States that give concealed-weapons licenses to anyone who meets basic criteria have higher homicide rates than those that require people to demonstrate good character or a real need to walk around armed, according to a new study. Public health scholars at Boston University found that states with “shall-issue” concealed-carry licensing standards had homicide rates 6.5 percent higher than states with “may-issue” standards.

Shall-issue refers to licensing standards that require authorities to provide a permit to carry a gun in public to anyone who meets certain minimum criteria, such as a clean criminal record and completion of firearms training. May-issue jurisdictions allow police to exercise discretion over whom to issue a permit, based on evidence of the applicant’s character or justifiable need for a gun.

The increase in homicide rates was even more pronounced when the authors looked at only killings committed with firearms, which were 8.6 percent higher in shall-issue states. When the researchers looked only at handgun homicides, distinct from killings committed with long guns or weapons other than firearms, they found an even bigger rise in homicide rates, to the tune of 10.6 percent.

Notably, the models used did not find statistically significant increases in non-firearm homicides or killings involving long guns, which typically cannot be concealed. That suggests the elevated levels of violent death resulted from more people carrying concealed weapons in public, rather than a broader increase in crime.

“That’s what’s novel about our study,” said Michael Siegel, one of the paper’s authors. “We disaggregated the type of weapon used. To our knowledge, that hadn’t been done before.”

Siegel and his co-authors examined 25 years worth of state-level injury data collected by the Centers for Disease Control and Prevention’s Vital Statistics registry of death certificates. They also used the FBI’s Supplementary Homicide Reports to its Uniform Crime Reports data, which specify the type of weapon used.

Another study published this summer by Stanford University legal scholars, led by Professor John Donohue, found that individual states that passed shall-issue concealed-carry laws had elevated levels of overall violent crime compared to those that retained may-issue standards. In contrast to the Boston University group’s results, however, Donohue did not find a consistent, statistically significant relationship between shall-issue standards and homicide across all statistical models used.

The difference comes down to the underlying data cited. As Donohue wrote in a response to the BU paper, the research showing elevated levels of homicide used injury data, not crime data. As a result, the BU researchers would have included justifiable homicides or those otherwise not recognized as crimes.

But as Phil Cook, a professor of public policy at Duke University, wrote in an email to Donohue, which was cited in Donohue’s response to the BU study, “we want to reduce the number of intentional killings, whether criminal or resulting from legal intervention or self-defense.”