ABSTRACT
Despite the devastating effects of firearm violence on individuals, families, and communities, research on the impact of the COVID-19 pandemic on firearm violence remains at a minimum. Our study contributes to this body of research by estimating the impact of two critical pandemic era timeframes on fatal and nonfatal shooting victimizations in Detroit, Michigan, using an innovative Bayesian Structural Time Series methodology. For each timeframe, we consider the impact of the pandemic era on total shooting victimizations, shooting victimizations that occurred at a residence (or at home), and shooting victimizations that occurred elsewhere. Our findings suggest that the pandemic era contributed to all three types of shooting victimizations in Detroit. We discuss the limitations of our study, along with directions for future research. Overall, we believe that our study underscores the importance of adopting a comprehensive and evidence-based strategy to prevent firearm-related fatalities and injuries.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1. In the discussions that follow, we primarily draw from research that was conducted prior to the pandemic in order to better understand the potential mechanisms through which the pandemic may have affected firearm violence. While our arguments are grounded in empirical evidence, we recognize the possibility that these mechanisms may operate differently during the pandemic era.
2. Government stimulus efforts have attempted to lessen the economic blow of the pandemic. However, the effectiveness of these efforts remains to be seen, with some early evidence suggesting the government’s response to be inadequate (Moffitt and Ziliak Citation2020; U.S. Census Bureau Citation2021).
3. Our decision to capture weekly instead of daily counts of fatal and non-fatal shooting victimizations was influenced by their forecastability. Daily counts are more likely to capture low volume, noisy data that are more difficult to forecast. This is especially the case for rare events, such as shooting victimizations. Prior research also suggests that models that include trend and seasonality components, such as our own, will likely explain more data variability for weekly measures than for daily measures (see Alarcon Falconi et al. Citation2020).