Neighborhood-Level Spillover Effects of Income Transfers on Health

About This Grant

PROJECT SUMMARY Cancer, mental health disorders, and substance use are leading causes of disease burden, disability, and premature death in the United States, with particularly large impacts among low-income populations. Identifying effective strategies to increase cancer screening and preventive care visits among low-income populations is critical to reducing this burden. Identifying effective strategies to increase cancer screening and preventive care visits among low-income populations is critical to reducing this burden. Because poverty is spatially concentrated in the U.S., income gains accruing to some residents of low-income neighborhoods could shift broader neighborhood economic conditions and generate health “spillover effects” among nonrecipients — through peer diffusion of healthy behaviors and other neighborhood improvements such as parks, transportation, and health clinics resulting from increased local economic activity. Evidence of neighborhood spillover effects on health could deepen understanding of how neighborhood economic conditions drive population health, equipping health systems and public health practitioners with evidence to better anticipate disease burden, prioritize community health investments, and improve health outcomes in low-income communities. We propose the first study to rigorously investigate neighborhood spillover effects of increased neighborhood income on health outcomes. We will study changes in neighborhood income resulting from neighborhood-level earned income tax credit (EITC) distributions, leveraging exogenous changes in distributions over time to rigorously estimate spillover effects. We will analyze individual level electronic health record data from 8.2 million patients in 1,222 primary care practices in all 50 states in the U.S. that is linked to U.S. Census Bureau data and tax returns. This unique dataset includes a large sample of individuals who live with or in proximity to each other, providing sufficient statistical power to estimate neighborhood spillover effects. We will investigate spillover effects stratifying by age to understand effects across the life course. Our Specific Aims are to: 1) estimate spillover effects of neighborhood economic conditions on health; and 2) test the robustness and heterogeneity of neighborhood spillover effects. Health outcomes from 2010-present will be characterized using electronic health records data and include cancer screening, preventive visits, mental health, and substance use disorders. We will estimate effects of state EITC distribution changes using staggered difference-in-differences with an event study framework. We will evaluate the reliability of our spillover effect estimates through multiple robustness checks and assess spillover effect heterogeneity by individual, neighborhood, and EITC characteristics. Our novel approach will serve as a template for future studies that investigate how health interventions targeting neighborhood economic conditions influence health.