Riots: Behavioral Aspects

As suggested by Janowitz's and Marx's typologies, the forms that riot events take are usually related to the particular incidents that spark them. In the case of communal riots, the most common sparks have involved perceived transgressions of racial/ethnic boundaries, either spatial, such as skirmishes that emerged over access to recreational facilities or religious sites, or sexual, such as accusations of rape made against a member of a minority racial/ethnic group by members of another group. In the case of commodity-type riots, the precipitants have ordinarily involved instances of police injustice, such as the excessive use of force against members of racial/ethnic minorities. Such was the spark of many 1960s-era riots as well as more recent events in Miami and Los Angeles. In the latter cases, the degree of injustice was even more palpable after police officers were acquitted of beating black motorists. Yet the 1919 Chicago riot also developed in reaction to police activity, or in this particular case, the lack thereof, when a white police officer refused to arrest Irish teens who had stoned a black teenager swimming in the "white" section of Lake Michigan. So communal riots, like commodity riots, may also be sparked in part by police behavior and corresponding perceptions of injustice. In the case of Miami and Los Angeles, anger at police and civil authorities was diffused toward Asian business owners, indicating that the targets of violence need not correspond to the immediate event that sparked rioting. Thus, while precipitating incidents may offer valuable insight into the motives for riot participation, by no means do they supply a full explanation of the origins of these events. Rather than one decisive spark, most riots have been preceded by a series of smaller incidents rooted in the social structure of riot areas.

Prior to the "spark" or precipitating incident, riot events are foreshadowed by a gradual escalation of resentments or grievances held by people who live in the riot area. These feelings, correlated with the structural characteristics of riot communities, represent the underlying conditions of riot genesis—the fuel that feeds the fire when the appropriate spark is provided. There are three general sets of structural explanations for the origins of riot violence: economic, political, and demographic. The first of these economic explanations suggests that people riot in response to conditions of abject poverty. A second and related economic explanation is that people riot when they see themselves as deprived relative to members of higher economic strata. Relative deprivation may spark rioting among those whose economic fortunes are improving but not fast enough to fulfill their rising expectations. By contrast, riots may also develop as a response to political disfranchisement, breaking out in areas where particular groups are politically underrepresented and angry about their lack of access to institutional power. Similarly, riots may represent a reaction to racial/ethnic segregation, which often is combined with economic deprivation and political exclusion. By this measure, places with the highest levels of minority racial/ethnic population will be more riot-prone. Finally, the origins of rioting might also be discovered in processes of demographic change that alter the racial/ethnic composition of communities, affecting cultural control of institutions, and the psychic well-being of longtime residents who fear such changes. Thus rioting may be related to a general process of ethnic succession and competition.

Empirical studies using census data and statistical models to examine the association of structural conditions with riot activity have yielded widely varying results. Comparing cities that had riots to those that did not, Downes found support for an economic deprivation explanation. Poverty, unemployment, and the quality of housing were statistically related to the presence of rioting in cities during the 1960s. Using a similar technique but a larger and more diverse sample of riot events from 1913 to 1963, Lieberson and Silverman found that cities where riots took place had political structures that minimized minority participation on city councils and police forces. By contrast he found no association between unemployment or dilapidated housing and the presence of rioting. Nor did he find any effects of population change on rioting. Seymour Spilerman's influential empirical studies of riot violence found that neither economic deprivation nor population changes were able to account for the frequency or severity of rioting in his multicity sample. Rather, he found that the only significant predictor of riot frequency and severity among cities was the size of a city's black population (Spilerman, 1970, 1976). Later research utilizing more refined statistical modeling has challenged these results. Using Spilerman's data, studies by Olzak, Shanahan, and McEneany and by Myers both found interactive effects of racial/ethnic population change and economic factors on the likelihood of multiple riots in cities, providing evidence that racial/ethnic composition was a significant factor in the outbreak of riot violence. Finally, comparing census tracts within cities that experienced rioting, Bergesen and Herman and Herman found that rioting was associated with processes of ethnic succession and competition, confirming Olzak et al. and Myers's findings at a local neighborhood level. Summarizing these studies, there is a clear consensus that rioting is most likely to take place in predominantly black and poor communities, but black neighborhoods undergoing ethnic succession are even more prone to riot than stable segregated areas. Furthermore, it remains evident that rioting is also a joint product of political, economic, and demographic factors. Political exclusion cuts off channels for the redress of grievances, acting in concert with segregation, population change, and economic competition to foster violence.