Assessing the Poverty Surge in Peru (2019-2023): A Spatial Econometric Analysis of Regional Disparities

Accurate territorial diagnostics are paramount for designing effective poverty eradication strategies, yet district-level studies in Peru remain scarce due to critical sampling limitations. The National Household Survey currently covers only 66.9% of Peru’s 1,874 districts, with rural coverage dropping to 50.9%, rendering direct estimators unreliable for local decision-making. This study fills this gap by providing the first longitudinal, spatially disaggregated analysis of district-level poverty spanning the pre- and post-pandemic periods (2019–2023). We implement a Modified Spatial Fay-Herriot model, a Small Area Estimation (SAE) technique that integrates survey microdata with auxiliary administrative sources. By leveraging spatial dependencies to borrow strength across neighboring domains, the model generates reliable estimates for districts with insufficient or null sample sizes. Our findings reveal profound territorial fragmentation. While poverty remains systematically above pre-pandemic levels in 18 of the 25 departments, the spatial analysis identifies a localized urban shock in Metropolitan Lima–where previously stable districts transitioned into high-poverty clusters (25–40%)–and a perverse resilience in the Amazonian frontier. Notably, in Madre de Dios, poverty reduction is paradoxically linked to illegal gold mining expansion, illustrating a trade-off between immediate income and long-term environmental degradation. Conversely, structural poverty persists in Loreto, Cajamarca, Puno, and the central Highlands, where infrastructure deficits impede economic convergence. These results produce the most detailed poverty maps available for post-pandemic Peru, underscoring the inadequacy of one-size-fits-all policies and the need for localized interventions to dismantle structural inequality.