A large literature in economics and public health studies the relationship between socioeconomic status (SES) and health, often relying on area-based measures when individual data are unavailable. Using comprehensive Dutch administrative data, we examine how data aggregation shapes estimated income-health gradients through ecological and atomistic biases. Individual and area income capture distinct but correlated channels, such as personal resources and local conditions. Spatial aggregation reduces variation and attenuates nonlinearities, leading area-level estimates to overstate gradients and reduce robustness to controls. Interaction analyses reveal unequal exposure: Health of low-income households appears substantially more sensitive to neighborhood conditions than that of high-income households.
Parker et al. (Fri,) studied this question.