Firearm Violence Following the Implementation of California's Gun Violence Restraining Order Law.

Importance: California's gun violence restraining order (GVRO) law, implemented beginning in 2016, allows for people at high risk of harming themselves or others with a firearm to be temporarily disarmed and prevented from purchasing firearms for 3 weeks to 1 year; many states have recently enacted similar laws. The research to date is on older and more limited risk-warrant laws. Objective: To determine whether implementation of the California GVRO law was associated with decreased rates of firearm assault or firearm self-harm in a large metropolitan county between 2016 and 2019. Design, Setting, and Participants: This serial cross-sectional study including data from 28 counties used the synthetic control method to evaluate differences in firearm violence between San Diego County and its synthetic control following implementation of the California law from 2016 to 2019. San Diego County was used as the treated unit because it issued substantially more GVROs than any other county in California during the study period. A total of 27 California counties that issued no or very few gun violence restraining orders from 2016 to 2019 and that had stable rates of firearm violence between 2005 and 2015 were included in the control pool. Data were analyzed from February 2021 to July 2021. Exposures: Implementation of the statewide GVRO law in 2016 in San Diego County. Main Outcomes and Measures: Annual rates of fatal and nonfatal firearm assault injuries and firearm self-harm injuries per 100 000 people. Results: In the study period, there were 355 GVROs in San Diego county, and a median (IQR) total of 8 (3-20) GVROs per donor county. The mean difference between the observed rate in San Diego County and the estimated rate in the synthetic San Diego County, 2016-2019, was -0.74 firearm assaults per 100 000 (-13% difference) and 0.13 firearm self-harm injuries per 100 000 (3% difference). Results from in-space placebo tests suggested that these differences cannot be distinguished from variation due to chance (pseudo-P values from a 1-sided test: P for assault = .35, P for self-harm = .67). Conclusions and Relevance: To our knowledge, this study was the first to analyze the association between GVRO implementation and firearm violence in California and the first to evaluate the association between risk-based firearm removal laws and firearm assault in any state. GVROs were not associated with reduced population-level rates of firearm violence in San Diego County, but this may change as the number of orders increases over time; the association between GVROs and firearm violence at the individual level cannot be inferred from our findings and should be the subject of future studies.