Correlation And Causality Are The Same Thing, Right?

Opponents of self-defense are becoming more desperate as they become more irrelevant. Advocates of self-defense have thoroughly crushed the claims of their ideological opposites over the years so you would think the issue would be put to rest. But it isn’t. Instead opponents of self-defense have been busily massaging data until it fits their narrative. Their latest exercise in massaging data was to look at the rate of firearm ownership and the number of officers killed per state:

Using a regression statistical analysis, the authors found that occupational homicide for law enforcement was correlated with higher rates of firearm ownership. The analysis controlled for the violent crime rate, which indicated that these higher rates of homicide couldn’t simply be attributed to more frequent violent crimes occurring in states with higher rates of gun ownership. Instead, higher rates of law enforcement homicides were associated with more frequent encounters with violent criminals and with more frequent exposure to situations where privately owned firearms were present.

However, there were limitations to this study related to the gun ownership rates. There is no standard measure of annual firearm ownership rates—while the Behavioral Risk Factor Surveillance System is widely considered to be the best measure available, questions about gun ownership were only included in the survey for three years: 2001, 2002, and 2004.

I think the first thing worth pointing out is there’s no way to know how accurate the study is because there is no standard measure of firearm ownership. The Behavioral Risk Factor Surveillance System is a survey so the answers are based on the information voluntarily divulged by participants. Firearm ownership, which the study has asked directly about, is something people are more likely to not volunteer information about.

The second thing that needs to be pointed out is that this study established a correlation:

The authors conclude that higher levels of private firearm ownership increase the likelihood that law enforcement officers will face life-threatening situations on the job. The authors state that a 10 percent increase in firearm ownership at the state level correlated to 10 additional law enforcement homicides over the 15-year period that was examined in this study.

Apparently the authors don’t understand that correlation does not imply causality. Correlation justifies further study of a phenomenon that appear related. But you shouldn’t state a conclusion based on a correlation. There are other possible explanations for a correlation between firearm ownership and the number of officers killed on the job. For example, officers being killed on the job may convince people to purchase firearms for self-defense. In that case a higher number of officer deaths could lead to a higher rate of firearm ownership.

So today’s lessons are, one, studies based on data of an unknown quality are questionable at best and, two, correlation does not imply causality.