The geography of misallocation: Spatial patterns and local distortions in Italian manufacturing

Productivity and resource allocation could be influenced by the occurrence of spillover effects due to geographical structure. This paper examines the spatial dimension of misallocation and productivity across Italian provinces, using firm-level microdata for manufacturing sector. Methodologically, we build province-level measures of misallocation and total factor productivity, and then we apply a spatial filtering approach to decompose them into two parts: a spatial component, capturing inter-regional spillovers and geographic patterns, and an idiosyncratic component, reflecting province-specific characteristics. By disentangling these two dimensions, the paper provides a clearer view of how geography and local structural conditions jointly shape the distribution of productive resources. This distinction is crucial to identify whether regional productivity disparities are primarily due to local peculiarities or spatially correlated distortions. We find significant spatial autocorrelation in both productivity and misallocation measures, where the geographical component is predominant. Eliminating distortions would raise output by over 20 percent, with roughly two-third of these gains associated with the spatial component.