This article is a collaboration between FiveThirtyEight and The Trace, a nonprofit newsroom reporting on gun violence.
In March 2020, as the first COVID-19 outbreaks rippled across the U.S., Americans flocked to gun stores. In total, civilians purchased some 19 million firearms over the next nine months — shattering every annual sales record. At the same time, shootings across the country soared, with dozens of cities setting grim records for homicides.
As the pandemic progressed, and gun sales continued to climb alongside shootings, researchers have puzzled over the connection between these two intersecting trends. Was the surge in violent crime related to the uptick in guns sold last year? We may not get a definitive answer to that question for years, but fresh data from the Bureau of Alcohol, Tobacco, Firearms and Explosives provides some of the first evidence that a relationship exists.
ATF data shows that in 2020, police recovered almost twice as many guns with a short “time-to-crime” — in this case, guns recovered within a year of their purchase — than in 2019. Law enforcement officials generally view a short time-to-crime as an indicator that a firearm was purchased with criminal intent, since a gun with a narrow window between sale and recovery is less likely to have changed hands. Altogether, more than 87,000 such guns were recovered in 2020, almost double the previous high. And almost 68,000 guns were recovered in 2020 with a time-to-crime of less than seven months (meaning they were less likely to have been purchased the prior year).
Put more plainly, thousands of guns purchased in 2020 were almost immediately used in crimes — some as soon as a day after their sale. That was the case of the 9mm Beretta pistol purchased by an Arlington man from Uncle Dan’s Pawn Shop and Jewelry in Dallas, according to police records. Officers seized the gun from its owner during a drug arrest 24 hours later. In another example, a Laredo, Texas, man assaulted his mother, then opened fire on police with his Smith & Wesson M&P 15-22 rifle in July 2020. The gun had been purchased at a Cabela’s in Ammon, Idaho, just three months prior.
“Overall, I think we can say that the gun sale surge may have contributed to a surge in crime,” said Julia Schleimer, a researcher in the Violence Prevention Research Program at the University of California, Davis, after reviewing the ATF’s data.
But it’s also possible, she added, that the increase in gun sales is not solely responsible for the increase in short time-to-crime recoveries. Schleimer suggested that better use of gun tracing technology by police departments could lead to more short time-to-crime traces, even if the number of guns recovered by law enforcement or sold to the public did not change.
Related: The U.S. Has A Lot Of Guns Involved In Crimes But Little Data On Where They Came From Read more. »
Gun tracing is a process in which police send details about a recovered firearm to the ATF, which then provides information like the weapon’s manufacturer, original retailer, date of purchase and time-to-crime, among other data. As The Trace has previously reported, the ATF has long struggled to convince local law enforcement to trace all of their recovered guns. Compliance has improved significantly over the last decade, but many agencies still fail to trace guns.
Researchers interviewed for this story cautioned that the number of guns recovered and traced by law enforcement does not always indicate the amount of gun crime in a given year. In other words, factors driving increases in the amount of short-time-crime guns in the ATF's data may be separate from the factors contributing to gun violence.
Still, no sales bump compares to 2020, when gun buying soared to unprecedented heights, Schleimer said, substantially widening the pool of recently purchased guns that could potentially turn up at crime scenes.
As our data shows, the absolute number of recovered and traced guns with a short time-to-crime increased dramatically in 2020. But since sales also surged last year, it’s logical to expect that number to rise as well. It’s helpful, then, to consider whether the rate of recovery of short-time-to-crime guns also increased. Put another way: Did a higher proportion of guns sold in 2020 wind up at crime scenes than in years past?
The data shows that while the share of short time-to-crime guns recovered and traced relative to total guns sold did increase, the size of the increase was not abnormal. In fact, it continues a trend dating back to 2013.
Historically, dips and spikes in gun sales have not strongly correlated with the number of recently purchased guns recovered by police in a given year. Daniel Webster, the director of the Center for Gun Violence Prevention and Policy at Johns Hopkins University, said that the limited data available suggests gun sales are an imperfect predictor of how many recently purchased guns police will recover: In several of the past 10 years, gun sales spiked without proportional increases in short time-to-crime recoveries; likewise, gun sales frequently dipped while recoveries were on the rise.
Webster said additional variables are likely contributing to variation in the number of short time-to-crime guns recovered and traced by police. But he noted that since researchers only have access to aggregate trace data, it’s difficult to understand what those variables might be. “Ideally, you’d want to tease apart the crimes committed by those who purchased the guns vs. guns diverted to other people,” he said. Federal law currently restricts the ATF from sharing granular trace data with the public.
Erik Longnecker, a spokesperson for the ATF, said it would be “inappropriate” for the agency to speculate on what was driving the increase in short time-to-crime recoveries.
Jim Bueermann, a former California police chief who serves as a senior fellow at the George Mason University Center for Evidence-Based Crime Policy, said that while the new data may not provide conclusive evidence of a causal relationship between gun sales and gun crime, it does signal the importance of additional exploration. “Data like this asks more questions than it answers, but this is a clarion call for criminologists to conduct research in this space.”
Second Amendment didn’t protect your right to own a gun until 2008 | FiveThirtyEight
The research we do have, though, shows that immediate booms in access to firearms almost always lead to corresponding spikes in violence.
Dr. Garen Wintemute, who directs the Violence Prevention Research Program at UC Davis, said he wasn’t surprised that the largest increase in access to firearms in history happened alongside a parallel surge in violence.
But so far, his team has struggled to find any causal association. In July, he and Schleimer co-authored a paper that sought to investigate what effect, if any, the pandemic sales boom had on violent crime. It found no relationship. Instead, they concluded that unemployment, economic disparity and physical distancing exacerbated by the pandemic were far more potent predictors of increased violence. The findings square with theories from researchers, violence prevention activists and former perpetrators about the root causes of gun violence.
Wintemute said that his findings should not undercut the significance of the new ATF data. His study only assessed state-level associations, he noted, and was published before the ATF’s release of 2020 time-to-crime data.
“It can be difficult or impossible statistically to sort out the contributions that any one of these [phenomena] made” to the rise in violence, Wintemute said. “But the bottom line is, if the prior research holds up and increases in access are associated with increases in violence, we’re in for a very rough time ahead.”
Daniel Nass contributed reporting.