One of the most commonly taught stories American schoolchildren learn is that of Ragged Dick, Horatio Alger’s 19th-century tale of a poor, ambitious teenaged boy in New York City who works hard and eventually secures himself a respectable, middle-class life. This “rags to riches” tale embodies one of America’s most sacred narratives: that no matter who you are, what your parents do, or where you grow up, with enough education and hard work, you too can rise the economic ladder.
A body of research has since emerged to challenge this national story, casting the United States not as a meritocracy but as a country where castes are reinforced by factors like the race of one’s childhood neighbors and how unequally income is distributed throughout society. One such study was published in 2014, by a team of economists led by Stanford’s Raj Chetty. After analyzing federal income tax records for millions of Americans, and studying, for the first time, the direct relationship between a child’s earnings and that of their parents, they determined that the chances of a child growing up at the bottom of the national income distribution to ever one day reach the top actually varies greatly by geography. For example, they found that a poor child raised in San Jose, or Salt Lake City, has a much greater chance of reaching the top than a poor child raised in Baltimore, or Charlotte. They couldn’t say exactly why, but they concluded that five correlated factors—segregation, family structure, income inequality, local school quality, and social capital—were likely to make a difference. Their conclusion: America is land of opportunity for some. For others, much less so.
A new working paper authored by the UC Berkeley economist Jesse Rothstein builds on that research, in part by zeroing in on one of those five factors: schools. The idea that school quality would be an important element for intergenerational mobility—essentially a child’s likelihood that they will one day outearn their parents—seems intuitive: Leaders regularly stress that the best way to rise up the income ladder is to go to school, where one can learn the skills they need to succeed in a competitive, global economy. “In the 21st century, the best anti-poverty program around is a world-class education,” Barack Obama declared in his 2010 State of the Union address. Improving “skills and schools” is a benchmark of Republican House Speaker Paul Ryan’s poverty-fighting agenda.
Indeed, this bipartisan education-and-poverty consensus has guided research and political efforts for decades. Broadly speaking, the idea is that if more kids graduate from high school, and achieve higher scores on standardized tests, then more young people are likely to go to college, and, in turn, land jobs that can secure them spots in the middle class.
Rothstein’s new work complicates this narrative. Using data from several national surveys, Rothstein sought to scrutinize Chetty’s team’s work—looking to further test their hypothesis that the quality of a child’s education has a significant impact on her ability to advance out of the social class into which she was born.
Rothstein, however, found little evidence to support that premise. Instead, he found that differences in local labor markets—for example, how similar industries can vary across different communities—and marriage patterns, such as higher concentrations of single-parent households, seemed to make much more of a difference than school quality. He concludes that factors like higher minimum wages, the presence and strength of labor unions, and clear career pathways within local industries are likely to play more important roles in facilitating a poor child’s ability to rise up the economic ladder when they reach adulthood.*
For Rothstein, there’s no reason to assume that improving schools will be necessary or sufficient for improving someone’s economic prospects. “We can’t educate people out of this problem,” he says.
His work, like Chetty’s, is not causal—meaning Rothstein is not able to identify exactly what explains the underlying variation in his economic model. Nevertheless, his work helps to provide researchers and policymakers with a new set of background facts to investigate, and signals that perhaps they should be reconsidering some of their existing ideas. (Both Raj Chetty and his co-author Nathaniel Hendren declined to comment for this story.)
Jose Vilson, a New York City math teacher, says educators have known for years that out-of-school factors like access to food and healthcare are usually bigger determinants for societal success than in-school factors. He adds that while he tries his best to adhere to his various professional duties and expectations, he also recognizes that “maybe not everyone agrees on what it means to be successful” in life.
Deborah Menkart, the executive director of Teaching for Change, agrees that Rothstein’s research supports what teachers have long known from their own experience on the front lines. “I think it also affirms the focus,” she said, on having “children not just see schools as a ticket out of poverty, as a way to ‘rise above’ your community, but as a way to [be] agents for change within their communities.” “Part of the problem,” she adds, “is that the whole conversation around education has become so focused on helping individuals ‘escape’ their bad circumstances, rather than helping them become part of the solution.”
Rothstein is quick to say that his new findings do not mean that Americans should do away with investments in school improvement, or even that education is unrelated to improving opportunity. Certainly the more that people can read, write, compute, think, and innovate, the better off society and liberal democracy would be. “It will still be good for us if we can figure out how to educate people more and better,” he says. “It might help the labor market, our civic society, our culture.” But Americans should be more clear, he says, about why they are investing in school improvement. His research suggests that doing so in order to boost a child’s chances to outearn their parents is unlikely to be successful. According to Rothstein, education systems just don’t go very far in explaining the differences between high- and low-opportunity areas.
Politicians rarely draw the same conclusions that Rothstein does, but his study certainly isn’t the first piece of evidence to suggest that increasing skills and educational attainment are often insufficient for disrupting economic trajectories.
According to Marshall Steinbaum, the research director at the Roosevelt Institute, economists have long believed that differing levels of skills and education (what the field refers to as “human capital”) is the most salient explanation for why individuals achieve such varied economic outcomes. “I think it’s becoming harder and harder to accept explanations like the so-called skills gap,” he says, referencing the popular idea that low-income people merely lack the necessary skills and training to thrive in the modern economy.
As a stronger explanation, Steinbaum points to the rise of “interfirm inequality,” a phenomenon in which even workers with very similar education histories, ages, and industries make very different amounts of money depending on which firms they work for.
Meanwhile, other studies have suggested that differences in local labor markets can affect economic outcomes and upward mobility. For example, in 2015, the left-leaning Center for American Progress, in conjunction with the economists Richard Freeman and Eunice Han, published a report building on Chetty’s work and found that union membership seems to be another critical factor helping poor people escape poverty. The researchers went beyond Chetty’s regional-level analysis to compare outcomes between individual union and nonunion households. They found that low-income children who grew up with parents in unions earned more as adults than the children of nonunion parents. They concluded that making it easier for individuals to collectively bargain would likely help boost economic mobility.
Rothstein’s work also comports well with some research findings out of Canada. The country has relatively high rates of intergenerational mobility—significantly higher than in the United States or Britain. That said, this past year, the University of Ottawa economist Miles Corak found that, similar to the U.S., where one grows up in Canada seems to largely determine one’s economic chances in adulthood.
Marie Connolly, an economist at the University of Quebec in Montreal who collaborates with Corak, told me that after studying geographic mobility across Canada, her team has identified similar patterns as Rothstein did in the United States.
“Education is just not a big part of the story,” she says. “You can see a little role for school quality, but the structure of the labor market seems to be a much bigger driver.”
Rothstein’s research is certainly not the final word on the subject, and his study comes with some caveats. The first, as mentioned earlier, is that his findings are not causal. It could be, for example, that whatever is going on in places where children are more likely to rise up the economic ladder, like Salt Lake City and San Jose, have low-income families that have some other unidentified characteristic—such as different parenting styles—that affects their children’s lives. It’s possible these families produce economically successful children for reasons that have nothing to do with where they live.
Another caveat, related specifically to Rothstein’s school-quality finding, is that low-income children who grow up in places that see lots of children transcend poverty could have some unobservable characteristic that leads them to have good employment outcomes but poor academic performance. And Rothstein does not identify specific schools in his paper when drawing his school-quality conclusions, meaning he’s making indirect inferences.
Ultimately, most Americans would probably agree that leaders should work to build great schools, and that individuals who work hard should be able to improve their economic earnings over time. Devoting the bulk of one’s attention to the former in the hopes that it causes the latter, however, might prove to be a real mistake.
* This article originally contained a graph showing that the "earnings gap with fixed skills" accounted for 16 percent of Jesse Rothstein's economic model. We regret the error.