clock menu more-arrow no yes mobile

Filed under:

Larry Summers makes the case for a government bailout of the American heartland

Summers and two co-authors offer proposals to help distressed parts of the country.

Dylan Matthews is a senior correspondent and head writer for Vox's Future Perfect section and has worked at Vox since 2014. He is particularly interested in global health and pandemic prevention, anti-poverty efforts, economic policy and theory, and conflicts about the right way to do philanthropy.

In 2016, only 5 percent of men ages 25 to 54 in Alexandria, Virginia (a rich DC suburb), were not working. In Flint, Michigan, the share was 51 percent.

That staggering fact frames a new paper by three Harvard economists — Benjamin Austin, Ed Glaeser, and former Treasury secretary/chief Obama economic adviser Larry Summers — released on Thursday as part of the Brookings Institution’s Papers on Economic Activity series.

Austin, Glaeser, and Summers argue that particular areas of the United States (specifically what they call the “Eastern Heartland,” defined as Michigan, Illinois, Missouri, Arkansas, Louisiana, Mississippi, Alabama, Tennessee, Kentucky, Indiana, Ohio, and West Virginia) have fallen behind on a number of economic indicators, but especially male joblessness.

Jobless rates among working-age men have been growing nationwide for decades now; the share of men ages 25 to 54 in the labor force fell from 97.1 percent at the start of 1960 to 89 percent at the end of 2017. Without that fall, there’d be 5 million more men in the workforce. The share of women working increased for most of that period, due to broader cultural changes, but has been stagnant since the mid-1990s.

Men aged 25-54 not working rates, 1980 and 2015 Austin, Glaeser, Summers 2018

While the whole country is experiencing this decline in male work rates, Austin, Glaeser, and Summers argue that the trend is geographically concentrated (as shown in the maps above, showing male joblessness in 1980 versus 2015) and that this justifies policies specifically targeting areas that are hardest hit. In particular, they argue that subsidies meant to encourage work, such as an expanded earned income tax credit, should perhaps be changed to be more generous for recipients in struggling regions, where they might do more to encourage additional employment and reduce joblessness.

The paper is notable not least for its authors. Austin is a PhD candidate and is thus enjoying a brief honeymoon in which he can’t be politically pigeonholed, but Glaeser has a generally free market orientation and is a huge booster of cities (especially coastal ones), which he argues have to grow larger to boost economic growth. Summers is practically synonymous with the center-left pro-market wing of the Democratic Party, drawing the ire of critics who argue that the policies Summers championed as an adviser to Presidents Clinton and Obama contributed to the decline of the Midwest and the rest of the “Eastern Heartland” the paper focuses on.

So it’s notable that Glaeser and Summers are now embracing an active government role in revitalizing struggling parts of the country, and trying to work through the best way to do it.

The case for “place-based” policies

While the idea that Youngstown, Ohio, needs more help than San Francisco might seem intuitive to a non-economist, the economic case for policies targeting certain areas, rather than certain kinds of individuals, is somewhat shakier.

As Austin, Glaeser, and Summers note, most variation in incomes is within regions, not between them. Only 1.2 percent of variation in individual incomes is explainable by what states people live in, and only 7.1 percent of variation is explainable by what sub-state region they live in. The US already has an individual income tax and income-based welfare policies that redistribute money based on what specific people make. Since region is at best a really imprecise proxy for how well individual people are doing, why implement policies to boost specific geographic areas when you could just direct money to poor individuals instead?

But Austin, Glaeser, and Summers argue that place-based policies are necessary because different regions respond to various public policies differently.

Suppose, for the sake of argument, that every $1 billion you spend boosting employment in West Virginia results in 50,000 people rejoining the labor force, but that $1 billion spent boosting employment in Wyoming results in only 10,000 people rejoining the labor force. It would make sense in such a situation to offer bigger employment subsidies in West Virginia than in Wyoming because doing so would bring more bang for your buck.

The authors roughly estimate how responsive workers in different areas are to employment subsidies that boost their wages by estimating how employment changes in different areas as wages rise and fall. Sure enough, they find that West Virginians are more than three times as sensitive to changes in the rewards to work as Wyoming residents are. That is: Employment subsidies targeted at struggling states like West Virginia are likely to be considerably more effective than subsidies to better-off states like Wyoming.

Possible reasons for skepticism

The idea of adopting policies to benefit specific parts of the US is hardly new, and typically runs aground due to concerns about effectiveness. As the authors readily concede, there’s little rigorous empirical evidence about how different areas respond to employment subsidies; their quick-and-dirty regression is informative, but you’d want a more sophisticated study to really confirm that West Virginians would respond more to subsidies than Wyomingites.

Other ideas to boost struggling areas have typically not proven particularly successful. Empowerment Zones — a common federal and state policy where businesses receive tax credits for hiring residents of distressed communities and local governments get block grants to fund business assistance, infrastructure, training, etc. in those areas — have had a very mixed track record at best when it comes to boosting employment and economic activity in disadvantaged areas. Most studies show little or no impact.

UC Irvine economist David Neumark and University of Bristol economist Helen Simpson argue that the evidence that big infrastructure projects (like the Tennessee Valley Authority) and education investments like public colleges and universities can boost struggling areas is better than the evidence for Empowerment Zones. They don’t review evidence for region-specific wage subsidies of the kind that Austin, Glaeser, and Summers suggest.

There’s also reason to be skeptical of how meaningful the divisions are that Austin, Glaeser, and Summers construct between the “Eastern Heartland,” “Western Heartland,” and “Coastal States”:

Divsion between Coastal States, Eastern Heartland, Western Heartland Austin, Glaeser, Summers 2018

The divisions they draw are necessarily a bit rough, given how much of the data they use is only available at the state level, but nevertheless, it produces regions that don’t look dramatically far apart. They find that male mortality is significantly higher in the Eastern Heartland than elsewhere, but GDP per worker isn’t dramatically different between the regions, and male joblessness in the Eastern Heartland is pretty close to that of the coasts:

Prime-age male mortality mortality Austin, Glaeser, Summers 2018
GDP per worker Austin, Glaeser, Summers 2018

As the maps at the start of this piece indicate, there are certainly areas of the country that are struggling much more than others with chronic joblessness. But they don’t seem to align very closely with the region definitions used in this paper, suggesting that more granular distinctions are necessary.

But perhaps the biggest weakness of the paper is its neglect of the place-based policy likeliest to boost growth: reducing zoning rules and other barriers to growth in rich coastal cities. That idea, which Glaeser himself has championed for years, would make people in those cities better off by lowering rents and encouraging more talented people to move in, but it could help struggling areas too. More people moving from Youngstown to Chicago means fewer workers left in Youngstown. That could create a labor shortage that forces employers in Youngstown to raise wages. That doesn’t always happen in practice, but it’s the explanation economists have typically offered for why migration helped in the past.

This paper, though, isn’t trying to figure out policies for big cities. It’s trying to figure out national policy and, in particular, a policy that can be compellingly pitched to states like Ohio and Michigan. That serves an economic need, but also a political need in a Democratic Party that found itself without a sufficiently compelling economic message in those areas in 2016. This paper is, in its way, a small step toward trying to fix that.

Sign up for the newsletter Today, Explained

Understand the world with a daily explainer plus the most compelling stories of the day.