IEI Insight: Common measure of income inequality seriously flawed

This IEI Insight is provided by Madelyn McGlynn, a Gail Werner-Robertson Fellow and author of a forthcoming paper on income inequality and the Gini Index.

Income inequality in America is growing quickly. Between the Occupy Wall Street phenomenon and presidential candidates debating solutions, this fact has started to creep into the popular consciousness and has sparked much concern. Income inequality is growing and this can be demonstrated mathematically. But what does that analysis really tell us?

According to Federal Reserve Chairman Janet Yellen, since 1973 the top ten percent of American incomes increased by about 30 percent. The bottom 50 percent of workers’ real income only rose by about five percent. This difference is significant, changes the dynamics of the American economy, and this inequality is not getting any better. Many studies have attempted to determine why this is happening. Researchers tend to conclude that income inequality is exacerbated by gaps in education, an aging labor force, and the presence of concentrated populations.

Before solutions or the morality of current income inequality can meaningfully be debated, we need to step back and evaluate how the numbers that are used to justify these debates are derived. Income inequality is most commonly measured usually the Gini Index that was developed in 1912. It measures discrepancies in income as a percentage of a whole, where 1.0 equals complete income inequality. A score of 1.0 would mean that a single individual has all of the income and everyone else has none and a score of 0.0 means income is distributed equally between all people.

The data collection for income in America is actually a reflection of taxable income. Thus income is measured differently based on the industry in which individuals are employed. Farm income is particularly hard to generalize because of the extent to which people involved in this industry take advantage of opportunities for tax credits and subsidies, making it a good example of the differences how income is understood. Farmers also often invest in large capital assets that amortize over time, creating an illusion of lost revenue. Often the revenue farmers make has to be reinvested into their land and equipment in order to keep their businesses going. These, however, are considered business expenses and so they are not included in the measurement of the farmers’ personal income. This might be a good reflection of farm work, however, it is not very helpful for comparison with income for other industries that may be less capital-intensive and may be regulated very differently.

The Gini Index is an inherently comparative tool because of the use of an index. When evaluating the Gini Coefficient in a single area, it must be compared to coefficients in other areas or over time in order to produce meaningful information. This is a problem because there is no mechanism to account for inconsistent data collection methods — as with the aberrations in reported farm income — within the equation.

Building off of earlier work with Nebraska market data, my research has now been extended to include the five states with highest farm income as a percentage of total income. These “farm states” in order of highest to lowest farm income concentration are: South Dakota, Nebraska, Iowa, North Dakota, and Idaho. Each of these states was assessed and compared on a county level. The counties of each state were assessed based on what people usually point to as the main causes of inequality: education (measured by the percent of the population with a high school degree or higher), median age, and population density, and our variable: percentage of farm income. Additionally, regression analysis was conducted to determine statistical relationships between these factors and the Gini Coefficient.

The analysis shows that in Nebraska and South Dakota, as expected, higher farm income is associated with greater income inequality. Contrary to expectations for the states Iowa and Idaho, farm income is correlated to lower levels of income inequality measured by the Gini Index. In North Dakota, the final state, the analysis indicated no relationship between farm income and income inequality.

The presence of a relationship between farm income and the Gini Index differed in significance, something that should produce further investigations beyond my current study. However, the present findings of my study seem to strongly undermine the reliability of the Gini Index for measuring income inequality for states heavily dependent upon farm income. More generally, this basic unreliability of the Gini Index calls into question its use as an analytical tool by social scientists and policymakers.