University of Texas Inequality Project
University of Texas Inequality Project
A new body of empirical literature on economic inequality emerged in the early- and mid-1990s that sought to establish cross-country and timewise analysis of the relationship between economic growth, income levels, and income distribution. At a fundamental level, this literature was concerned with exploring questions posed by Simon Kuznets in the mid-1950s about the relationship between income and inequality levels. His hypothesis was that if one were to plot inequality levels against income levels over the long run, this relationship could be described by an inverted U curve. As countries grew from an agricultural to an industrial economy, inequality would go up initially, with few people working in the high-productivity and high-income industrial sector. As more and more people transitioned from agriculture to industry, the increase in inequality would decelerate, stop, and eventually reverse when the majority of the population worked in the industrial sector.
The new empirical studies of the early 1990s related to the Kuznets hypothesis but also to newer conjectures that predicted either positive or negative relationships between inequality and growth. Some theories, such as those of Alberto Alesina and Dani Rodrik, were based on political economy arguments, while others, such as the work of Abhijit Banerjee and Andrew Newman, were based on the effect of inequality in impeding access to credit.
This work generated a large demand for data on inequality that were internationally comparable and that spanned as long a time series as possible. While such data had long been available for income and growth, there was a paucity of global data on income distribution. Klaus Deininger and Lyn Squire, then at the World Bank, were pioneers in setting up and making available a compilation of inequality measures that were broadly comparable across countries and over time. Their dataset, based on household surveys, reported Gini coefficients—a summary measure of income inequality that ranges from 0, for perfect equality, to 100, when all income goes to a single individual, which is the most unequal of all possible distributions of income. Other similar, more comprehensive efforts followed, including that by the United Nations University’s World Institute for Development Economics Research (WIDER) center.
As Andrea Bradonlini and Anthony Atkinson noted in the late 1990s, these comprehensive datasets had promise but also pitfalls. Some of the main shortcomings involved the lack of long and dense time series of inequality for many countries, especially developing countries. Cross-country comparability was also an issue. These factors made empirical work using these datasets difficult. Under the leadership of James K. Galbraith, the University of Texas Inequality Project (UTIP) was created initially as an effort to offer other sources of measures of inequality that addressed some of the shortcomings of these datasets.
UTIP’s initial proposal was to derive measures of within-country inequality using internationally comparable pay data, collected for industrial statistics released by the United Nations Industrial Development Organization (UNIDO). An issue with this dataset was that pay data were reported only at the industry level, and thus the within-industry distribution was unknown. Thus the use of the Theil index—the most widely used measure of a family of inequality measures that are perfectly decomposable into a within-group and between-group distribution—emerged as the most natural choice. For between-industry pay, the Theil index generated long and dense measures of inequality that were broadly comparable across countries.
In addition to releasing this global dataset, the work of UTIP involved exploring the methodological issues associated with the chosen measure of inequality. Using this dataset as well as data from national sources, Galbraith led innovative work on the interactions between inequality and several economic and political issues, ranging from conflict to financial crises. In fact, the use of the Theil index was expanded beyond global measures of interindustry inequality pay into other fields where issues of aggregation made the Theil index the natural measure to use. For example, when income or pay data are available at several levels of geographic aggregation (as in the United States, for example, at the national, state, and county levels), the Theil index enables the construction of decomposable aggregate measures into the contributions of the lower levels of aggregation. Insights into the dynamics of the contributions to aggregate inequality from lower levels of aggregation can then be established. Galbraith and colleagues at UTIP did pioneering work in using the Theil index to “aggregate upward.” For example, while inequality in the United States is often compared with that of other countries, it has rarely been compared with inequality in, say, Europe or European Union countries. By aggregating national measures of inequality upward, it is possible to construct a European-wide measure of inequality that may be more analytically appropriate to use in comparisons with the United States than that of individual countries.
More recently the UTIP-UNIDO measures of manufacturing-pay inequality were used, with other information, to estimate measures of household income inequality. This was accomplished by taking advantage of the systematic relationship between the UTIP-UNIDO estimates and those of Deininger and Squire. The residuals from this exercise provided a map to problematic estimates in the Deininger and Squire data, and the estimated coefficients enabled the construction of a new panel dataset of estimated household income inequality.
SEE ALSO Deininger and Squire World Bank Inequality Database; Income Distribution; Inequality, Income; Kuznets Hypothesis; Theil Index
Alesina, Alberto, and Dani Rodrik. 1994. Distributive Politics and Economic Growth. Quarterly Journal of Economics 109 (2): 465–490.
Atkinson, Anthony, and Andrea Brandolini. 2001. Promise and Pitfalls in the Use of “Secondary” Data-Sets: Income Inequality in OECD Countries as a Case Study. Journal of Economic Literature 39: 771–799.
Banerjee, Abhijit, and Andrew F. Newman. 1993. Occupational Choice and the Process of Development. Journal of Political Economy 101: 274–298.
Deininger, Klaus, and Lyn Squire. 1996. New Ways of Looking at Old Issues: Inequality and Growth. World Bank. Unpublished paper.
Galbraith, James K., and Pedro Conceição. 1998. Constructing Long and Dense Time-Series of Inequality Using the Theil Index. UTIP Working Paper No. 1. http://utip.gov.utexas.edu/papers/utip_01.pdf
Galbraith, James K., and Hyunsub Kum. 2005. Estimating the Inequality of Household Incomes: A Statistical Approach to the Creation of a Dense and Consistent Global Data Set. Review of Income and Wealth 51 (1): 115–143.
Kuznets, Simon. 1955. Economic Growth and Income Inequality. American Economic Review 45 (1): 1–28.
United Nations University. 2005. WIDER World Income Inequality Database V 2.0b. http://www.wider.unu.edu/wiid/wiid.htm.