State and Local Prevalence of Firearms Ownership Measurement, Structure, and Trends

• NCJ Number: 205033
• Journal: Journal of Quantitative Criminology Volume: 20 Issue: 1 Dated: March 2004 Pages: 43-62
• Author(s): Deborah Azrael; Philip J. Cook; Matthew Miller
• Editor(s): David McDowall
• Date Published: March 2004
• Length: 20 pages
• Publication Format: Article
• Publication Type: Report (Study/Research)
• Language: English
• Country: United States of America

Annotation

In determining the best readily computed proxies for analyzing the influence of gun prevalence on gun use in criminal violence and suicide, this paper analyzes the validity of several possible proxies for gun prevalence for which the data are readily available at the State and county level over a number of years.

Abstract

The use of a good proxy that is consistently available at the desired level of aggregation is seen as having potential in analyzing the effects of gun prevalence. This paper analyzes several plausible proxies for gun prevalence for which the data are readily available at the State and county level over a number of years. It was found that among these proxies was a simple measure that performed as well or better than the rest; namely, the percentage of suicides committed with a firearm. This paper validates this proxy measure, firearms suicides divided by suicides (FS/S), 1995-1997, and then uses it to describe the geographic structure of gun prevalence, documenting the wide differences among States and the remarkable stability of these differences over recent decades. It also demonstrates that this stable structure is becoming “flatter” over time, with a trend toward greater geographic homogeneity. FS/S, the percentage of suicides committed with a gun performed consistently better that the others in cross-section comparisons. It readily computed for States and counties and had a high degree of validity when tested against survey-based estimates. However, even though the FS/S proxy performed well in the cross-section variation, there was no guarantee that it would also perform well in a time-series analysis. It was found that it did correlate with inter-temporal variation in regional gun prevalence over the period 1980 to 1998. In conclusion, FS/S provides the best of the readily computed proxies for analyzing the influence of gun prevalence on gun use in criminal violence and suicide. Appendix and references