Mobility and Mean Reversion in the Dynamics of Regional Inequality

Abstract

The literature on regional growth convergence and economic disparities has tended to confound four interwoven measurement phenomena: 1) mean reversion (so-called beta convergence)—richer regions move towards the average from above and poorer regions from below; 2) diminishing inequality (so called sigma convergence)—the horizontal or spatial distribution of income becomes more equal; 3) mobility—the rank of a region in the overall distribution of income changes either upwards or downwards; and 4) leveling—the richer regions become poorer (leveling-down) or the poorer regions become richer (leveling-up). We use a new statistical methodology that treats these four phenomena on an integrated basis. The methodology is applied to Israeli regional earnings. We show that regional earnings are Gini divergent, but after adjusting earnings for regional cost-of-living differential, this picture is reversed. In the absence of genuine cost-of-living data, a simple and practical method is proposed, whereby regional house price data are used to proxy regional cost-of-living differentials.