Measuring growth and convergence at the mesoscale

Abstract

Global inequality has shifted inward, with rising dispersion increasingly occurring within countries rather than between them. Using 8,790 newly harmonised Functional Urban Areas (FUAs) - micro-founded labour-market regions encompassing 3.9 billion people and representing approximately 80% of global GDP - we show that national aggregates systematically, and increasingly, misrepresent the dynamics of growth, convergence, and structural change. Holding the underlying nighttime-lights GDP raster (1992-2019) fixed while varying the unit of aggregation (ADM0-ADM3, FUA), we isolate the contribution of the Modifiable Areal Unit Problem directly in the growth literature. Three results follow. First, where inequality is located is unit-dependent: FUAs recover the most stable income-inequality relationship, corroborated by independent wealth data. Second, estimated beta-convergence is scale-sensitive, and FUAs exhibit a discrete jump in convergence strength relative to administrative units of comparable population. Third, we find no poverty trap at the urban scale: expected growth remains positive throughout the income distribution, while the middle-income acceleration flattens over time. The cross-country convergence debate has been conducted on national aggregates, yet nations are political containers rather than economic units, and measured relationships, including the convergence coefficient, depend on the spatial unit of analysis.