Shocks to the Local Political System: Violent Protests and Changes in Policy Preferences

This piece, written by Adam B. Lerner, covers Dr. Ryan D. Enos, Harvard University, Dr. Aaron R. Kaufman, New York University, Abu Dhabi, and Dr. Melissa L. Sands’, University of California, Merced, new article, Can Violent Protest Change Local Policy Support? Evidence from the Aftermath of the 1992 Los Angeles Riot” 


In the case of political riots, does the sound and the fury signify nothing?

Political science has long understood riots—defined as “political acts in which participants engage in violence to express grievances and attempt to spur policy change”—as politically-motivated, but pinpointing their impact on policy preferences and political activity has proven more difficult. Recent media coverage of riots and protests in cities like Ferguson, Missouri and Baltimore, Maryland, speaks to the attention they can bring locally and nationally to issues like police brutality and urban poverty, but determining whether riots actually change political preferences and behavior or simply highlight existing tensions has remained difficult.

In a recent article in the American Political Science Review, Ryan D. Enos, Aaron R. Kaufman, and Melissa L. Sands find compelling evidence that violent protests have significant impact on local politics. Drawing on sample data surrounding the 1992 Los Angeles Riots, they find that individuals who lived closest to the riots’ epicentre were more likely to support liberal policies in the November election following the outbursts. Perhaps most remarkably, the researchers found that this local mobilization was no flash in the pan—it persisted more than a decade later.

The 1992 Los Angeles Riots began in response to the acquittal of four white police officers who were videotaped violently beating Rodney King, an unarmed African-American man. The videotape’s release led to outrage in predominantly African-American areas of the city, as did the decision to move the officers’ trial to predominantly white nearby Ventura County. Within hours of the all-white jury’s decision, a week of violent incidents broke out in South Central Los Angeles, resulting in 54 deaths, more than 2,300 injuries, over 11,000 arrests and material damage of over $1 billion.

Political science has long understood riots—defined as “political acts in which participants engage in violence to express grievances and attempt to spur policy change”—as politically-motivated, but pinpointing their impact on policy preferences and political activity has proven more difficult.

Enos, Kaufman, and Sands ask whether the Los Angeles Riots had an impact on political preferences or behaviour. They tackle this question by comparing support for parallel ballot initiatives that occurred before and after the Los Angeles riots. They compare changes in support for local public-school funding—an issue intimately tied to the rioters’ grievances around local economic and social conditions—with changes in support for university funding, which appeals less directly to the riots’ central issues. This comparison allows them to control for unrelated changes and isolate the riots’ true impact.

In their initial analysis, the researchers demonstrated not only that the riots led to a significant increase in voter registration among those closest to the riots’ epicentre, but also that these voters were more likely to support increased funding for local public schools. This increase in support was significant for both nearby white and African-American voters, though it was more pronounced among African-Americans.

In addition to the shift the riots caused in voting on specific policy referenda, the authors also found that the riots substantially increased local voter registration and turnout, a change which persisted over time.

Voter turnout increased 10.6 percent between 1990 and 1992 in Los Angeles County, home of the riots’ epicentre, compared to an average increase of only 6 percent in California’s other largest counties. In addition to this immediate impact, by merging 1992 voter files with those from 2005, the authors were able to demonstrate that increased registration was not isolated to a single election cycle. Indeed, large majorities of those who registered in the riots’ immediate wake kept their registration and same party affiliation for more than a decade. This included 82 percent of white voters who registered as Democrats after the riots and 77 percent who registered as Republicans. Among African-Americans, a significant majority registered as Democrats after the riots, and 87 percent of them maintained this affiliation through to 2005.

Enos, Kaufman and Sands have established rigorous evidence that riots change local policy preferences in the short-term and that some of these changes persist over the long-term. Though the 1992 Los Angeles Riots were a massive and complex political event, their aggregate effect on local communities was both politically significant and robust, warranting the attention of policymakers at multiple levels of government.

At least on a local level, the authors demonstrate that riots can signify the beginning of long-term policy changes.