Finite sample distributions of studentized inequality measures differ substantially from their asymptotic normal distribution in terms of location and skewness. We study these aspects formally by deriving the second-order expansion of the first and third cumulant of the studentized inequality measure. We state distribution-free expressions for the bias and skewness coefficients. In the second part we improve over first-order theory by deriving Edgeworth expansions and normalizing transforms. These normalizing transforms are designed to eliminate the second-order term in the distributional expansion of the studentized transform and converge to the Gaussian limit at rate O(n-1). This leads to improved confidence intervals and applying a subsequent bootstrap leads to a further improvement to order O(n-3/2). We illustrate our procedure' with an application to regional inequality measurement in Côte d'Ivoire.

Mobility indices are popular tools designed to quantify the extent of income changes by aggregating “local” distributional change into a “global” scalar according to some rule. For some mobility measures, this aggregation rule is only implicit in their standard definition. We derive an insightful approximation to the (statistical) aggregation rule for the important class of mobility indices introduced by Shorrocks (Journal of Economic Theory 19 (1978), 376–93) and further generalized by Maasoumi and Zandvakili (Economic Letters 22 (1986), 97–102), which enables us to characterize their normative properties. We also develop methods for estimation and inference. A substantive empirical contribution emerges from the comparison of mobility between the United States and Germany. Our methods reveal why income mobility is higher in Germany than in the United States: Higher German mobility in the bottom of the distribution is combined with an implicitly higher weighting by the mobility index at the bottom.

No abstract is available for this item.

This article is about statistical inference for inequality and poverty measures when income data exhibit contemporaneous dependence across members of the same household. While much empirical research is based on household survey data such as the PSID, standard methods assume that income is an independent and identically distributed random variable. Applying them to contemporaneously dependent data produces biased results, and Monte Carlo experiments reveal that their confidence intervals are too narrow. By contrast, our proposed distribution-free estimators perform well. Copyright Economics Department of the University of Pennsylvania and the Osaka University Institute of Social and Economic Research Association