The blog of the Urban Institute
December 6, 2012

Location and Inequality: Are the Income Gaps Real?

December 6, 2012

Debates about income inequality and the shrinking middle class have largely focused on globalization, the declining share of middle-wage jobs, the eroding role of unions, technological change that benefits more educated workers, tax policies, and the share of income going to the top 1 percent. Often ignored is the question of whether we’re really measuring inequality accurately.  Do standard measures of money income really capture inequalities in living standards?  Not really.  Because of differences in living costs across communities, higher incomes don’t necessarily translate into higher living standards.  Housing costs reduce purchasing power in some communities more than in others.  At the same time, income gaps can become more pronounced when low-wage workers are discouraged from moving to areas with high housing costs.  Where people live, it turns out, matters a lot in measuring and accounting for the inequality of living standards.

Analysts of poverty trends have long recognized that cash income does not tell the whole story, since it ignores the importance of noncash public benefits, such as food stamps, housing assistance, and health coverage. Indeed, the Census Bureau is now taking into account noncash benefits and differences in housing costs when measuring poverty. Recently, researchers have begun analyzing geographic differences to explain trends in income inequality.

Several mechanisms are potentially at work, as Enrico Moretti of the University of California, Berkeley points out. Living costs may increase faster in areas where high-income, highly educated people are concentrated. The rise in living costs may come from more rapid growth in housing prices and in the prices of other goods and services linked to rising land values. High-income people may have moved to metropolitan areas where housing costs are especially high. By contrast, low-wage, less-educated workers have been less likely to move to areas where they would earn higher wages but not higher living standards. Moretti finds all of these factors at work, showing that the rising inequality in money income didn’t completely translate into rising inequality in purchasing power. According to Moretti, more than 20 percent of the rising money advantage of college graduates over high school graduates between 1980 and 2000 did not represent an advantage in living standards.

Locational differences can make income inequality appear worse than the actual inequality of living standards.  Economists have long viewed local zoning requirements as harmful to low-income families by limiting their access to attractive suburban neighborhoods. By requiring large lot sizes, towns have priced low-income families out of their housing markets. Now, as highlighted by the New York Times Economix blog, Peter Ganong and Daniel Shoag of Harvard University have demonstrated that differential housing regulations are a major culprit in slowing the convergence of regional income gaps, thereby lessening the migration into high cash income, high cost areas communities and adding to the inequality in money incomes. High-wage areas used to attract all types of workers.  As the supply of workers, including low-skill workers, went up in high-wage metro areas and fell in low-wage metros, wage differences between metro areas declined.  In recent years, because high housing costs have increasingly offset higher wages, fewer workers have chosen to migrate within the United States. Thus, low-income families lose in two ways from restrictive regulations—those in highly regulated areas face higher prices because of limitations on supplies and those in other locations lose access to better paying jobs because they cannot afford the high-priced housing.  Housing subsidies can shield some low-income families from increased housing costs in high-priced areas, but most low- and middle-income families receive no housing benefits at all.

To see how money income fails to capture purchasing power differences, compare the ability of low- and middle-skill workers and of median-income families to buy homes in four high-priced and four low-priced metropolitan areas. To simplify, let’s look only at the burden of a 30-year mortgage at a 4 percent interest rate. As the table shows, workers at moderate education levels face enormous mortgage burdens trying to buy homes in the four high-priced metro areas. But, homes in low-priced areas are quite affordable even among workers without a college degree. The gaps in affordability are far less in the case of family income. Still, a median-income family would have to spend more than double their share of income on the median-priced home in the Los Angeles or San Francisco metro areas than in the four low-priced metro areas. As a result, family income inequality across cities looks far higher when we don’t account for differences in housing costs. Median family incomes are 40 percent higher in San Francisco than in Oklahoma City, but the gap in income after mortgage payments falls to 9 percent. Since other living costs are higher in San Francisco, the differences in purchasing power are even smaller. The figures illustrate how inequality in purchasing power is often lower than inequality in money incomes.  

Metropolitan Area Median Wage High School Graduate Buys at the 25th Percentile of Home Values Median Wage Worker with Some College Buys Median Priced Home Median Income Family Buys Median Priced Home Value of Median Priced Home
Percent of Income Needed to Buy Home
Boston, MA 53% 53% 18% $364,300
Los Angeles, CA 66% 71% 28% $438,300
San Francisco, CA 64% 81% 36% $719,800
Washington, DC 44% 48% 20% $422,400
Kansas City, MO 21% 27% 11% $133,800
Oklahoma City, OK 19% 25% 13% $135,200
Pittsburgh, PA 17% 23% 12% $90,500
South Bend, Indiana 16% 22% 12% $113,600


As an organization, the Urban Institute does not take positions on issues. Experts are independent and empowered to share their evidence-based views and recommendations shaped by research.


Thanks to Rolf for his interesting comments. Although I believe restrictive regulations do play a role in the size of housing price differences, this issue is not the main point of the blog. Instead, the blog's focus is on how wide differences in prices (home values and other prices related to land values) lead to an overstatement of real differences in purchasing power. Whatever the causes of the differences in home values, the question of how much current inequaliy measures overstate differences in purchasing power remains important as well.
Did you mean "Kansas City, MO" in the table?
Ganong and Shoag's database of appellate court cases is a very clever way to identify places where land use is highly contested. It's probably also a pretty good proxy for the stringency of land-use regulations. But there are positive feedbacks between land-use restrictions and high house prices. People living in states with higher house prices are more likely to impose regulations to protect their asset values (and to elect legislators who will allow those regulations), and landowners seeking to build housing are more likely to file land-use related lawsuits to build there. Regulations also have benefits that are usually not weighed against their costs. All this makes the policy implications of articles like the one by Ganong and Shoag and blogs like this one unclear. Exhorting cities and states to relax their land-use regulations hasn't worked in the past half-century of fulmination about it. In fact, with fiscal crisis deepening, local land-use restrictions, moratoria, and infrastructure-related regulations (e.g., impact fees) are likely to spread because the costs of growth to individual jurisdictions so often outweigh the benefits to the jurisdiction. Even if deregulation occurred, it's not clear that restriction would simply cease, since it's partly a consequence of underlying fiscal, political, and economic motivations. Reforms to infrastructure funding--channeling more funds to congested places, rebuilding older transport and water infrastructure, and VMT-based transportation pricing, for example-- could create incentives for some of the most regulated big cities and inner suburbs in high-cost metro areas to accommodate more new residents. Reducing exclusionary zoning in suburbs is important for reasons of social justice, but it's not likely to yield enough new construction to reduce housing price inflation by much.