In this paper we conceptualize inequality in a population as having two basic types—individual inequality and structural inequality—with the former generated by individual variations whereas the latter responding to the social structure in a society represented by social groups. We propose several structural inequality measures, and a structural Gini and a structural Theil measure and its standard error estimator are particularly useful implementations. The proposed measures are examined in three applications. The first is a simulation study, where a range of two-class situations of varying degrees of inequality are analyzed. Two empirical data analyses using the 1991 British Household Panel Survey and the 2005 Panel Study of Income Dynamics from the US are conducted. The applications taken together suggest some useful properties of the new measures that the conventional Gini and Theil measures including Theil decompositions do not possess. Used together with the conventional measures, the structural inequality measures provide a complementary picture of the form of inequality in a society; standing alone, these measures give a single index of the degree of stratification in the form of structural inequality sometimes ignored by conventional measures.
Presented by:
Tim Liao (University of Illinois)
Date & time:
June 15, 2009 3:00 pm - June 15, 2009 4:30 pm
External seminars home