TY - JOUR
AU - Thistle, Paul D.
AB -  I. INTRODUCTION Inequality and welfare comparisons of income distributions are closely related. Welfare comparisons incorporate 'efficiency preference' or preference for higher mean income, as well as inequality aversion. In welfare comparisons, greater inequality may be compensated for by sufficiently higher mean income. To completely order a set of income distributions requires specifying a welfare index.' Sen (1974, 1976) provides an axiomatic basis for a welfare index based on the Gini coefficient. The purpose of this paper is to provide statistical procedures that allow óne to make probabiistically valid statements regarding Sen's index. Much of the data used in empirical analyses of income inequality and welfare are based on samples of the population. Mean incomes, income shares, and inequality and welfare indices calculated from such data are sample statistics, and are therefore subject to sampling variability. Comparisons of the calculated values of such statistics (i.e., point estimates), commonplace in the literature, are hypothesis tests where the probability of type I error is one. Application of sound statistical inference procedures requires knowledge of the sampling distribution of the statistic. In this paper we examine the large sample distribution of Sen's welfare index. Section II briefly discusses the interpretations of Sen's 
TI - PRACTITIONER'S CORNER: An Asymptotically Distribution‐Free Test for Sen's Welfare Index
JF - Oxford Bulletin of Economics & Statistics
DO - 10.1111/j.1468-0084.1990.mp52001008.x
DA - 1990-02-01
UR - https://www.deepdyve.com/lp/wiley/practitioner-s-corner-an-asymptotically-distribution-free-test-for-sen-hArTioN8Pi
SP - 105
VL - 52
IS - 1
DP - DeepDyve
ER -