Sorry Krugman, Piketty and Stiglitz: Income inequality for individual Americans has been flat for more than 50 years
Carpe Diem
In the comment section of two of my recent posts on income inequality here and here, Ironman of the Political Calculations Blog sent along some links to his excellent three-part analysis of income inequality in the US ( here, here and here) using the annual Gini index of income inequality for: a) individual Americans, b) US households and c) US families from 1947-2012. It’s one of the best and most comprehensive analyses of income inequality trends in the US over a long 66-year period, and deserves much greater attention. Here’s one of the main and most important conclusions of Ironman’s analysis of income inequality in the US (taken from two of his posts).
In the chart above, we see that a steady increase in the Gini coefficient, the most common measure of income inequality in the U.S., begins to take place for both households (blue line) and families (red line) in 1970 after bottoming in the late 1960s. Meanwhile, we see that income inequality among individual American income earners has been essentially flat during for more than 50 years since 1961 (green line).
That observation is significant because if income inequality in the United States was really rising as a result of economic factors that concentrate an increasing amount of income into progressively fewer hands, we would not observe this outcome because income payments are made to individuals, not to households and not to families. That basic reality means that a rising level of economically-driven income inequality would be most prominently evident among the distribution of income for individuals as measured by the Gini coefficient if it were actually taking place – just like it did in the post-World War 2 recovery years from 1947 through 1960.
We do however observe such a rising trend in measured income inequality in the Gini coefficients calculated for U.S. households and for U.S. families. Interestingly, we don’t see much of that change occurring when economic factors were actually driving up the level of income inequality among individual American income earners in the years from 1947 through 1960. Instead, we see that the overall trend for household and family inequality was basically flat during this time, which then continued through the 1960s. It’s not until 1970 that we find that a rising trend in the amount of income inequality for U.S. households and families begin to take hold.
The only way that can happen is if the composition of U.S. households changed so that they consisted of greater numbers of lower income earning households and families. In this case, the change was driven by an increasing number of single person households, and specifically by an increasing number of single person households consisting of women over the Age of 65.
Bottom Line: The combination of a flat Gini coefficient index for individual income inequality for more than 50 years along with rising Gini coefficients for US households and families means that social, rather than economic factors, are responsible for the most frequently reported rise in income inequality for households and families.
MP: This is a very important finding that: a) individual income inequality has been flat for more than 50 years, b) it’s only household and family inequality that have risen, and c) the rise in household and family income inequality can be explained by demographic factors, rather than economic reasons. In other words, there has been no increasing concentration of income for individual Americans during the last 50 years as is frequently reported by Krugman, Piketty, Stiglitz, and the media. As I have mentioned before, the “problem” of rising income inequality might be what Mencken called an “imaginary hobgoblin,” which is certainly keeping the populace and politicians alarmed, and looking for solutions like higher taxes to address what is basically a non-problem.
Related: See Don Boudreaux’s post today Household Inequality Is Not Individual Inequality.
aei.org |
