On October 25, the Congressional Budget Office (CBO) published an analysis of changes in the distribution of household income between 1979 and 2007. CBO argues that the 62 percent gain in average household income over this 28-year period mostly went to households in the top 20 percent of the income distribution, where average income grew by 65 percent. Average household income in the top 1 percent of the distribution grew by 275 percent after taxes. Those households in the bottom 20 percent of all households saw income gains of 18 percent across this period.

These are dramatic findings, but they are not above criticism. The CBO has made progress in addressing some past issues, and, for example, now does a much better job of defining total compensation. But it still leaves other issues untouched, from the changing demographic composition of the quintiles to the imputed income value of Medicaid and the treatment of elderly income.

This year’s report—like last year’s—also skates around a clear statement of exactly why CBO is interested in the changing distribution of income at all: Is the CBO describing changes in the distribution of income with a vision of a “correct” distribution in mind, or is it merely giving us the latest statistics without any notion of what we should do with them? Finally, the CBO gives us an inadequate discussion of how the distribution of income has been affected by fundamental changes to the structure of our economy and the U.S.’s evolving role in global economic activity.

What is CBO’s point? While it might seem odd to ask this question in today’s highly charged debate over the distribution of income, CBO attempts to stay above the fray. Unfortunately, its analysis clearly implies that growing unevenness in the distribution is a bad thing. It lays an important foundation for a possible and unfortunate public policy response, and that hammer could beat away at the economy before finding an appropriate nail. If the distributional changes are due to demographic shifts (older workers, more women working, more immigrants), how does public policy alter this momentum? If distributional changes reflect the enormous alteration in the structure of the U.S. economy from manufacturing to services, is it CBO’s intent to lay a foundation for reversing that trend? Since the median-income worker in the bottom quintile is so much better off today than in 1979, is the purpose of this report to argue that the federal government should act against the economic forces that have produced those income benefits? In short, the whole report begs the question of why it was written.

Only slight changes in labor income. If one looks only at labor income (what we earn through wages and salaries), the distribution of income is little changed between 1979 and 2007. Figure 6 of the report shows that labor income is slightly less equal, but CBO leaves us guessing if this change is due to the baby boomers’ being at their peak earning ages or to other systematic factors, like the growth of females in the labor force or the expansion of low-income workers, partially due to the explosion in immigration across this period. The CBO report unfortunately fails to narrow the factors affecting this distributional change, which I would assume are largely demographic: After all, the largest single generation in U.S. history could be affecting this feature of American life as much as it has nearly everything else, and immigrants now account for 50 percent of U.S. labor force growth over this 28-year period.

Larger changes in capital and business income. However, significant distributional differences between 1979 and 2007 do exist for income from businesses and from the ownership of capital. There can be no doubt that the main force producing these differences in income is the structural change in the U.S. economy that occurred during this period—primarily the growth in financial, legal, health, and technical services. CBO touches on this when it discusses the change in income between 1979 and 2007 for executives and professional workers. Oddly, it leaves untouched the effects on the income distribution from the explosion in federal government spending and the wealth that spending has created among government contractors in these sectors.

In addition to its uneven treatment of different types of income, the CBO report fails to address several issues with low-income data:

  • The report counts the value of Medicaid and Medicare at their “fungible value,” which is far less than their actual cost to the taxpayer. The fungible method assigns a low or zero value to Medicaid and Medicare benefits received by poor people and a much higher value to the same coverage given to a higher-income person. So medical coverage which costs the taxpayer on average $8,000 per person per year is counted as little or nothing if given to a poor person but is worth $8,000 when given a higher income person. Thus, this form of income is undercounted at low-income levels.
  • The report fails to take into account the fact that the bottom quintile has far fewer people in it than the top quintile, which has more than any of the other quintiles. The high-income quintile contains the most families, thus more children and dependent workers. The low-income share of the bottom quintile is, to a substantial degree, due to the fact that there are fewer people in it and even fewer workers.
  • Another issue that is becoming increasingly important is the withdrawal of funds from IRAs and KEOGH accounts by the elderly during retirement. CBO treats these withdrawals as income, not dissaving. While payouts from IRAs are taxed when they are taken, they constitute income earned prior to retirement. Payouts from pensions are counted as current income. Because of the decline in guaranteed-benefit pension plans and the rise of contribution-type IRA plans, we can expect the share of “income” of the elderly to drop significantly in the future.

The author gratefully acknowledges the help of Robert Rector in preparing the brief overview of the CBO study on changes in household income.