This paper discusses inequality orderings based explicitly on closing up of income gaps, demonstrating the links between these and other orderings, the classes of functions preserving the orderings and applications showing their usefulness in comparison of economic policies. JEL: D31, D63 Keywords: Inequality, income distribution * I am grateful for helpful comments from Tim Besley, Chris Gilbert, Terence Gorman, Chris Harris, Peter Lambert, James Mirrlees, Stephen Nickell, Hyun Shin, seminar participants and an anonymous referee. The paper draws in large part upon my doctoral thesis, for the funding of which I am grateful to the Economic and Social Research Council.† Address: Ian Preston, Department of Economics, University College London, Gower Street, London WC1E 6BT, UK. Email: i.preston@ucl.ac.uk 1
Executive SummaryIt is a truism to say that inequality is about gaps between incomes and that reducing inequality is about closing these gaps up.Common means of comparison between income distributions all use criteria which do show inequality as falling when gaps close. However explicitly asking whether the gaps reduce throughout the whole distribution in concertina-like fashion is a rare criterion to apply. This paper seeks to investigate the related orderings. The most common criteria for inequality comparison are those based on Lorenz curves, made plausible most persuasively as indicators of inequality by their link to progressive transfers of income. Progressive transfers are often seen, since the arguments of Pigou and Dalton, as uncontentiously inequality reducing but this view could be challenged if there are more than two people. A transfer from the top to the middle of the income distribution reduces inequality between the top and the middle but increases it between the middle and the bottom. Regarding inequality as having fallen overall involves giving priority to the former effect -the effect on the gap between incomes of those involved directly in the transfer -for which there may be good reason, but it is not obvious that it would not be sensible to say inequality simply could not be compared. A minor function of the current paper is to bring some overlooked but highly germane mathematical literature to the attention of inequality theorists. The major function, though, is to tell a rounded story about the ratio and difference dominance concepts, and associated orderings and welfare properties.