On measuring deprivation adjusted for group disparities

Abstract

This paper is concerned to present a procedure for `adjusting' a real-valued index of deprivation in such a way that the resulting measure is a summary statistic of both the average level of deprivation and the extent of inequality which obtains in its distribution. Copyright Springer-Verlag Berlin Heidelberg 2002

Using the Mean of Squared Deprivation Gaps to Measure Undernutrition and Related Socioeconomic Inequalities

“…While relational inequalities are embedded in the social structure in the form of relations of 'superordination' and 'subordination', distributional inequalities allude to interpersonal differences in wealth or income or outcome indicators such as health and educational status (Chakraborty, 2001). Whichever space we deal with, income or important human capabilities, we find a host of studies that dwell on the issue of inequality across various population subgroups within the same geographical area (Anand and Sen, 1995;Subramanian and Majumdar, 2002;Jayaraj and Subramanian, 1999;Majumdar and Subramanian, 2001). Kakwani et al (1997) have discussed how health inequalities can be studied using grouped data, when groups are formed on the basis of socioeconomic status, social class, levels of educational attainment or income.…”

Reckoning Inter‐group Poverty Differentials in the Measurement of Aggregate Poverty

“…Poverty indices which satisfy these restricted versions of symmetry and transfer also possess the virtue of a certain sort of ‘flexibility’, in terms of which, for example, an interpersonal transfer across members of different subgroups which reduces inter‐group disparity in the distribution of poverty may or may not reduce aggregate poverty, depending on how regressive the transfer is; a more rather than less regressive transfer could be partial to the transfer axiom at the expense of the subgroup sensitivity axiom and the other way around with a less, rather than more, regressive transfer. Examples of such poverty indices are available in a poverty related version of the Anand and Sen (1995) ‘gender‐adjusted human development index’, and in the measure advanced by Jayaraj and Subramanian (1999), and discussed in Subramanian and Majumdar (2002). These poverty measures are presented below.…”

“…This is the basic idea underlying the construction of ‘group inequality sensitive’ indices of aggregate poverty. A specific version of the Anand and Sen index (suitably adapted to the present context) is the index P A ; and the poverty measure advanced in Subramanian and Majumdar (2002) is the index P B , which – given z , x and g – can be written as…”

“…Let be the squared coefficient of variation in the inter‐group distribution of the π j , andbe an alternative measure of inequality for this distribution (introduced and discussed in Subramanian and Majumdar 2002). Subramanian and Majumdar (2002; p. 268) point out that ‘[b]ecause of the rank‐order weighting formula employed in [the expression for I *], one can discern that [ I *] bears a close family resemblance to the familiar Gini coefficient of inequality . .…”

“…Related approaches to social evaluation have been examined by Anand and Sen (1995) in the context of deriving a 'gender-adjusted human development index'; variations on this theme have been explored by Jayaraj and Subramanian (1999), Majumdar and Subramanian (2001), Subramanian and Majumdar (2002) and Subramanian (2006). The Subramanian and Majumdar (2002) paper provides an axiomatic justification for a specific 'group inequality-adjusted' aggregate index of deprivation. A similar index is derived in the present paper and a diagrammatic link to this index, in the form of a graph called the 'group poverty profile', is advanced.…”