Income inequality is an important aspect of the economic and social well-being of a nation. The axiom “the rich are getting richer and the poor are getting poorer” is synonymous with rising income and wealth inequality. Both affluent and poor countries experience some degree of income inequality.
Income inequality rose in the United States between the mid-1970s and early-2000s based on several measures. The United States generally exhibited the highest degree of income inequality of the industrialized OECD (Organization for Economic Cooperation and Development) countries, although research by Peter Gottschalk and Timothy Smeeding (2000) did not discern any universal trend. According to opinion polls by Demos, a U.S. public policy organization, American public opinion toward income inequality tends to shift in response to media reports on excessive executive compensation, corporate downsizing, and the state of the U.S. economy.
Opinions are divided on whether society should be concerned about income inequality. Economists Martin Feldstein and Anne Krueger argue that policymakers should focus on reducing poverty, not inequality. This is because an increase in inequality will occur even though all income-receiving units (IRUs) are better off in absolute terms if the incomes of richer IRUs rise by higher proportions than those of their poorer counterparts. An alternative view, articulated by Simon Kuznets (1901–1985) and Kenneth Arrow, winners of the 1971 and 1972 Nobel Prize in Economics, respectively, is that income inequality is an important determinant of aggregate savings and growth, and the perception of unfairness it entails may cause political instability and social conflict.
One way of analyzing income inequality trends is to examine the income shares accruing to some segment of the population (e.g., the poorest and richest 10 percent). A falling (rising) income share of the poorest (richest) group indicates rising inequality.
A summary inequality measure is a single number computed from the incomes of several IRUs (individuals, households). This number summarizes the degree of income inequality, with higher values indicating greater inequality.
The Gini index or Gini coefficient —named after Corrado Gini (1884–1965), the Italian statistician who developed it in 1912—is a popular summary measure. It ranges from zero for perfect equality (all IRUs receive identical incomes) to one for perfect inequality (one IRU receives all the income). For example, the World Bank (2005) reports the 2000 Gini index for the United States and Canada as 0.38 and 0.33, respectively, indicating that the United States had greater inequality. Such international comparisons are strictly valid if they are based on common definitions of IRUs and income (gross income, net income) and similar geographical coverage (national, urban).
Frank Cowell (1995) provides details (including formulas) of the Gini index and other summary measures. Cowell describes the generalized entropy (GE) family, which includes the mean logarithmic deviation (GE(0)), Theil’s index (GE(1)), and half the squared coefficient of variation (GE(2)) as special cases, and Atkinson’s measure. Atkinson’s measure ranges from zero to one and GE from zero to infinity.
There is no single best summary inequality measure. The choice among summary measures is influenced by: (1) computational convenience; (2) satisfaction of desirable properties, including scale independence (multiplying all incomes by a constant should not affect inequality), the Pigou-Dalton transfer principle (an income transfer from a richer to poorer IRU that does not reverse their ranking should reduce inequality), anonymity/symmetry (identities of IRUs are irrelevant), population independence (doubling population size by replicating every IRU should not affect inequality), and decomposability (total inequality should be conveniently broken down by population subgroups or income components); and (3) the portion of the income distribution to emphasize.
The availability of software for computing inequality measures has rendered computational considerations less important. The Gini index, GE, and Atkinson’s measure satisfy scale independence, the transfer principle, anonymity/symmetry, and population independence. The Gini index cannot be conveniently decomposed and is more sensitive to income changes in the middle of the distribution. The sensitivity problem is circumvented by computing a generalized/extended Gini, whose formula, like that of Atkinson’s measure and GE, incorporates an inequality aversion parameter that can be changed to stress different portions of the income distribution. The GE is renowned for its additive decomposability (total inequality in a population that is partitioned according to race, gender, or other characteristics is the sum of within-group inequality and between-group inequality).
The degree of income inequality can also be deduced graphically from a Lorenz curve. When comparing non-crossing Lorenz curves, the one with greater curvature away from the perfect equality line indicates greater inequality. The Gini index is twice the area between the Lorenz curve and the perfect equality line, which may be inaccurate if the Lorenz curve is constructed from data that are grouped into income brackets, as statistical agencies often do.
The Gini index for U.S. household income inequality, reported by the U.S. Census Bureau, increased from 0.397 in 1975 to 0.466 in 2001. Both the GE and Atkinson’s measure also increased during this period, indicating rising inequality.
Two factors complicate the analysis of the observed inequality trends. First, the trend may depend on whether a narrow definition of income (market income) or a broader definition that takes into account taxes, social transfers (e.g., child benefits), and noncash benefits (e.g., food stamps) is considered. Rising social transfers may offset an increase in earnings inequality. Second, since summary measures are based on household surveys, researchers should test whether observed inequality changes are statistically significant by computing an appropriate standard error. Until the mid-1990s, many practitioners eschewed this issue because of the complexities of most standard error formulas. Advances in computing have facilitated significance tests for inequality changes using the bootstrap standard error, as Martin Biewen (2002) has demonstrated.
Economic historian Peter Lindert (2000) has summarized the vast literature explaining the rise in U.S. income inequality. The rise is ascribed to a complex mix of economic and social/demographic factors. Since earnings constitute the largest component of income for most IRUs, explanations of rising inequality in the United States have focused mainly on earnings inequality. They include the weakening of unions, technological changes requiring highly skilled workers, outsourcing of jobs, and immigration. Other explanations include import competition from low-cost countries, government tax and income transfer policies, and the rise in single-parent families. Similar factors have been used to explain inequality trends in other industrialized OECD countries, with variations in the relative importance of the factors.
Does economic growth lead to rising inequality? In 1955 Kuznets postulated the existence of an inverted U-curve relationship between economic growth and income inequality by tracking the historical experiences of England, Germany, and the United States. This U-curve hypothesis contends that the intersectoral shifts associated with the early stages of economic growth exacerbate inequality (the rising portion of the inverted U-curve). At some threshold level, inequality peaks and then falls (the falling portion of the inverted U-curve).
The vast literature on the U-curve hypothesis, reviewed by economist Ravi Kanbur (2000), reveals mixed empirical results. Furthermore, government policy is important in influencing the direction of inequality. Regarding the reverse causation from income inequality to economic growth, empirical evidence by Klaus Deininger and Lyn Squire (1998) reveals that inequality in land distribution is a more important determinant of future growth than income inequality.
It is now widely recognized (e.g., World Bank 2005) that the root cause of persistent earnings and income inequality is unequal opportunities. For example, sustained discrimination in employment or education against a particular racial or ethnic minority group could result in persistently low earnings for the group’s members, resulting in an inequality trap.
The link between ethnicity and income inequality is convoluted. Economists William Darity and Ashwini Deshpande (2000) articulate how overall inequality could drive interethnic inequality and vice versa. Shelly Lundberg and Richard Startz (1998) explain how community influence and loyalty could shape interethnic inequality even without discrimination. Some have argued that ethnically homogeneous societies should have less inequality since assimilation by ethnic groups is easier and support for income redistribution is more likely. Fractionalization indexes, which measure the probabilities that two randomly selected individuals from a population belong to different ethnic, linguistic, or religious groups, have been used in empirical investigations of the ethnic-diversity income inequality nexus, with mixed results.
The causality between wealth inequality and income inequality may potentially run two ways. On the one hand, those with greater wealth accumulation are more likely to invest and generate higher incomes. On the other hand, those with higher incomes have higher potential to accumulate more wealth.
Are the children of poor parents destined to remain poor? This is the issue of intergenerational mobility, the relationship between a person’s socioeconomic status and that of his or her parents. The rich are capable of providing their offspring with a better education, increasing their chances of earning higher incomes. Also, the offspring of the rich are likely to inherit greater wealth, which aids the wealth accumulation process for the next generation. Thomas Piketty (2000) has surveyed theoretical and empirical literature on intergenerational mobility. Economists and sociologists measure the degree of intergenerational mobility by computing an intergenerational elasticity in income and wealth. Unfortunately, intergenerational elasticity estimates are sensitive to the methodology employed. Irrespective of the exact magnitude, there is no doubt that the level of opportunities allowed by society determines intergenerational mobility.
SEE ALSO Gini Coefficient; Income; Income Distribution; Inequality, Gender; Inequality, Political; Inequality, Racial; Inequality, Wealth; Interest Rates; Poverty, Indices of; Profits; Rent; Wages
Biewen, Martin. 2002. Bootstrap Inference for Inequality, Mobility, and Poverty Measurement. Journal of Econometrics 108 (2): 317–342.
Cowell, Frank A. 1995. Measuring Inequality. 2nd ed. Wheatsheaf, U.K.: Prentice Hall.
Darity, William, Jr., and Deshpande Ashwini. 2000. Tracing the Divide: Intergroup Disparity Across Countries. Eastern Economic Journal 26 (1): 75–85.
Deininger, Klaus, and Lyn Squire. 1998. New Ways of Looking at Old Issues: Inequality and Growth. Journal of Development Economics 57 (2): 257–285.
Gottschalk, Peter, and Timothy M. Smeeding. 2000. Empirical Evidence on Income Inequality in Industrialized Countries. In Handbook of Income Distribution, eds. Anthony Atkinson and François Bourguignon, 261–307. Amsterdam: Elsevier.
Kanbur, Ravi. 2000. Income Distribution and Development. In Handbook of Income Distribution, eds. Anthony Atkinson and François Bourguignon, 791–841. Amsterdam: Elsevier.
Lindert, Peter H. 2000. Three Centuries of Inequality in Britain and America. In Handbook of Income Distribution, eds. Anthony Atkinson and Francois Bourguignon, 167–216. Amsterdam: Elsevier.
Lundberg, Shelly, and Richard Startz. 1998. On the Persistence of Racial Inequality. Journal of Labor Economics 16 (2): 292–323.
Piketty, Thomas. 2000. Theories of Persistent Inequality and Intergenerational Mobility. In Handbook of Income Distribution, eds. Anthony Atkinson and François Bourguignon, 429–476. Amsterdam: Elsevier.