Are Politicians Responsive to Mass Shootings? Evidence from U.S. State Legislatures

On August 3, 2019, a white nationalist manifesto titled “The Inconvenient Truth” was posted on the imageboard 8chan. The author, Patrick Crusius, claimed to be defending the State of Texas from a cataclysmic “Hispanic invasion” by “remov[ing] the threat of the Hispanic voting bloc.” Roughly twenty minutes after posting manifesto, Crusius walked into a Walmart near the Cielo Vista Mall in El Paso carrying a semi-automatic AKM assault rifle and opened fire, killing 23 people and wounding another 22 in the deadliest anti-Latino attack since the 1910-1920 paroxysm of violence against Mexicans in Texas known as “La Matanza.”

Figure 1: U.S. State House and Senate Districts with Mass Shootings, 2011-2022. Red districts are those affected by a mass shooting, shown as the small, light red points at the location of the shooting. The map uses district boundaries adopted after the redistricting cycle following the 2010 Census, and only plots shootings that affected districts under these new maps (accounting for variation in timing of adoption of these maps across states). Nebraska has a unicameral state legislature, so we do not report lower chamber districts and only show upper chamber districts on the map.

Sadly, this hate-inspired mass shooting is one of many. The U.S. witnesses more mass shootings annually than any other country, many of which are manifestations of group-based hate (Peterson & Densley, 2021). Over the past few years, political scientists have conducted a string of empirical studies on hate-motivated violence, but so far this research has yielded conflicting evidence on a number of key issues. As yet there is no consensus on how mass shootings impact public opinion on gun control (Newman & Hartman, 2019; Barney & Schaffner, 2019; Rogowski & Tucker, 2019), for example, nor what collective behavioral responses these hate crimes elicit (Hassell et al, 2020) nor even how they impact state-level regulation of firearms (Luca et al, 2020). 

Our project investigates the responsiveness—or lack thereof—of state legislators to mass shootings, which includes shootings that have been identified as hate-motivated (as is the case with El Paso) and also those that have not been identified as such  (i.e. shootings that are not motivated by hate against a particular demographic or community, such as the 2018 Parkland, FL shooting).  We estimate the latent positions of 264 California state legislators on gun policy from their roll-call voting records on firearm-related bills from 2011 to 2022—a novel approach to measuring how strongly legislators support or oppose firearm-related policy. 

Figure 2: Gun Control Score of California State Senate and Assembly Members by Party, 2011-2022. Each point is an individual legislator’s score (N = 264, including legislators who switched party affiliation) for the corresponding year of the legislative session. Points are colored by the party affiliation of the legislator, and linear regression smoothers (with dotted standard errors) are fit to legislators in either party and across parties to show general trends in partisan voting behaviors across session years.

Using this scoring regime, we investigate how eleven mass shootings perpetrated in California between 2011 and 2022 affected the voting behavior of the legislators representing the districts in which the shootings occurred. Our estimated effects of mass shootings on treated legislators’ support for restrictive gun policies are statistically insignificant regardless of the legislator’s party, with 95% confidence intervals spanning opposite directions and indicating no discernible effect.

When we extend our analyses to an additional 5 states covering 1,593 legislators and 17 shootings, we find similarly null results. We conclude that, on average, mass shootings fail to produce changes in a legislator’s support for (opposition to) restrictive (permissive) firearms bills. Our findings suggest that even the most heinous acts of mass violence—that are squarely in the domain of events that state legislators might respond to—fail to produce any measurable effects on legislator’s positions on firearm-related policy.