The decline in gun homicides is part of an overall decrease in violent crime in the United States that has puzzled researchers. Wonkblog reported Thursday on five trends that likely contributed to these declines: more officers and better computers in police departments, reduced alcohol consumption, the removal of lead from gasoline and an improving economy.
In response to questions from readers, here are five more theories on what caused the gun violence decline.
1. Mass incarceration
The United States locks up more people than any other country in the world, both in sheer numbers and relative to the general population. It even has more inmates than China, an autocratic state that's three times larger than this country. And that vast population behind bars is relatively new: Incarceration increased dramatically beginning about three decades ago, just as crime was declining.
That coincidence led some of our readers to ask how the increase in the prison population affected rates of violent crime and gun homicides. It's much more difficult for criminals to commit crimes once they're behind bars. Economist Steven Levitt and journalist Stephen Dubner, in their best-selling book "Freakonomics," were among those who concluded that incarceration might have been a main driver of the decline in crime and violence.
Researchers are still investigating the relationship between crime and incarceration, but many now think these earlier studies overstated the benefits of locking people up.
Increasing incarceration accounted for about 6 percent of the decline in property crime between 1990 and 2000, according to a recent report from the Brennan Center for Justice. That report concluded that incarceration had no discernible effect on violent crime. Other recent reviews from the Brookings Institution and the National Academy of Sciences, both nonpartisan research organizations, have reached similar conclusions.
A couple of things might be limiting the effect of incarceration on violence. For one, people tend to commit fewer crimes as they get older. As prisoners age, the number of crimes they are prevented from committing declines, so locking people up for longer periods of time does not have much effect on how many crimes they commit over their lifetimes.
Further, time in prison can actually make people more likely to commit crime, if a sentence turns a youth who has broken the law into a hardened criminal with experience behind bars. One study of decades of data on the correlation between crime rates and releases from prison found that criminals leaving prison seemed to be more likely than those entering. Another study, published in 2006, concluded that increasing incarceration may have actually made crime more common, counteracting the overall decline.
2. The ban on assault weapons
Congress outlawed certain assault rifles, pistols and shotguns in 1994. Guns with components designed for use by the military in combat became illegal, including grenade launchers; mounts for bayonets; and flash suppressors, folding stocks and silencers to help keep weapons hidden. Crucially, magazines with capacity for more than 10 rounds were also banned.
Until the ban expired 10 years later, it coincided with the decline in crime, but the law is unlikely to have been a significant factor.
Large magazines might make shootings more dangerous, according to a federal report on the ban published in 2004. According to that document, most people are not very accurate with guns, hitting their targets with only about a third of the bullets they discharge. Large magazines might result in more victims in any shooting, and in more bullets striking each person who is hit. As a result, large magazines could increase the rate of injury and death from gun violence.
That said, the law passed in 1994 exempted magazines manufactured before that year, and there were about 25 million guns with such magazines already in civilian possession at the time. In other words, at a minimum, a quarter of a billion rounds in high-capacity magazines were exempted from the law. Meanwhile, after the ban took effect, U.S. buyers imported more weapons from abroad that would have been banned except that they were manufactured beforehand.
"The ban’s effects on gun violence are likely to be small at best and perhaps too small for reliable measurement," the report concluded.
3. Concealed-carry laws
While federal gun laws became more restrictive as crime rates declined, many state laws became less so. In 1990, just 15 states had laws mandating that law enforcement issue concealed-weapon permits to residents. By 2013, 38 states required that residents be able to obtain such permits.
Some economists have posited that these laws might deter criminals from committing robberies and other crimes. In states that allowed more citizens to carry concealed firearms in public, the argument goes, criminals might worry more about potential victims being armed, and they would hesitate to commit crimes. On the other hand, people carrying their guns in public might resort to deadly force more quickly in an argument with a stranger on the street, ending a dispute with a bullet when it might have ended with a few obscenities.
Studies of this question have produced conflicting results, suggesting that if the laws do have an effect on crime, it might be too small to measure accurately. People holding concealed-carry permits typically live in rural and suburban areas, where crime is rare to begin with, said David Hemenway, an economist at Harvard University. Any change in the rate at which they carry guns in public will not have a large effect on the overall crime rate either way, he said.
4. Crack cocaine
The crack epidemic that gripped urban areas in the 1980s also has been linked to gun violence and crime, but researchers haven't conclusively a causal connection.
Harvard economist Roland Fryer Jr. worked with Levitt, the author of "Freakonomics," and several collaborators to develop what they called a "crack index," incorporating the number of people arrested for using cocaine, the number of hospitalizations involving cocaine, the number of raids by the Drug Enforcement Administration, and the number of times the word "crack" appeared in newspaper articles.
Using this index, the authors estimated that crack increased the crime rate in major cities by about 5 percent between 1984 and 1989. However, crack use did not decline sharply after 1991, when the overall rate of violent crime began to decrease. Indeed, the number of deaths involving crack continued to increase at about the same rate.
One explanation the authors offer is that disputes between criminal organizations that distributed crack fueled violence in the early years of the epidemic. But as time went on, informal agreements settled those conflicts over turf and other aspects of the business.
Levitt also argued that the legalization of abortion in 1973 could have reduced crime, an intensely controversial theory. He reasoned that children born as a result of unwanted pregnancies would be more likely to commit crimes when they were older, but if they were never born, they would not be able to commit those crimes.
There are a couple of problems with the theory. First, Levitt's original paper, which he wrote with the Stanford University legal scholar John Donohue, did not consider abortions in states where the procedure was illegal before Supreme Court's decision in Roe v. Wade. Data on abortions before they were legal are hard to come by. That said, demographers estimate that about two out of three abortions after that ruling would have occurred even if abortion had remained unlawful, as Ted Joyce, an economist at Baruch College, City University of New York, noted in a critique of the theory. When Joyce focused on the number of abortions following Roe v. Wade, a period for which reliable data is available, he found no correlation with crime rates.
He also observes that the birth rate actually did not decline much following the decision. Perhaps an increase in abortions might have reflected an increase in conceptions, not a decrease in the number of unwanted pregnancies carried to term. Alternatively, the passage of Medicaid in 1965 might have made bearing and raising children less expensive for poor women, giving them less of a reason to seek abortions even after they became legal. Whatever the reason, the figures on fertility provide little evidence that a smaller group of children was born following Roe v. Wade who could grow up to become criminals.
Finally, there is the issue of timing. The rates of homicide among young people began to decline after about 1993, 20 years after the court's decision. Using data on offenders' ages from the Federal Bureau of Investigation, Joyce showed that homicide rates declined for 15 to 19 year olds at the same time as they declined for 20 to 24 year olds. If legal abortion accounted for the declines, then the rate of homicides among 15 year olds should have declined beginning sometime around 1988 -- five years earlier. Instead, homicide declined in all age groups simultaneously, suggesting that the year of birth was irrelevant. Nor did the timing of the decline in crime vary between states that legalized abortion earlier, before the court's ruling.
Overall, the most important factors in the decline in gun violence probably include better equipped and better staffed police forces, the removal of lead from gasoline and the declining consumption of alcohol. The improving economy helped, though the relationship between the economy and gun violence specifically is unclear. Incarceration is also a factor, though not as important a factor as you might expect given how many Americans are behind bars.
More from Wonkblog: