The Changing Nature of Homicide in the Twin Cities: An Interrupted Time Series Analysis of the Police Murder of George Floyd

This study tests whether the 2020 police murder of George Floyd produced measurable changes in the quantity and qualities of homicide in Hennepin and Ramsey Counties, Minnesota—the epicenter of the ensuing social unrest—and whether these effects were moderated by neighborhood disadvantage or mediated by changes in police activity. We constructed a weekly panel of 472 census tracts from 2018–2022 ( N = 123,192) using homicide data from the Minnesota Bureau of Criminal Apprehension. Interrupted time-series (ITS) models estimate post-event effects, controlling for COVID-19 caseloads, policy interventions, mobility, weather, and autoregressive structure. Random-coefficient ITS models assess spatial heterogeneity; a causal mediation sub-analysis of Minneapolis evaluates the contribution of de-policing to post-event homicide changes. The homicide rate increased 2.5-fold—from 0.11 to 0.28 per 100,000 residents—immediately after Mr. Floyd’s murder and remained elevated for 136 weeks, yielding an estimated 183 excess homicides. This effect persisted after adjustment for pandemic covariates. Argument-related killings showed the strongest post-event increase, followed by felony-crime homicides; domestic cases were unchanged. Adult perpetration and victimization rose significantly, while juvenile-involved homicides did not. The post-murder effect varied across space, with larger increases in tracts characterized by higher concentrated disadvantage. In Minneapolis, reduced police stops mediated 25% of the total effect. The police murder of George Floyd triggered a sustained escalation in homicide in the Twin Cities, driven by situational and interpersonal violence concentrated in disadvantaged neighborhoods. De-policing explained part of the increase. Findings highlight how high-profile police violence can generate collateral criminogenic effects and reinforce spatial inequality.