Measuring income inequality under restricted interpersonal comparabilit

The standard approach in empirical analyses of income distributions is to estimate income inequality in a country under the assumption of full interpersonal comparability of income. To be meaningful, this method requires that prices and qualities of goods as well as consumption habits are uniform across individuals in different regions of the country. In this paper, we pursue two alternative approaches to measure inequality under restricted interpersonal comparability of income. First, we estimate regional price indices, transform observed incomes into real incomes in an attempt to incorporate relevant non-income heterogeneity, and then aggregate across individuals. Second, we use the observed income data and account for non-income heterogeneity at the aggregation stage. Results based on a Norwegian register household panel data set indicates that both levels and trends in overall inequality as well as the inequality contributions of various income factors are robust to whether the income inequality analysis is based on the standard approach or the methods developed to cope with comparability problems within a country.