print logo

The Myth of Middle-Class Stagnation

Friday, September 16, 2011

It’s time to correct a misconception regarding middle-class income.

Conventional wisdom says that the middle class hasn’t caught a break for at least a decade and that incomes have stagnated or declined. But new research corrects a misconception regarding middle-class income, and therefore should come as a pleasant surprise—not just to members of the middle class, but also to pundits, journalists, and of course, politicians.

Speaking to an audience in Racine, Wisconsin, on June 30, 2010, President Obama repeated a winning theme from his 2008 campaign—the theme that the economic policies of the previous administration had failed those of us who consider ourselves middle class:

Nearly a decade of tax breaks for millionaires and billionaires led to little more than sluggish growth [and] a shrinking middle class. Your paychecks flatlined. Wages and incomes did not go up. Even when the economy was growing, it wasn't growing for you.

It was the same message we had been hearing for two years: the benefits of U.S. economic growth after the Bush tax cuts had allegedly been going to the rich and bypassing the middle class.  By today, that mantra has become unchallenged conventional wisdom.

Today’s economy makes the problem even worse: the growth rate is somewhere between zero and pathetically anemic. So, even if the middle class did start getting its fair share of growth, that “fair share” wouldn’t amount to much. The rut just seems to get deeper and deeper for the middle class.

Wishful thinking?

Imagine a better outcome. For example, in the wake of the Bush tax cuts, wouldn’t it have been nice to see the middle class get its fair share of the overall economy’s growth? Wouldn't it have been even nicer if middle class income had grown at a faster pace than that of the rich?

All four versions of ‘middle class’ outperformed every definition of ‘rich’; in short, the gap between the rich and the middle class got smaller, not larger.

If the seven years before 2008 had yielded results like that, it might have driven that year’s political campaigns and subsequent debates away from the rear-view-mirror exercise of assigning blame for unfair distribution of past income growth and towards a forward-looking focus on our society’s choices for enhancing future economic growth. We could today be focusing on the single, overwhelmingly important problem facing our economy—jump-starting overall growth—armed with the confidence that the middle class, as before, would get at least its fair share of that growth.

But imagining those middle-class results is just wishful thinking, isn’t it?

No. In the seven years from 2001-2007 (inclusive), not only did the middle class get at least its fair share of overall income growth, the income gap between the rich and the middle class actually got smaller. In an apparent paradox, the same Census Bureau database that told us that median household income was essentially unchanged in 2007 versus 2000 also tells us that the middle class enjoyed a higher income growth rate than did either the overall economy or the rich—and therefore that their income gap versus the rich had actually decreased.

How is that possible? How could the same official database—the March Supplement to the Current Population Survey—lead us to the following two, seemingly polar-opposite, conclusions:

(a) median household income “flatlined,” underperforming the overall economy; and

(b) middle-class income growth outperformed that for the overall economy as well as that for the rich.

Although it seems impossible, both statements are true. The key lies in the difference between the “median household” versus the “middle class.” The median household is a single theoretical household exactly in the middle of the entire income-ranked list of U.S. households. Conversely, the “middle class” has no official definition, but it is certainly tens of millions of households in size and presumably centered around the median household.

If the median household was an accurate proxy for the “middle class” then both of the statements above could not be true simultaneously. On the other hand, if both statements are true at the same time, then the median household did not accurately represent the middle class. And so it turns out that there’s more to the story than we thought.

Clues strongly suggested that median household income was not representative of the middle class.

For several years, just about everyone has implicitly equated the median household’s performance with middle-class performance. Why? Well, let’s face it: with no official definition of middle class, it’s easy (and plausible) to assume that as the median household goes, so goes the middle class. It’s easier, anyway, than going to the trouble of defining the range of households that comprises the middle class, defining the range that comprises the rich, defending those two definitions, calculating the income growth rates for those two ranges, then, finally, comparing the results. Just hanging our hats on the median household and calling it “the middle class” is a much quicker path to judgment about the success or failure of the policies of the past. Unfortunately, relying on the median as a proxy for the “middle class” can lead to false conclusions about past policies—which in turn can lead to misguided future policies.

Poker hands, middle linebackers, and median households

Why doesn’t the median always tell us what we need to know? A few examples will help illustrate the point.

First, let’s say a poker dealer deals two five-card hands face down in front of her, looks at them, then flips the middle-ranking card face up in each hand: the seven of spades and the seven of clubs, respectively. Do I now have enough information to conclude that both hands have identical value in a poker game? Of course not; I need to see all ten cards, not just two.

Another example: Let’s say a sports fan asks me whether my local pro football team’s defense has improved over the year before, but the only thing I know for sure is that the middle linebacker is the same guy performing at the same level; no change from last year. Do I know enough to reply, “Not a chance; the defense is no better than last year”? Of course not; I need to know about all 11 players, not just one. I need to know about the whole team, not just the middle guy.

Lastly, let’s say an economist (or a politician, or a journalist) points out that the middle U.S. household didn’t enjoy any increase in real income during the seven years from 2000-2007, even though the economy had grown. And indeed, the chart below, adapted from page five of the Census Bureau’s 2007 report (published in mid-2008), unmistakably confirms that point.

Steve Conover RMHI

But does that give me sufficient information to conclude that middle-class income flatlined, that the benefits of growth had gone to the rich, and that the economic policies of the past had therefore failed the middle class? Of course not; I need to analyze the millions of households making up the middle class, not just the middle household.

Ominous clues about the median household

The Census Bureau booklet published in August 2008 was the source for the median household income statistic, but it also contained some clues that the median might not be telling us everything. For example, the Gini index of income inequality, comparing 2000 versus 2007, indicated no discernable change in overall inequality (0.462 vs 0.463, respectively, with 0.003 standard error; see page 40 of the booklet). Would an increase in rich-versus-middle income inequality be likely, given that overall inequality was unchanged? Hardly.

A further clue was the Census Bureau’s table titled "Share of Aggregate Income Received by Each Fifth and Top 5 Percent of Households," which clearly showed that the middle-income quintile (as well as the middle three quintiles—whichever comes closer to one's preferred definition of "middle class") equaled or increased their share of aggregate income in 2007 versus 2000, while both the top quintile and the top 5 percent lost income share. When the middle class maintains, or increases, its share of total income, is that “unfair” to them? Of course not; instead, it is indicative of a “fair share” of income growth. Indeed, the household income statistics for 2010, released this month and about which there has been much gnashing of teeth, confirm that the middle three quintiles maintained or increased their income share, while that of the top 5 percent decreased.

A closer look at the ‘middle class’

For my doctoral dissertation at the University of Texas at Dallas, I looked deeply into the Census Bureau’s numbers. There are more than 100 million households in the United States; ranking them by income (using proper database tools) enables us to select any portion of them to define income groups. To cover a wide range of plausible definitions of “middle class,” I chose four possible definitions: the middle 20 percent, 40 percent, 60 percent, and 80 percent of households. In a similar manner, three possible definitions of “the rich” were selected: the top 5 percent, 10 percent, and 20 percent of households. (Interestingly, when survey respondents have been asked which income class they’d place themselves in, close to 80 percent have tended to place themselves somewhere in the middle class. In contrast, the various definitions used by social scientists indicate that the middle 40 to 60 percent is a more-defensible range. Hardly anyone thinks they are part of “the rich.”)

I also took advantage of the detailed data and painstaking design of the government’s database, the March Supplement to the Current Population Survey, made available jointly by the U.S. Census Bureau and the Bureau of Labor Statistics. Its meticulous design, fine-tuned by experts over a period of decades, enables researchers to control for factors that might otherwise cause misleading inferences. Controlling for such factors helps to isolate the income return to a household for a given amount of work contributed to the economy by that household, and my research has incorporated the nuances.

One example of the results of my research is shown below; a chart that compares the growth rates of income per full-time-equivalent worker for the several possible definitions of “middle class” and “rich.”

Steve Conover Growth Rates

Although the recession/recovery during that period resulted in unimpressive aggregate income growth, the observations in the chart above nonetheless contradict conventional wisdom because three of the four chosen definitions of “middle class” outperformed the overall economy. In other words, the middle class got at least its fair share of overall growth, and arguably more. Moreover, all four versions of “middle class” outperformed every definition of “rich”; in short, the gap between the rich and the middle class got smaller, not larger.

Will we use the good news to our advantage?

The middle class finally caught a break. Figuring it out took us longer than it should have, but at least we now know that the middle class has in fact been getting its fair share of whatever income growth has resulted from the work we put into the economy.

And that’s not just good news for the middle class, that’s good news for everybody. Why? Because we now know that, if past is prologue, the middle class should automatically get its fair share of future growth under current tax policy, as it did during the Bush years of the recent past—which means it shouldn’t be necessary to spend any of our limited time for national discourse on whether or how to redistribute the benefits of whatever growth our economy can muster. The time we’ll save by sidelining the redistribution argument and the class-warfare rhetoric can be much better spent on the overwhelmingly important problem of how to kick the economy into higher gear, back to robust growth rates approaching 4 percent or more. The debate about our economy can now focus freely on which policies are best in the long run for creating jobs, jobs, jobs.


Steve Conover retired recently from a 35-year career in corporate America. He has a BS in engineering, an MBA in finance, and a PhD in political economy.

FURTHER READING: Conover’s blog is “The Skeptical Optimist,” which primarily covers economic growth and the political and economic implications of fiscal and monetary policy. John H. Makin writes more on these topics in “Systemic Risk Returns: Is the Current Crisis Worse Than the Lehman Collapse?” and “The Limits of Monetary and Fiscal Policy.”

Steve Conover, Doctoral dissertation, “U.S. Middle Class Income Stagnation, 2000-2007: Reality or Illusion?” August 2011, University of Texas at Dallas.

Income, Poverty and Health Insurance in the United States: 2007

Income, Poverty and Health Insurance in the United States: 2010

Table 678, Stat Abstract 2010

March supplement, Current Population Survey:

Image by Rob Green | Bergman Group

Most Viewed Articles

‘The American Banking System Might Not Last Until Monday’ By Alex J. Pollock 08/23/2014
Learning from the crises you’ve forgotten.
Peanut Butter’s Many Inventors By Edward Tenner 08/15/2014
The popular product illustrates both the opportunities and the risks of intellectual property.
A Flawed E-Cigarette Regulation By Sally Satel and Alan D. Viard 08/16/2014
The FDA's proposed regulation should not go forward in its current form, or it will undermine ...
The Long-Hours Luxury By Tino Sanandaji 08/04/2014
One factor that is often overlooked in the debate over causes of income inequality is a shift in ...
100 Years of the Panama Canal By John Steele Gordon 08/15/2014
One of the supreme engineering feats of the early 20th century, the canal has been an immense boon ...