Editors' Summary

The Brookings-Wharton Papers on Urban Affairs presents new research to a broad audience of interested policy analysts and researchers. The papers and comments contained in this volume, the fifth in the series, were presented at a conference at the Brookings Institution on October 23 and 24, 2003. The conference amply displays the breadth of issues that fall under the rubric of urban economics and includes topics as diverse as the influence of the 1960s riots on long-term outcomes in urban labor markets, alternative explanations for the increasing attraction of people to cities with greater human capital, the role of educational policies in determining residential location choices, the effects of immigrants on the housing opportunities available to native renter households, the impact of states' parole practices, and the effects of larger urban markets on the diversity of television programming.

The Labor Market Effects of the 1960s Riots

William J. Collins and Robert A. Margo investigate the impact of the 1960s riots on long-term labor market outcomes. Although many studies have examined the determinants of race-related civil disturbances in the 1960s, few have explored the consequences.

Collins and Margo analyze riots as natural disasters with direct and indirect effects on the level and location of economic activity. While direct effects of destroyed physical capital may be small, the economic impact may be magnified by indirect effects felt over a long time. The authors envision a process of potential self-reinforcing decline as a result of a riot. For a business owner whose property is damaged, potential costs might rise because of higher security expenses and insurance premiums. Potential benefits of remaining in business may decline if customers avoid the area. As businesses decide to close, business revenues and personal income may fall; tax revenues and thus public services may decline; and falling rental and property values may increase poverty. For the population unable to relocate, a new labor market equilibrium may ensue with fewer employed workers, lower wages, and fewer residents with high human capital.

To identify the effects of riots, they cite the work of Seymour Spilerman, who shows that black population size and region were the two most important explanatory factors in predicting the incidence and severity of riots. They also use a two-stage approach, with the number of days of rain in April 1968 used as an instrument. The authors show that just 6 of 752 riots accounted for roughly 60 percent of the total fatalities.

Drawing from both city and individual-level data, the authors conclude that riots were associated with slower subsequent income growth for blacks and that cities that experienced riot-related relative declines in income during the 1960s did not catch up during the 1970s. Cities that had riots had relatively large subsequent increases in the share of the city population that was black. The onset of severe riots did not affect the likelihood of employment for black men between 1950 and 1970 but caused a 3 to 4 percentage point decline in the 1970s in the employment rate of black men in cities affected by severe riots.

Several hypotheses may explain the relative declines in income. The riots may have hurt the labor market outcomes of people living in the riot cities during the 1960s and 1970s; relatively high-income individuals may have moved out, leading to a decline in average income; and family units may have disintegrated faster in cities stricken by riots than elsewhere.

The Rise of the Skilled City

Between 1980 and 2000, the population of metropolitan areas where less than 10 percent of adults had college degrees in 1980 grew on average by 13 percent. In areas where more than 25 percent of adults in 1980 had college degrees, average population growth was 45 percent. Similar trends have occurred during the past one hundred years in both Britain and the United States. Edward L. Glaeser and Albert Saiz explore alternative explanations of these trends.

The first question is whether the relationship demonstrates that growth causes increases in skills or that increases in skills cause growth. The authors test for causation...