States that have enacted right-to-carry (RTC) concealed handgun laws have experienced higher rates of violent crime than states that did not adopt those laws, according to a Stanford scholar.

Right-to-carry laws are linked with higher violent crime rates according to research by Stanford Law School Professor John Donohue. (Image credit: Ron Bailey / Getty Images)

Examining decades of crime data, Stanford Law Professor John Donohue’s analysis shows that violent crime in RTC states was estimated to be 13 to 15 percent higher – over a period of 10 years – than it would have been had the state not adopted the law.

The working paper, released this week by the National Bureau of Economic Research, challenges the effectiveness of RTC laws and could have a significant impact on pending litigation between the National Rifle Association and the state of California.

Making a ‘synthetic state’

Donohue’s paper builds on the National Academies’ National Research Council’s 2004 report investigating guns and violence.  While that report debunked claims that RTC laws had been shown to reduce crime, the 16 experts on the panel were not able to definitively conclude that carrying concealed weapons had an effect – positive or negative – on violent crime. Their uncertainty was rooted in the fragility of estimates that were derived from differing statistical models applied to panel data available at the time.

“The committee found that answers to some of the most pressing questions cannot be addressed with existing data and research methods, however well designed,” the report stated.

The most convincing comparison would take two otherwise identical states and observe violent crime when one of them adopts a RTC law. Donohue and his team employed a new statistical technique that creates a “synthetic control,” which attempts to find the best possible comparison for the RTC-adopting state drawn from among other states that had no RTC law at the time.

The synthetic control approach, a research method now widely applied in economics and political science, uses an algorithm that combines crime patterns from several non-RTC states – or during the time before states adopted RTC – to create an artificial or synthetic state.

Take Texas, which passed RTC laws in 1996. Donohue’s comparison for Texas came from combining data from California – a non-RTC state – and Nebraska and Wisconsin, which hadn’t pass RTC laws at that time. By weighting the violent crime data from these three states for the period from 1986 to 1996, he produced a synthetic crime rate similar to Texas’ crime rate in the 10 years prior to adopting RTC laws.

Donohue then projected the synthetic state’s crime rate for the next 10 years and compared it against Texas’ crime rate post-RTC passage. He performed the same analysis on the 33 states that enacted RTC laws over his data period and found a strikingly consistent picture.

On average, RTC states had aggregate violent crime rates around 7 percent higher than the synthetic states five years after RTC law passage. After 10 years, the gap increased to almost 15 percent.

“All this work is based on statistical models,” Donohue said. “When the models all generate similar estimates, it increases your confidence that you have captured the true effect.”

Donohue had further reasons for that confidence. Compared to the 2004 report, he was able to study an additional 14 years of crime data and include 11 additional states that adopted RTC laws. While the earlier panel data results were sensitive to changes in the explanatory variables (incarceration, population, poverty and unemployment rates among others) used in the statistical model, such changes had little effect on the synthetic controls estimates, which again increases confidence in the estimates, Donohue said.

RTC laws increase violent crime

Donohue applied the synthetic control approach using four previously published statistical data models that had generated conflicting panel data estimates of the impact of RTC laws on violent crime. In all four cases, the synthetic control estimates showed increases in overall violent crime of 13-15 percent.

“There is not even the slightest hint in the data that RTC laws reduce overall violent crime,” Donohue stated in the paper.

To put the significance of a 15-percent increase in violent crime in perspective, the paper notes that “the average RTC state would have to double its prison population to counteract the RTC-induced increase in violent crime.”

Donohue’s team engaged in an array of different tests to ensure that the findings were sound. For example, Donohue noticed that Hawaii was included as part of a synthetic control more than any other single state. So, he re-ran the entire synthetic controls analysis while excluding Hawaii to see if there were any major changes; there weren’t. He then did the same for every other state that contributed to the synthetic controls for any of the 33 adopting states, and the resulting estimates showed very little variation: in all cases RTC laws were linked with higher violent crime rates.

“That was a comfort,” he said.

Another comfort was the increased rates of incarceration and hiring of law enforcement personnel Donohue noticed among RTC states.

“This suggested that RTC states were not simply experiencing higher crime because they decided to lock up fewer criminals and hire fewer police,” Donohue said. “The relatively greater increases in incarceration and police in RTC states implies that, if anything, our synthetic controls estimates may be understating the increase in violent crime, which was pretty persuasive to me.”

Guns and value

The debate over RTC laws comes at a crucial time for the state of California, which in April was sued by the National Rifle Association, challenging state gun control laws.

Because the heart of the case is whether there is a constitutional right to carry a gun, which would make RTC laws moot, Donohue said there is a high likelihood the case will ultimately be resolved by the U.S. Supreme Court. His paper has been included in the court filings in federal district court.

Having a gun can generate a benefit under certain circumstances and will impose costs in other circumstances, and sound policy must consider the overall magnitude of these conflicting effects, Donohue said.  RTC proponents often overlook how often gun-carrying leads to lost and stolen guns, which are then in the hands of criminals.

Moreover, one can incur all of the costs of buying and carrying a gun, only to find that a criminal attack is too sudden to effectively employ the gun defensively.  Donohue cites a 2013 report from the National Crime Victimization Survey that showed in 99.2 percent of the violent attacks in the United States, no gun is ever used defensively – despite the nearly 300 million guns in circulation in the country today.

For most Americans, said Donohue, carrying a gun to avoid a criminal attack is similar to thinking that having a weekly brain scan will save your life, without considering the potential hazardous effects.

“If we gave 300 million people a brain scan, we would save a certain number of lives,” Donohue said. “But you wouldn’t want to advocate that treatment without considering how many lives would be lost by exposing so many to radiation damage.  One needs to consider both the costs and benefits of any treatment or policy.  If the net effect of more gun carrying is that violent crime is elevated, then RTC laws seem much less appealing. This paper may have an impact in making people think differently about these issues.”

This work was supported by Stanford Law School. The paper’s co-authors are Abhay Aneja, a law student at Stanford and a graduate student in economics at the University of California, Berkeley, and Kyle Weber, a graduate student in economics at Columbia University.