New study evaluates validity of proxy variables for county-level firearm ownership

New study evaluates validity of proxy variables for county-level firearm ownership

A newly published study evaluates how well proxy variables for firearm ownership used in county-level studies measure firearm ownership.

By: Alexandra de Leon Date: April 28th, 2021

A newly published study evaluates how well proxy variables for firearm ownership used in county-level studies measure gun ownership.

Due to legislative restrictions and infrastructure limitations, accurate and updated data on firearm ownership at the state and county level do not exist. As a result, researchers must rely on proxy variables when they wish to evaluate firearm ownership in their research.

Prior research has identified valid proxies at the state level, but limited research exists on the validity of county-level proxy variables, especially among rural counties.

This study was published online this month in the Journal of Preventive Medicine.

“In the absence of well-maintained firearm registries, it is essential for researchers to have a reliable proxy variable for firearm ownership to understand how the policies and programs they are evaluating are impacting firearm ownership. This is true whether we are evaluating firearm ownership as an outcome or a covariate in our studies.” says Miriam Haviland, epidemiologist at UW Medicine’s Harborview Injury Prevention & Research Center.

Using Behavioral Risk Factor Surveillance System (BRFSS) data from 2013-2018, researchers were able to obtain information on household firearm ownership for 305 out of the 3141 counties in the U.S. They compared proxy variables for firearm ownership used in county-level studies against this BRFSS estimate for firearm ownership.

This study found that of the county-level proxies available, the five-year aggregated per capita rate of federal firearm licenses had the strongest correlation with the BRFSS estimate of firearm ownership; however, none of the proxies was very strongly correlated with the BRFSS estimate. Findings from this study indicate that identification or development of a robust proxy variable for firearm ownership—especially among rural counties—is warranted.

This study is part of HIPRC’s Firearm Injury & Policy Research Program. The program’s mission is to reduce the impact of firearm injury and death on people’s lives through interdisciplinary research and collaboration with institutional, community, and governmental partners.

Other authors on this work include, Emma Gause, MS, MA, Harborview Injury Prevention & Research Center, Frederick Rivara, MD, MPH, Firearm Injury & Policy Research Program (FIPRP) Director and Department of Pediatrics with the University of Washington School of Medicine, Andrew G. Bowen, Harborview Injury Prevention & Research Center, Amelia Hanron,MPH, MLS Harborview Injury Prevention & Research Center and Ali Rowhani-Rahbar, MD, PhD, MPH, FIPRP Co-Director and Associate Professor of Epidemiology at the University of Washington School of Public Health.

This work was funded by the State of Washington.