Abstract: Place-based policies and environmental shocks can have heterogeneous distributional effects, yet empirically assessing these consequences is difficult because inequality measures remain spatially coarse. We develop a satellite-based framework to measure neighborhood-level inequality in durable assets and propose a bottom-up aggregation scheme based on the Law of Total Variance to recover inequality for any policy-relevant geography. Using Mexico’s 2020 census data on 35 million households, we demonstrate that our approach recovers inequality across spatial scales with substantial accuracy, improving as predictions aggregate from neighborhoods to municipalities. We illustrate the method’s utility by constructing the first annual municipal panel of inequality (2017–2024). Using synthetic difference-in-differences, we estimate that a border-region tax-and-wage reform (the Northern Border Free Zone) reduced relative inequality by 0.18 standard deviations, while Hurricane Otis increased it by 0.13 standard deviations. We also find that while pre-existing inequality does not limit the benefits of progressive policy, it significantly amplifies vulnerability to adverse shocks.
Risk and Insurance Department