In this interview, Janet Gornick discusses how the size of the middle class in the United States and other rich countries has shifted in recent decades, and the implications for inequality.
In recent years, the shrinking of the middle class has been frequently cited in articles and public debates about rising inequality in the United States. But a recent essay by Janet Gornick, coauthored by Graduate Center Ph.D. candidate Nathaniel Johnson and published in the Items series from the Social Science Research Council (SSRC), shows that the increase in U.S. inequality was in fact driven by changes above the middle class’s upper boundary.
A key point in the essay is how the middle class is defined. Unlike many studies that define the middle class as a fixed percentage of the income distribution, Gornick and her coauthor chose a definition that allowed the size of the middle class to change. We recently spoke to Gornick about the pressures the middle class is facing and the findings that she reported in “Income Inequality in Rich Countries: Examining Changes in Economic Disparities.”
I was surprised to see that, even back in 1985, the United States had a substantially smaller middle class than the other countries you studied. Yet it seems that there’s a widespread belief in — or nostalgia about — the American middle class. Is this entirely misguided?
Let me start with two technical points. First, there is no single definition of the middle class: not in research, nor in policymaking. That’s different than with poverty, where governments, including ours, issue official poverty definitions. So while Americans have ideas and perceptions about our middle class, those are based on a diversity of concepts and definitions. Some focus on family structure; others on education or occupation. In this study, we rely on income to define the middle class. It’s a fairly common way to define the middle, but it’s only one approach.
Second, the comparatively small middle class in the United States, both 35 years ago and now, goes hand-in-hand with high levels of inequality. We, like many scholars, define the middle class as households whose incomes fall in the middle of their country’s income distribution. So a small middle class implies a large share of poor households, or a large share of affluent households, or both. The size of the middle class, in itself, is a measure of inequality.
It’s definitely true that, earlier in the 20th century, especially in the decades following the Second World War, the U.S. and other industrialized countries saw the rise of a new middle class. The idea of the middle class was intertwined with the idealized family: one in which men, on their earnings alone, could comfortably support a family. And that was real. But by 1985, where our study begins, that middle class was already struggling. Starting in the 1970s, male earnings fell, and families increasingly needed two incomes to reach the middle class. Employment in general became less stable. In most locations in the U.S., the prices of housing, education, and healthcare rose. The middle class got squeezed.
I was also surprised to see that, from 1985 to 2016, the size of the middle class has not changed in the United States. There have been so many articles about the shrinking of the U.S. middle class. But your study found that “the overall increase in inequality in the United States was driven mainly by changes that occurred among the affluent — changes that arose above the upper threshold of the middle class.”
A common way to define the middle class — not what we did in this study, but what many have done — is to define the middle class as the middle 60 percent of the income distribution. A sort of irony, given the public discussion, is that the middle class, if defined as the middle 60, can’t shrink. It can get better or worse off, meaning its average income level can rise or fall and its share of the total income pie can grow or shrink, but its size can’t change. There’s a lot of talk about the shrinking middle, but in fact what people generally mean is that families who fall into the middle are hurting, and they’re hurting more over time.
A good bit of this discussion, this anxiety about the harm that’s been done to the middle, really hasn’t made a lot of sense. This messiness is part of what motivated our study. We decided to choose a definition where the size of the middle class could actually change. Then we took a close look at what happened.
If you define the middle class by income alone, as we did, there are still two approaches. One is to choose a portion of households and lock them in place, selecting, as I mentioned, the middle 60; some choose other groupings — say, the middle 50 or middle 40. We did something different, turning to a method we call “drawing bands around the middle.” Specifically, we defined the middle class as households whose income falls between 50 and 150 percent of their country’s median household income. The point is, if you take that approach, the size of the middle class can rise or fall over time. In fact, as income inequality has increased in many countries, the middle class has shifted. Some households have fallen into poverty; others have moved into affluence. The balance of those two shifts determines what happens to the size of the middle class.
You found that, in about half of the countries you studied, the size of the middle class fell substantially — in fact, by about 10 percentage points. But the middle class in the U.S. did not follow that pattern. Did that surprise you?
First, a technical point. National income distributions are highly diverse; their shapes are complex and vary across countries and over time. So two countries can have the identical level of inequality, captured by the Gini coefficient or another inequality measure, while the underlying distributions might be entirely different. One might have a large share that is extremely poor, while another might have a large share that’s exceptionally rich.
Given that diversity, I was not especially surprised when we found that, over our 30-year period, the size of the U.S. middle class remained the same: 59 percent. I was not especially surprised because I knew that any pattern of change was possible. Also, note that, in 1985, the U.S. middle class was the smallest among our study countries. Although I don’t think there’s any economic reason that there would be a floor, I had an instinct that perhaps, by 1985, the U.S. middle class had already shrunk to its limits.
It may seem inconsistent that U.S. inequality increased between 1985 and 2016, but its middle class size was unchanged. In fact, the explanation is rather simple. The main change, in the U.S., took place above the upper threshold of the middle class; in other words, the affluent grew more affluent. In other countries, as our study showed, inequality took a different shape: elsewhere, rising inequality went hand-in-hand with a shrinking middle.
What’s important about our U.S. finding — that the top jumped sharply upward — is that prior research has shown that, if the gap from the top to the middle grows too large, the result is a weakening of institutions that are required for a strong and stable middle class. The super rich often withdraw from several crucial institutions: public schools, health insurance, security, transportation, even the energy grid. When the rich become super rich, that causes a deterioration in the infrastructures that strengthen the middle. So, in the U.S., during the decades that we studied, the size of the middle class was stable, but the fact that these households were becoming more distant from the ultra rich brought costs that made life more difficult, more risky, and more precarious.
Now that high and rising inequality has become a mainstream issue in the U.S., do you think policymakers and the public are willing to enact changes that would reduce inequality? And if so, what kinds of policy reforms would be most effective?
First, political scientists have argued for a long time that the U.S. public has more “taste” for reducing inequality than existing policymaking would suggest. That said, public opinion about inequality is not my expertise; that would be the expertise of the Stone Center’s Leslie McCall.
What I have studied extensively are the policies themselves that can reduce income inequality. Basically, we know what could reduce income inequality in the U.S. First, we could reform policies and institutions that reduce earnings inequality: that would mean, for example, raising minimum wages, strengthening unions, and protecting employed workers from insufficient work hours. Second, we could make our income tax system much more progressive. And we could increase income transfers — payments from governments to households — and make them more progressive as well, especially for households with children.
The U.S. is an outlier among rich countries with respect to nearly every one of these policy tools. We, in the United States, face exceptionally high inequality not because we don’t know how to reduce inequality, but because we have lacked the political will.
In the conclusion of your SSRC essay, you say: “We write this brief during a time of sudden, unfathomable, social and economic upheaval. We can only imagine how the outcomes that we have reported here will evolve during the coming months and years.” I can’t help but wonder if you think the pandemic and the economic fallout will open opportunities for change? Or is inequality likely to become even higher?
I think we have to assess this in temporal phases. First, in the short term, we know that the virus has struck unequally, with disproportionate levels of illness and death among the poor and people of color. If that’s not distressing enough, Covid-19 has already increased income inequality. Several economists have been assessing who’s been hit hardest via the employment crisis. Yonatan Berman, a colleague associated with the Stone Center, just released a study in which he found that the Covid-19 outbreak in the U.S., as of mid-April, has had a major negative impact on earnings, and that has been concentrated among the bottom fifth of earners. That’s due to a combination of massive layoffs and reductions in working hours, both disproportionately affecting lower-paid workers.
I think that the key question remains open: After the acute public health crisis has receded and labor markets have recovered, will we see reforms in policies and institutions that could fundamentally reshape our income distribution?
Historians have been reflecting on this question. If you look at a long sweep of history, lessons are mixed. Some crises catalyzed large-scale changes that reduced inequality, while others did not. So, really, it’s impossible to predict what lies ahead in terms of structural change.
My crystal ball is foggy, but I can tell you what I hope for. My hopes now are focused close to home. I am deeply concerned about the impending budget cuts that seem to be on the horizon for New York State and New York City. Our own university is looking at extreme austerity measures and those cuts will greatly harm our students, potentially stripping both salary and health insurance from hundreds of them, possibly thousands. I would like to see New Yorkers demand that these and other budget cuts be dramatically lessened by tax increases, specifically increases targeted on the most affluent. There is an absurd amount of income and wealth held by those at the top in our state and especially in our city. Regardless of how the fantastically well-off have gained their fortunes, I hope that we will call upon them to pay it forward. It will be a tragedy if this crisis is met only with austerity and budget cuts, without some counter-balancing measures to increase taxes.