In just four years in office, former Brazilian President Jair Bolsonaro introduced reforms that increased gun ownership six-fold. But on Jan. 1, on his first act in office, the newly inaugurated Luiz Inacio Lula da Silva signed an executive order banning sales of guns and ammunition and limiting people to a maximum of three guns. Lula also banned concealed carry and is moving to take away gun ownership licenses issued under Bolsonaro. Naturally, gun-control supporters are thrilled, but none of the leading academics who support these policies are willing to put their money where their mouths are and bet that murder rates will fall.
Over the last five years, many academics and journalists predicted Bolsonaro’s dramatic loosening of gun-control laws, which started in 2019, would cause murder rates to soar in the country. Now, they predict that Lula’s severe crackdown on gun ownership will reduce crime.
A Washington Post article on Brazil in December claimed: “Research consistently shows that when private gun ownership goes up, killings follow.” Back in 2018, immediately before Bolsonaro became president, outlets worldwide such as The New York Times warned: “[Bolsonaro’s policy] is worrying some experts who argue that more guns fuel more violence.”
Unique One-Country Case Study
Brazil provides a unique experiment because of how radical the changes in law are. In the U.S., the handgun bans in Chicago and Washington, D.C., provide the closest case studies. But many gun-control activists dismissed the post-ban increases in murder in those cities, arguing that the ban could only work if the entire country instituted the same rules. With Brazil, we have a country-wide case to examine.
When Bolsonaro became president on Jan. 1, 2019, Brazil had one of the highest homicide rates of any developed country.
In 2018, the year before Bolsonaro became president, the homicide rate stood at 27.8 per hundred thousand people — 5.5 times higher than the U.S. rate. The murder rate then fell in each consecutive year. By 2021, the third year of Bolsonaro’s presidency, it had dropped to 18.5 — a 34 percent drop and a rate not seen in Brazil since the early 1990s.
In December, Dan Webster of Johns Hopkins University claimed that for Brazil, “every 1 percent increase in firearm ownership is associated with a 0.6 percent increase in overall homicide rates.” If that were true, a more than 600 percent increase in gun ownership should have resulted in a more than 360 percent increase in homicides, not a 34 percent drop. Instead, researchers attributed the drop to a slight decline in the 12-to-29 age group’s population share and “a decade of investment in policing” in some parts of the country, even though murders fell in all but one of the country’s 26 states.
What Will Result from Brazil’s Gun Ban?
With the dramatic changes earlier this year in Brazil’s gun-control laws, the country again offers a good test of the more guns, less crime hypothesis. So I contacted 12 prominent gun-control advocates in academia, including Hopkins’ Webster. I offered a $1,000 bet on whether Brazil’s homicide rate would go up or down during the first two years of Lula’s presidency. If it went down, I would pay them. If it goes up, they would pay me. The ultimate test of a theory is whether it accurately predicts what will happen.
Seven of the 12 didn’t respond to my emails. Duke University’s Phil Cook wrote back a nice response, saying: “I like the idea of a bet, but am not going to take this one, since I have no confidence that guns and ammo will actually become scarcer in the neighborhoods with high rates of violence.” Indiana University’s Paul Helmke turned down the bet but wrote: “Happy to read/review any data and conclusions you come to.” University of California at Berkeley’s Frank Zimring expressed some interest, but he stopped responding after asking me to sketch out the different theories for why the homicide rate fell.
Stanford’s John Donohue insisted that we already had all the evidence we need to support gun control and that we should look at past data rather than making predictions. I pointed out that he hadn’t shied away from making many predictions in the past, even based on less dramatic changes in gun-control regulations. For example, Donohue wrote in The Washington Post last year after the Supreme Court’s New York State Rifle and Pistol Association v. Bruen decision in favor of concealed carry: “We unfortunately know what effects this ruling will have in the relatively few states that still restrict the carrying of weapons. … It will cause a spike in violent crime.”
Donohue also argued that you could only compare one place that changes its gun-control laws relative to another that didn’t (e.g., he argued we should compare past changes in Houston, Texas, to New York City). I offered to adjust the bet to compare changes in Brazil’s homicide rate to those of neighboring countries, but he declined.
UCLA’s Adam Winkler wanted changes in the bet so I wouldn’t have to pay him the money but would write an op-ed with him if I lost. However, Winkler wouldn’t have to write up a similar op-ed if he lost. I noted that it might be hard for us to agree on an op-ed’s content, but I suggested that we both agree to write an op-ed no matter who won. When I asked Winkler to explain the asymmetrical bet, instead, as an explanation, he offered a string of personal attacks saying he didn’t take the discussion seriously. I responded by noting that I had “similar” feelings about his work, but I had hoped he “would view this offer as a chance to prove that you are right and show everyone I am wrong.” He never responded.
These academics have no problem confidently making predictions for the press or legislative committees about the future effects of gun-control laws. But they aren’t willing to put their money where their mouths are in a way that would make people remember their bad predictions. Maybe that’s because they already know the crime-fighting benefits of private gun ownership.