NCPA - National Center for Policy Analysis

Distorting the Data to Show Income Inequality

November 17, 2014

French economist Thomas Piketty's book "Capital in the Twenty-First Century" sparked a nationwide discussion of income inequality, as he offered readers a plethora of graphs and statistics purporting to show that America's rich have gotten exceptionally richer over the last 40 years, while the rest of America has not seen similar gains. But is that really what has happened? Former Senator Phil Gramm (R-Texas) and Michael Solon of US Policy Metrics say no, arguing that the inequality debate is full of misleading and distorting data that masks the real gains made by Americans of all income levels.

In 2003, Piketty and economics professor Emmanuel Saez of the University of California, Berkeley, authored a study on income inequality in the United States. The study made headlines, depicting growing inequality in the United States since 1979 in which the bottom quintile of households saw a 33 percent decline in income from 1979 to 2007, the second quintile saw a mere 0.7 percent rise in income and the middle class saw a 2.2 percent income increase. But those numbers, say Gramm and Solon, are faulty:

  • Piketty and Saez's study only looked at pretax income, failing to take into account taxes, nonmonetary compensation (such as employer-provided health insurance), Social Security, benefits from Medicare and Medicaid, or means-tested welfare programs. It also did not count IRAs or 401(k)s unless the money had been taken out in retirement.
  • The study only looked at individual income, not household incomes, despite that many families have multiple earners.
  • By leaving out Social Security, Medicare and Medicaid payments, Piketty and Saez conclude that many of American's senior citizens are poor. In reality, say Gramm and Solon, they have incomes above the national average.
  • Similarly, failing to account for nonmonetary benefits, 401(k)s and IRAs means that the Piketty-Saez data ignores large portions of the wealth of Americans in the middle income brackets.

A study recently published in the Southern Economic Journal took those missing figures into account. When accounting for taxes, transfer payments and other benefits, instead of the 31 percent decrease in income for the bottom quintile that Piketty and Saez found, the bottom quintile actually saw a 31 percent increase in income. Similarly, the second quintile saw a 32 percent increase in income -- far higher than Piketty and Saez's 0.7 percent increase -- and the middle quintile saw a 37 percent increase, a far cry from the Piketty-Saez increase of just 2.2 percent.

The distortion does not stop there. Gramm and Solon point out that federal tax reform in 1986 lowered the personal income tax rate down to 28 percent, while the corporate income tax rate was set at 34 percent. This is significant, because businesses that are not registered as C-corporations pay the individual income tax, rather than the corporate tax. In response, many businesses restructured into subchapter S corporations, LLCs and the like, in order to have their income taxed at the individual level. Remarkably, a full third of the growth in income by the top 1 percent from 1979 to 2012 is due to businesses restructuring in this way.

The way that data is used can drastically change the inequality numbers. According to Gramm and Solon, in 1992, the Census Bureau changed the way that it collects information from high-income Americans when conducting population surveys. That change in reporting, they write, is responsible for 30 percent of the rise in inequality from 1979 to the present.

For more on income inequality, see Jeff Lerner's post on wealth taxes at the NCPA Retirement and Taxes blog. Lerner contends there is little sense in attacking income gaps, noting that gains by those at the top do nothing to reduce the wellbeing of those at the bottom. 

Source: Phil Gramm and Michael Solon, "How to Distort Income Inequality," Wall Street Journal, November 12, 2014. 

 

Browse more articles on Tax and Spending Issues