Gun Show Loopholes And Social Science Caveats

Brian Doherty cites a recent academic study examining the impact of gun shows on gun-related deaths. While the study's conclusion, according to the press release, matches my ideological biases and expectations - "The absence of gun show regulations does not increase the number of gun-related deaths as proponents of these regulations suggest" - the methodology used to reach this conclusion has two pretty glaring loopholes/caveats:

The study focused on the geographic areas surrounding the gun shows, and would not capture the effect when weapons were transported more than 25 miles away. In addition, the data tracked the effects only up to four weeks after the gun shows, which would exclude later gun-related deaths.

Suppose the study had only looked at gun-related deaths within 25 feet of a gun show, and only from the time the gun show began until the time that it ended. Would anyone care if the results of this study showed no evidence of substantial increases in either gun-related homicides or suicides within 25 feet of gun shows at the time the gun shows take place? What about within one mile and one day of gun shows? Five miles and one week? 100 miles and one year?

I understand that the further out you get - in both time and space - from the source of the gun show, the more data you have to collect and process, the more expensive your study, and the more overlap you get between observational groups. So there are many reasons to keep your study limited.

But these limits seem pretty arbitrary, and thus highly susceptible to selection bias, whether intentional or not. What percentage of gun show purchases are made by customers who live within 25 miles of that gun show? What percentage of guns purchased at gun shows remain with the original customer (for how long?) and are not later resold? What percentage of gun-related deaths involve guns purchased within four weeks of that death? Unless we have good reason to believe that these percentages are particularly high (how high?), we have no good reason to think this study tells us anything useful or interesting.

Share this

I Bet You Won't See much of This Study

Interesting. For libertarians the results are irrelevant for the same reason studies of deaths outside of car dealerships or stores that sell knives are irrelevant to restricting purchases of cars and knives.It is wrong to restrict the use of legal products,some would say any products,by people who use them properly inorder to prevent others from using them improperly. It was interesting that homicide rates in Texas actually declined after gun shows. That would be consistent with the known peace keeping and crime preventative tendencies of guns in private hands.

I respectfully disagree with your caveat about time and local factors since CA has strict gun show rules and TX had no rules.The ten day delay in CA would be most relevant to near term acts,such as people going to a gunshows to get a weapon to kill or commit suicide. This was exactly what was not observed. The real complaint against gun shows is they are alleged to be places where criminals obtain guns to sell in far off locales where they are illegal to possess.The study couldn't adress this but the idea of restricting honest citizens in an ineffectual attempt to prevent criminals activity by someone else is the cause of massive loss of liberty,yet is seldom questioned by the sheep like public.