Two Statistical Issues in the Study of Income Inequality Open Access
Downloadable ContentDownload PDF
In this paper, several indices designed to measure income inequality used to analyze the data from several countries to evaluate and compare the trend towards increased inequality that has occurred in the last forty years. In addition to the Gini and Theil index, their median-modified versions proposed by Gastwirth (2014) and two ratios of the share of income going to the upper and lower portions of the income distribution (the 8020 or ratio of the top 20% to bottom 20% and the Palma measure) are applied to the data.Because government agencies report income data in grouped form, the second chapter compared the estimates of the indexes of inequality using the split histogram and a simple linear interpolation method. It turned out that estimates of the Lorenz curve and several indices of income inequality derived from it obtained from both methods were very similar. Thus, researchers can rely on either one.The third chapter begins an exploration into some recent indices proposed by Gastwirth (2016) and Zenga (2016). Our preliminary results indicate that the Zenga index of inequality applied to the U.S. data grew more slowly during the study period than the other indices.