Scales of Inequality: Applying a Two-Stage Theil Decomposition Approach to Examine the Changing Spatial Dimensions of Regional Inequality in the US, 1970–2020

A substantial body of research has documented the shift from convergence to divergence in inter-regional income disparities and identified the key forces driving this transformation, including educational sorting, technological change and globalization. Yet, despite growing recognition that spatial patterns of income inequality are sensitive to the choice of geographical scale, relatively few studies have explored scalar dynamics in the U.S. In this paper, we apply a two-stage nested Theil decomposition method to examine the role played by the choice of geographical scale in describing patterns of inter-regional income inequalities over the 1970 to 2020 period. We use three geographical scales in our analysis: quasi-states, commuting zones, and counties. We find that inter-regional income inequalities increase over time across all three scales, but the speed of this increase varies significantly across the different geographical scales. In 1970, inter-regional income inequalities between states explained most of the overall levels of inter-regional income inequalities. Over the ensuing decades, however, inter-regional income inequalities within commuting zones subsequently rise more rapidly (than inequalities between states or commuting zones) and by 2020 explain the largest portion of overall inter-regional inequalities. This suggests that smaller-scale divergence, especially inequalities within local labor market areas, is a key driver of the recent rise in inter-regional inequality. By highlighting the growing importance of intra-metropolitan or intra–labor market inequalities, this study contributes to theoretical debates on spatial divergence and urban sorting, emphasizing that the mechanisms driving regional inequality must be understood not only across regions but also within them.