Economists have noted America’s growing economic and geographic polarization for some time. Enrico Moretti at the University of California, Berkeley, called it “the great divergence” between upscale and downscale cities. It’s also well known that some cities tax more and spend more per resident than others.
The question is why, and what it means. In her research, Rebecca Diamond, an assistant professor of economics at Stanford Graduate School of Business, gets answers by delving into data on geography, wages, housing costs, the movement of workers, and the presence of amenities such as restaurants, bars, and stores.
In a recent paper on why some local governments tax more but don’t seem to produce better government services, Diamond used geographic data to reach a surprisingly simple answer: because they can.
Her theory was that a city’s residents will often “vote with their feet” if they can. If it’s easy to move from a central city to a lower-tax suburb, a central city risks losing residents and ultimately revenue if it raises taxes sharply. But if a city has geographic barriers that make it hard to add housing, officials may be able to extract higher “rents” — higher taxes — without any repercussions.
Consider San Francisco and Phoenix. Both cities are highly sought after and have seen their populations increase. But due to a shortage of open land in San Francisco, as well as restrictive land-use regulations, that city’s housing supply hasn’t kept up with demand and prices have soared. By contrast, Phoenix has lots of available land and has increased housing construction. As a result, housing prices in Phoenix have climbed much more slowly.
San Francisco’s shortage of new housing supply allows the local government to raise taxes. Higher taxes may prompt some families to live in other cities, but higher taxes won’t have much effect on population growth — which determines the tax base — because that is already being limited by the scarcity of land for more housing. If Phoenix tried to raise taxes, however, the decline in housing demand would substantially reduce the growth of both its population and its tax base. That risk makes cities such as Phoenix more reluctant to raise taxes.
How Geography Affects Taxes
Sure enough, a comparison of more than 250 metropolitan areas confirmed Diamond’s hunch: Cities that were hemmed in by geographic barriers tended to impose higher taxes. Diamond estimates that San Francisco’s shortage of land for new housing enabled it to levy taxes 23 percent higher than Phoenix’s.
How local governments chose to spend their additional tax revenue varied. In cities that allowed collective-bargaining by public-employee unions, some of those tax dollars flowed into higher wages for government employees.
Diamond calculated that if San Francisco had the same land availability as St. Louis, it could be saving an average of $11,000 in salary per municipal employee. That would amount to an annual saving of $309 million, or 12.8 percent of San Francisco’s tax revenue.
At first blush, that might seem to justify moves by Wisconsin and other states to ban collective bargaining by government workers. On closer examination, though, Diamond found a startling twist: In cities that prohibited collective bargaining, a shortage of land available for real estate development led to higher levels of corruption. Diamond thinks that collective bargaining may give workers a formal way to bargain for a share of the extra tax revenues, while corruption provides an “informal” alternative way for government workers to capture rents when collective bargaining isn’t allowed.
College Graduates are Clustering Together
Geography is also at the heart of Diamond’s paper on the factors that seem to be dividing the nation into upscale white-collar cities and downscale blue-collar ones.
It’s already well established that income inequality has risen sharply over the past two decades, and that higher-earning college graduates are living farther apart from lower-income blue-collar workers.
What hasn’t been clear is why this geographic class sorting is occurring, whether it is compounding the effects of income inequality, and what it implies for cities trying to speed up growth.
Analyzing two decades of data from more than 200 cities, Diamond found that college graduates are increasingly clustering in more expensive cities that offer more amenities — more restaurants and cultural attractions, better parks, less crime, and less pollution.
An increase in the ratio of college graduates also correlated with increases in wages for workers across the board, but those wage increases were disproportionately higher for college graduates.
That disparity seems to have increased the geographic sorting between college and high school graduates. An increase in the ratio of college graduates was very tightly correlated with an increase in rents, and higher rents led to a disproportionate out-migration of non-college graduates. Everybody may like the amenities of an upscale city, but lower-income workers are less willing to tolerate increases in the cost of housing.
That might not seem surprising on a gut level — lower-paid workers are much more likely to live in Hoboken than Manhattan, or Oakland than San Francisco. But economists have been unsure how the psychic tradeoffs play out in practice. It’s possible, for example, that Manhattan’s sky-high housing costs are outweighed by the value of its social benefits.
Diamond’s results show that those amenities are, in fact, a net benefit to college graduates, even after discounting for higher costs of living, and that they magnify what is already a widening income gap. Diamond estimates that “welfare inequality,” the net benefits of living in an upscale city, effectively adds 25 percent to the simple wage gap between college- and high-school graduates.
Put another way: The wage gap probably understates the real level of economic inequality. Diamond also discovered that the sorting between rich and poor cities seems to be a self-reinforcing trend. Cities that had above-average shares of college graduates back in 1980 were also more likely to keep attracting disproportionate numbers of college graduates – and boosting their prosperity — in the 20 years that followed. That was true even in some cities that lost many traditional employers.
In other words, college graduates attracted more college graduates. Their presence spurred increases in amenities, but a high percentage of college graduates was an “amenity” in itself. That allowed some cities to re-invent themselves more easily, but it also widened the gap between rich and poor cities.
Diamond points out that Detroit and Boston both suffered from sharp declines in factory work in the 1970s, but Boston entered the 1980s with a much higher share of college graduates and never stopped attracting them. As a result, Boston continued to thrive while Detroit decayed to the point of near collapse.
New Questions About Inequality and Growth
Diamond is now digging into a host of other questions about the relationship between inequality and economic growth. It’s not clear, for example, whether geographic class-sorting increases an upscale city’s overall productivity over the long term.
Diamond is also taking her research down to the neighborhood level. She wants to know about the costs and benefits of mixed-income neighborhoods, an issue that could affect decisions on whether to build affordable housing in wealthy neighborhoods. She also wants to know whether a mixed-income neighborhood can stay that way if the government doesn’t intervene in some way. Her findings could have implications for policymakers and city planners for years to come.