What Do We Know About Permitless Carry?

by David

The changes in states’ right-to-carry (RTC) laws over the last few decades has been truly phenomenal. In 1986, many conservative states such as Florida, North Carolina, Tennessee, and Texas were “no-issue” states, meaning that a person was not legally permitted to carry a concealed weapon in public. Now, in 2023, every state allows citizens to carry weapons in public with a permit, while 25 states do not require any permit to carry. This dramatic change in laws over the years has fueled intense academic and public debate over RTC laws and their effects on crime. More recently, since many states have been moving to permitless carry laws, a new debate has sparked: What effect do permitless carry laws have on crime rates?

The first and most important thing that I will note is that there is currently no evidence that permitless carry laws increase the rate of gun carrying among criminals or non-criminals alike. Therefore, we have no real reason to believe that there is a causal relationship between permitless carry laws and crime rates, whether the laws supposedly increase or decrease murder rates. The problem of measuring permit growth, which is important to the shall and may-issue debate, obviously does not apply here because permitless carry states don’t require permits. 

 

Firstly, an unpublished study by Adams (2022) uses the generalized synthetic control method to analyze the effect of a state enacting a permitless carry law on their murder rate. He controls for confounding variables known to be correlated with the enactment of permitless carry laws and related to the murder rate. He finds a statistically insignificant negative relationship between the two variables 10 years post-adoption (p = 0.28).  

 

Another study by Moody and Lott (2022) published in Academia Letters finds a statistically significant 14.7% decrease in murder rates following the adoption of permitless carry laws. However, Bondy, Cai, and Donohue (2022) write that the two authors ignore the parallel trends assumption and are heavily influenced by changes in murder rates in small-population states, such as Alaska and Wyoming. They write, “Moody and Lott (2022) use a panel of 2,249 state-year observations in their regression of homicide rates on permitless carry. Of these, only 18 observations have an active permitless carry regime…”

 

While I agree with Bondy et al. about the Lott and Moody estimates being pretty fragile. I also take issue with one of the claims that they make. On page 18, they make the claim that policymakers should “…consider the large body of evidence showing that RTC laws increase violent crime.” The studies that show that RTC laws increase violent crime often have serious limitations. In regards to the most measurable and serious crime, murder, Donohue et al. have yet to produce a single study that shows a clear evidence of a murder-increasing effect from RTC laws. According to the RAND analysis I talked about in my last post, the IRR for homicide in Donohue’s famous 2019 study shows a no statistically significant effect. Robert VerBruggen notes that, “...of the 32 models covering the whole time period for murder, only three generate statistically significant results. Of the 16 for violent crime, seven do” [My emphasis added]. (VerBruggen informs me that he took a stab at assessing the permitless carry research here). Even in Donohue’s recent study, which examined the effect of RTC laws on homicide rates and violent crime across US cities, they found no statistically significant effect on homicide rates; they write, “Homicide is estimated to rise by 9 percent and firearm homicide by 13 percent, and nonfirearm homicide is estimated to drop by 3.4 percent, but these coefficients are not statistically significant at conventional levels (with p-values of 0.39, 0.29, and 0.70 respectively)” [My emphasis added]. However, they do find a violent crime increasing effect, which appears to be driven by firearm-related violent crime, as one would expect. For one critique, see here.

 

As I discussed in my last post, Fridel (2020) found that shall-issue or permitless carry laws were associated with an increase in firearm homicides. However, she fails to distinguish between the two types of laws, and she doesn’t analyze the total homicide rate, which, in my opinion, is a better dependent variable. As Gary Kleck writes, “There is little doubt that where there are more guns, there are more acts of violence that, by definition, require gun possession, but on the more significant question of whether more guns cause more people to be murdered, Fridel’s research has nothing to say.”

 

One state that we now have a good amount of data on their pre- and post-permitless carry adoption homicide/crime trends is Arizona. A study by Michael Smith and Matthew Petrocelli attempted to estimate the effect of this law by comparing crime trends in Tucson, Arizona and El Paso, Texas. They choose these cities because they have similar populations, are in very similar in region, and share “a common heritage and culture that historically has been influenced by immigration and trade with Mexico.” They use the difference-in-difference method to analyze the effect in the “treatment” (Tucson) and the “control” (El Paso). Using homicide, robbery, and aggravated assault counts by quarter of the year pre- and post-treatment, authors found no change in quarterly homicides or robberies in both cities. For aggravated assaults, they found a non-statistically significant 24% increase in Tucson (t=1.682), but a statistically significant 35% increase in El Paso. This means that aggravated assaults actually increased more in El Paso after the law in Arizona than they did in Tucson, which means that any increase in Tucson, statistically significant or not, could be due to other factors besides the passage of their permitless carry law. They do another test comparing change in reported crimes at different time periods in the two cities. In every follow-up period, they find no statistically significant difference in the change in reported crime levels.

 

One potentially harmful effect of permitless carry laws is that they might increase the anxiety of police officers, who might be more inclined to shoot a civilian they see reaching for an object if that civilian lives in a state that doesn’t require them to obtain a permit before carrying a firearm in public. To find out whether these laws actually increase office-involved shootings (OIS), Doucette et al. (2022) look at a period from 2014-2020, which is a pretty small time series. Using the Augmented Synthetic Control + Fixed Effects model (ASCM+), they find that a state adopting permitless carry leads to 1.56 people per million per year being shot by the police, a 12.9% increase. They say that this is consistent with other studies showing that there are more police killings in states with greater firearm prevalence. Specifically, they mention Nagin (2020). who found that a 1% increase in gun ownership increased the rate of fatal police shootings by 0.5%. 

 

The last thing I will note is that the RAND review had a section on permitless carry and homicide. All of these studies found no effect of permitless carry on total homicide rates. However, RAND chose to classify these studies as having “serious methodological problems.” 

They also looked at the research on permitless carry and mass shootings. As one would expect, none found any significant effect. They concluded that permitless carry laws had inconclusive effects on total homicide rates and mass shootings. Fridel (2020) also found no effect of permitless carry laws on rates of mass shootings here.