The Exaggerated Death of the Middle Class. By Ron Haskins and Scott Winship
Brookings, December 11, 2012
The most easily obtained income figures are not the most appropriate ones for assessing changes in living standards; those are also the figures that are often used to reach unwarranted conclusions about “middle class decline.” For example, analysts and pundits often rely on data that do not include all sources of income. Consider data on comprehensive income assembled by Cornell University economist Richard Burkhauser and his colleagues for the period between 1979—the year it supposedly all went wrong for working Americans—and 2007, before the Great Recession.
When Burkhauser looked at market income as reported to the Internal Revenue Service (IRS), the basis for the top 1 percent inequality figures that inspired Occupy Wall Street, he found that incomes for the bottom 60 percent of tax filers stagnated or declined over the nearly three-decade period. Incomes in the middle fifth of tax returns grew by only 2 percent on average, and those in the bottom fifth declined by 33 percent.
Things appeared somewhat better when Burkhauser looked at the definition of income favored by the Census Bureau which, unlike IRS figures, includes government cash payments from programs like Social Security and welfare, and looks at households rather than tax returns.
Still, the income of the middle fifth only rose by 15 percent over the entire three decades, much less than 1 percent per year. The Census Bureau reports that from 2000 to 2010, the income of the middle fifth actually fell by 8 percent. With numbers like these, it’s understandable why so many people think the American middle class is under threat and in decline.
But there are three reasons why even the Census Bureau figures are deceiving. The size of U.S. households, which has been declining, is not taken into account. The figures ignore the net impact on income of government taxes and non-cash transfers like food stamps and health insurance, which benefit the poor and middle class much more than richer households, and the value of health insurance provided by employers is also left out.
Burkhauser and his colleagues show that if these factors are taken into account, the incomes of the bottom fifth of households actually increased by 26 percent, rather than declining by 33 percent. Those of the middle fifth increased by 37 percent, rather than by only 2 percent. There is no disappearing middle class in these data; nor can household income, even at the bottom, be characterized as stagnant, let alone declining. Even after 2000, estimates from the Congressional Budget Office (CBO) show the bottom 60 percent of households got 10 percent richer by 2009, the most recent year available.
Making sense of income trends
Aside from the brighter picture presented by the Burkhauser and CBO analyses, there is a more complicated trend emerging in the United States. Four factors, both inside and outside the market, explain those trends.
The first market factor affecting middle-class income is a longtime trend of low literacy and math achievement in U.S. schools, which partially explains why conventional analyses of income show stagnation and decline. Young Americans entering the job market need skills valuable in a modern economy if they expect to earn a decent wage. Education and technical training are key to acquiring these skills. Yet the achievement test scores of children in literacy and math have been stagnant for more than two decades and are consistently far down the list in international comparisons.
It is true that African American and Hispanic students have closed part of the gap between themselves and Caucasian and Asian students; but the gap between students from economically advantaged families and students from disadvantaged ones has widened substantially—by 30 to 40 percent over the past 25 years.1
In a nation committed to educational equality and economic mobility, the income gap in achievement test scores is deeply problematic. Far from increasing educational equality as an important route to boosting economic opportunity, the American educational system reinforces the advantages that students from middle-class families bring with them to the classroom. Thus, the nation has two education problems that are limiting the income of workers at both the bottom and middle of the distribution: the average student is not learning enough, compared with students from other nations, and students from poor families are falling further and further behind.
It is difficult to see how students with a poor quality of education will be able to support a family comfortably in our technologically advanced economy if they rely exclusively on their earnings.
The second market factor is the increasing share of our economy devoted to health care. According to the Kaiser Foundation, employer-sponsored health insurance premiums for families increased 113 percent between 2001 and 2011. Most economists would say that this money comes directly out of worker wages. In other words, if it weren’t for the remarkable increase in the cost of health care, workers’ wages would be higher. When the portion of market compensation received in the form of health insurance is ignored in conventional analyses, income gains over time are understated.
Turning to non-market factors, marriage and childbearing increasingly distinguish the haves and have-nots.
Families have fewer children, and more U.S. adults are living alone today than in the past. As a result, households on average are better off since there are fewer mouths to feed, regardless of income. At the same time, single parenthood has grown more common, thereby increasing inequality between the poor and the middle class. Female-headed families are more than four times as likely to be in poverty, and children from these families are more likely to have trouble in school as compared with children in married-couple families. The increasing tendency of similarly educated men and women to marry each other also contributes to rising inequality.
The most important non-market factor is the net impact of government taxes and transfer payments on household income. The budget of the U.S. government for 2012 is $3.6 trillion. About 65 percent of that amount is spent on transfer payments to individuals. The biggest transfer payments are: $770 billion for Social Security, $560 billion for Medicare, $262 billion for Medicaid, and nearly $100 billion for nutrition programs. In addition to these federal expenditures, state governments also spend tens of billions of dollars on programs for low-income households. Almost all of the over $1 trillion in state and federal spending on means-tested programs (those that provide benefits only to people below some income cutoff) goes to low-income households.
Thus, taking into account the progressive nature of Social Security and Medicare benefits, the effect of government expenditures is to greatly increase household income at the bottom and reduce economic inequality.
Similarly, federal taxation—and to a lesser extent state taxation—is progressive. Americans in the bottom 40 percent of the income distribution pay negative federal income taxes because the Earned Income Tax Credit and the Child Tax Credit actually pay cash to millions of low-income families with children.
IRS data on incomes incorporate only the small fraction of transfer income that is taxable. Census data includes all cash transfer payments but leaves out non-cash transfers—among which Medicaid and Medicare benefits are the most important—and taxes.
The bottom line is that market income has grown, and government programs have greatly increased the well-being of low-income and middle-class households. The middle class is not shrinking or becoming impoverished. Rather, changes in workers’ skills and employers’ demand for them, along with changes in families’ size and makeup, have caused the incomes of the well-off to climb much faster than the incomes of most Americans.
Rising inequality can occur even as everyone experiences improvement in living standards.
Even so, unless the nation’s education system improves, especially for children from poor families, millions of working Americans will continue to rely on government transfer payments. This signals a real problem. Millions of individuals and families at the bottom and in the middle of the income distribution are dependent on government to enjoy a decent or rising standard of living. While the U.S. middle class may not be shrinking, the trends outlined above make clear why this is no reason for complacency. Today’s form of widespread dependency on government benefits has helped stem a decline in income, but far better would be to have more people earning all or nearly all their income through work. Getting there, though, will require deeper reforms in the structure of the U.S. education system.
1 Sean F. Reardon, Wither Opportunity? Rising Inequality and the Uncertain Life Chances of Low-Income Children (New York: Russel Sage Foundation Press, 2001).