January 23, 1997
University Gazette


Full contents
Police Log
Gazette Home
Gazette Archives
News Office


  The Economics of Urban Unrest

By Eileen K. McCluskey

Special to the Gazette

When pondering which scholarly disciplines grapple with controversial social issues, economics may not be the first field that jumps into your mind. But for Edward L. Glaeser and Denise DiPasquale, economics provides the systematic lens they need to peer into such mysteries as why riots burst out in some crowds but not in others, and why riots occur in the first place. These topics are addressed in their coauthored working paper, "The L.A. Riot and the Economics of Urban Unrest," which has been accepted for publication in The Journal of Urban Economics.

Perhaps not surprisingly, a study of riots from within an economic framework is a rare undertaking. As Glaeser noted, "There hasn't been a whole lot of economics work on riots." Glaeser, recently named Paul Sack Associate Professor of Political Economy, added, "What's fascinating about riots is they're connected with a lot of very important issues, like crime and revolution."

DiPasquale, who taught at Harvard until one year ago when she left to become a visiting associate professor in the Collegiate Division of the Social Sciences at the University of Chicago, spoke of her impetus for the study on riots: "You could not watch the events that unfolded in Los Angeles in 1992 without being overwhelmed by what you were seeing. The popular press alluded to the underlying economic conditions, but no one tried to test the impacts of these conditions" in an attempt to explain why riots occur.

Disturbing Findings

The working paper's title does not capture the full range of riots studied by DiPasquale and Glaeser. Their study "include[s] international riots data, evidence from Los Angeles in the 1990s, and cross-city U.S. evidence on race riots from the 1960s."

Some of the scholars' major findings are that the outbreak of riots "are most closely linked to the size of the nonwhite community. . . . The nonwhite unemployment rate is also significantly correlated with riot occurrence."

When asked to comment on the why of their results, DiPasquale said, "It's crucial to put the findings in context. We're trying to explain a very complicated phenomenon and at best this paper is suggestive."

Why are nonwhite communities more apt to riot? "In our country, the nonwhite community is, by and large, oppressed," said Glaeser. Elaborating on this point, he noted that a sense of having been wronged can be a predictor of riotous behavior. Individuals in a crowd turning violent, he noted, "may be seeking retribution for a perceived aggrievance."

The Anonymous Crowd

A feeling of being the victim of social inequities can sow enough seeds of discontent, according to Glaeser and the study, to inspire an otherwise peaceful individual to, for instance, throw a brick at an authority figure -- like a police officer. The very existence of a crowd may also make individual members feel safer from retribution than if they stood in isolation. "The presence of crowds," said Glaeser, "makes law enforcement difficult." Glaeser noted that from the perspective of members of the crowd, "there's a sense of anonymity."

A Look at Differing Social Norms and Community Structure

In the paper's abstract, Glaeser and DiPasquale declared, "In our results, ethnic diversity seems a significant determinant of rioting, while we find little evidence that poverty in the community matters."

Glaeser expanded on the element of diversity: "People with different ethnicities who live in close proximity to each other" can more easily get into conflicts where the "perception is that the different group is violating social contracts." These diverse groups, Glaeser noted, may be operating under incompatible social norms, so that seeds of hostility may find fertile ground. "Misunderstandings," said Glaeser, "are some of the stuff of riots."

DiPasquale noted that a community whose fabric is torn seems more vulnerable to riots: "Some community-level variables do seem to matter in our results, but we really weren't able to measure these variables very well." She emphasized that this paper's results call for more in-depth studies of community structure -- analyses that may attempt to measure, for instance, community cohesiveness, and the impact of community institutions, as variables in violent outbursts. The paper, DiPasquale commented, "points us well beyond the question, 'Is it a poor community?' We need to look instead at how connected people are to each other."

No Easy Answers

The scholars note that questions about why riots happen are by no means brought to closure with this economic analysis. Glaeser said "We're expanding our knowledge base about why human beings behave the way they do, but this study points only to how complex an area this is, not to definitive conclusions."


Copyright 1998 President and Fellows of Harvard College