On Comprehensively Intermediate Measures of Inequality and Poverty, with an Illustrative Application to Global Data

Abstract The dominant convention in the measurement of inequality and poverty is to employ scale-invariant and replication-invariant measures, that is, measures that are thoroughgoingly relative. This is a routine feature of both the theoretical and applied literature in the area, despite weighty arguments that have been advanced by certain practitioners in favor of centrist measures which avoid the “extreme” values of both income-relative and income-absolute measures. The present paper extends these arguments in favor of measures which are both income-centrist and population centrist. A comprehensively centrist Gini coefficient of inequality is proposed, and likewise a comprehensively centrist class of poverty measures which are counterparts of the well-known Foster-Greer Thorbecke class of relative poverty measures. It is suggested that our diagnosis of the problems of inequality and poverty is likely to be a profoundly variable function of the precise types of inequality and poverty measures we employ in order to assess the magnitudes and trends of the phenomenon.