An Exploratory Analysis of the United States Firearm Transfer Regulatory System

The Bureau of Justice Statistics Firearm Inquiry Statistics (FIST) Program has data for every year between 1996 and 2010. The datasets contain the two primary outcome variables relevant to the proposed analysis: total number of firearm transfer and permit applications and the total number of firearm transfer and permit application denials. These are Level-1 time varying variables. The datasets also contain characteristics of states where the background checks on firearm transfer and permit applicants are conducted, such as approval system type (i.e. instant check, purchase permit, exempt carry permit, mixed permit and instant check, and other approval system), whether or not the state is an ATF-designated Brady state or Brady alternative state, and U.S. Census geographic region (i.e. Northeast, Central, South, and West) where the state is located. These are Level-2 time-invariant variables. A complete breakdown of states is provided below (Table 1).
We will also utilize the Brady Campaign's and the Law Center to Prevent Gun Violence's annual scorecard for 2007 to 2010. These scorecards rank-order states according to their strength of firearm regulation. Twenty-eight laws are in these scorecards of which we have selected 10 as relevant to the proposed study. These include 1.) whether or not the state requires a universal check for all firearms, 2.) whether or not the state requires a universal check for handguns only, 3.) whether or not a background check is conducted or a permit is required for all firearm gun show purchases, 4.) whether or not a background check is conducted or a permit is required for all handgun gun show purchases, 5.) whether or not a background check is conducted or a permit is required for all long gun gun show purchases, 6.) whether or not fingerprinting is required to obtain a permit, 7.) whether or not safety training is required to obtain a permit, 8.) whether or not the permitting agency is allowed to extend the check time period beyond 3 days, 9.) whether or not the checking process involves a law enforcement agency, and 10.) whether or not the state has a one handgun per month requirement. We will sum these categories for a total state-by-state score that can vary between 0 and 10. This is a Level-1 time-varying variable. These scorecards will be validated by a legal analyst.
Additionally, several state-level covariates will be included in the final dataset. These covariates include the total number of records in the Interstate Identification Index (III) by state between 2007 and 2010 (Bureau of Justice Statistics, 2008; 2010; 2012). It will also include the total number of records available in the NICS index by state between 2008 and 2010 (Schauffler, Strickland, Mason, Greenspan, and DeBacco, 2010; Schauffler, et al., 2011; Schauffler, Strickland, Gallegos, Greenspan, & DeBacco, 2012). The total number of records in the National Crime Information Center which covers protection orders, active felony, and misdemeanor warrants is not available by state by year from the Federal Bureau of Investigation. These are Level-1 time-varying covariates. Additionally, we have obtained state-by-state total population size (United States Census Bureau, 2010).
ca/ncf