Beyond the Farm: How Employment Shifts and Infrastructure Shape Regional Welfare in Uzbekistan

Abstract

As Uzbekistan undergoes rapid economic transformation, this study investigates how the structural shift toward a diversified, non-farm labour market alleviates poverty across its 14 administrative regions. We quantify how this transition is conditioned by localised thresholds in physical infrastructure, digital connectivity, and human capital. Analysing a 2010–2024 macro-panel dataset, our empirical strategy employs a rigorous general-to-specific spatial econometric framework. Following exploratory spatial data analysis that confirms significant regional clustering, we primarily utilise a Two-Way Fixed Effects Spatial Durbin Model (SDM) to robustly handle cross-sectional dependence and explicitly quantify inter-regional spillovers, while deploying a Spatial Autoregressive (SAR) model to evaluate core-periphery heterogeneity. While initial estimates suggest non-farm employment (NFE) directly reduces poverty, robust spatial models reveal that this structural shift is highly conditional and often spatially conditioned. Crucially, a core-periphery heterogeneity analysis shows that expanding the NFE share yields significant poverty reduction only within Urban Core regions. In contrast, structural labour shifts in the Rural Periphery fail to reduce local poverty without a foundational baseline; instead, these isolated regions rely heavily on regional capital investment for inter-regional welfare spillovers and digital infrastructure expansion for local economic inclusion. Consequently, to sustainably bridge the spatial welfare divide, policymakers must abandon uniform interventions in favour of spatially tailored strategies, transitioning from basic digital and transport capacity-building in peripheral regions to advanced economic absorption strategies in developed hubs.