. Data alone does not solve disparities; policy action does.

Data Availability

CDC PLACES and ACS data are publicly available. Analysis code is available at: github.com/cschuman/resilience-mapping