The United States has a much higher homicide rate than just about any other highly developed country, and it also has much more civilian gun ownership. Further, within the U.S., the simplest analysis — the type journalists can do, using data that are readily available online — reveals tight correlations between states’ gun deaths and factors such as gun ownership and gun-control laws. The obvious conclusion seems to be: Get rid of the guns, get rid of the violence. It’s that simple, everyone knows it, and yet redneck conservatives in thrall to the NRA stubbornly resist handing over their weapons.
That’s a seductive line of reasoning — especially in the wake of horrifying, high-profile incidents such as the recent mass shooting in Colorado Springs — but it’s false. In fact, the United States’ overall high rate of homicide is largely explained, perhaps entirely explained, by problems unrelated to gun ownership. There are policies concerning gun ownership that could reduce homicide, but the reductions would most likely be modest.
There is actually no simple correlation between states’ homicide rates and their gun-ownership rates or gun laws. This has been shown numerous times, by different people, using different data sets. A year ago, I took state gun-ownership levels reported by the Washington Post (based on a Centers for Disease Control survey) and compared them with murder rates from the FBI: no correlation. The legal scholar Eugene Volokh has compared states’ gun laws (as rated by the anti-gun Brady Campaign) with their murder rates: no correlation. David Freddoso of the Washington Examiner, a former National Review reporter, failed to find a correlation even between gun ownership in a state and gun murders specifically, an approach that sets aside the issue of whether gun availability has an effect on non-gun crime. (Guns can deter unarmed criminals, for instance, and criminals without guns may simply switch to other weapons.)
For good measure, I recently redid my analysis with a few tweaks. Instead of relying on a single year of survey data, I averaged three years. (The CDC survey, the best available for state-level numbers, included data on gun ownership only in 2001, 2002, and 2004. Those were the years I looked at.) And instead of comparing CDC data with murder rates from a different agency, I relied on the CDC’s own estimates of death by assault in those years. Again: no correlation.
Left-leaning media outlets, from Mother Jones to National Journal, get around this absence of correlation by reporting numbers on “gun deaths” rather than gun homicides or homicides in general. More than 60 percent of gun deaths nationally are suicides, and places with higher gun ownership typically see a higher percentage of their suicides committed with a gun. Focusing on the number of gun deaths practically guarantees a finding that guns and violence go together. While it may be true that public policy should also seek to reduce suicide, it is homicide — often a dramatic mass killing — that usually prompts the media and politicians to call for gun control, and it is homicide that most influences people as they consider supporting measures to take away their fellow citizens’ access to guns.
There are large gaps among the states when it comes to homicide, with rates ranging all the way from about two to twelve per 100,000 in 2013, the most recent year of data available from the CDC. These disparities show that it’s not just guns that cause the United States to have, on average, a higher rate of homicide than other developed countries do. Not only is there no correlation between gun ownership and overall homicide within a state, but there is a strong correlation between gun homicide and non-gun homicide — suggesting that they spring from similar causes, and that some states are simply more violent than others. A closer look at demographic and geographic patterns provides some clues as to why this is.
The first major factor is race. Blacks lacked the government’s protection from violence through most of American history and even today have higher rates of homicide than other racial groups do. Despite being 13 percent of the general population and owning guns at just half the rate of whites, blacks commit about half of murders, overwhelmingly against other blacks. Drawing on recent CDC data, the website FiveThirtyEight has reported that while blacks suffer homicide at a rate of 19.4 per 100,000, the rate for non-Hispanic whites is just 2.5 — “not so much of an outlier” in the international context, FiveThirtyEight notes. To the extent that the legacy of slavery and Jim Crow affects homicide rates among black Americans, it prevents meaningful comparisons with countries that lack a comparably lamentable racial history.
Race is not the end of the story, though, because a rate of 2.5 per 100,000 would still put America solidly above most of Western Europe, where rates tend to be around one per 100,000. Further differences become apparent when, in addition to focusing on the death rates of non-Hispanic whites, we break the CDC data down by state, combining the years 2009 through 2013 to make the rates more reliable at this more local level.
Whites in 14 states face a roughly European level of violence, with an annual homicide risk of no more than 1.5 per 100,000. Confusingly, these states don’t appear to have much in common culturally. They include liberal states known for gang violence in poor minority areas (New York, New Jersey), mostly rural red states teeming with firearms (Idaho, North Dakota), Upper Midwest states with strong hunting cultures but also large urban areas (Wisconsin, Minnesota), and gun-loving states either purple (New Hampshire) or solidly blue (Vermont).
But there is, in fact, a common thread: They are all in the northern United States. Cognitive psychologist Steven Pinker, documenting this pattern in The Better Angels of Our Nature, ascribes differences in state violence to a North–South divide reflecting “historical routes of migration.” He notes that “the Chesapeake colonies of Maryland and Virginia started out more violent than New England.” As Pinker and countless others have discussed, inhabitants of the North were relatively quick to establish the rule of law and allow the government a monopoly on violence, while the South developed a “culture of honor” in which individuals took personal slights seriously and handled their disputes themselves, sometimes resorting to violence.
This difference suggests that there will be higher homicide rates in the South, regardless of the prevalence of guns. In fact, whites in the most violent states — mainly in the South and Southwest, where gun ownership is high — have non-gun-assault death rates around 1.5 to 2 per 100,000, enough to put them above the total rates of the least violent foreign nations and of white Americans in peaceful northern states. Similarly, to return to the previous topic, blacks nationwide have a rate of non-gun-assault death above 3.5 per 100,000.
Of course, it’s possible that America’s gun laws (which are permissive compared with those in Europe), while not fundamentally driving these broad patterns, are dumping gasoline onto racial and cultural fires. In one theory, for example, the Second Amendment is basically tangled up with everything else that’s wrong with America. As the historian Jill Lepore once summarized it, in reviewing Pieter Spierenburg’s A History of Murder:
By the time European states became democracies, the populace had accepted the authority of the state. But the American Revolution happened before Americans had got used to the idea of a state monopoly on force. Americans therefore preserved for themselves not only the right to bear arms — rather than yielding that right to a strong central government — but also medieval manners: impulsiveness, crudeness, and fidelity to a culture of honor. We’re backward, in other words, because we became free before we learned how to control ourselves.
To assess more nuanced arguments, such as this one, we need an analysis that takes numerous variables into account and enables us to isolate the effect we’re looking for. Do such analyses reliably find that, all else equal, an unarmed population is a safer one? Or do guns not make much of a difference — which might be the case if, for example, civilian guns deterred as much crime as they enabled, or if criminals without ready access to guns simply used other weapons instead?
You’d certainly get the impression that guns increase violence from reading public-health academic journals, which have been churning out findings along those lines for decades. Some of these studies are well done and worthy of consideration. But when we look beyond public-health studies and also consider the findings of economists and criminologists, the picture becomes a lot blurrier.
Recently, the criminologist Gary Kleck reviewed about 40 studies that probed the link between gun ownership and crime. These compared a wide range of geographies, from cities to entire nations. Half of the findings indicated a link tying guns to higher homicide rates; the other half found none. But the more rigorous a study was, the less likely it was to find such a correlation. “It must be tentatively concluded that higher gun ownership rates do not cause higher crime rates, including homicide rates,” Kleck wrote.
Despite the impression left by a spate of media reports earlier this year, this finding is consistent with the experience of Australia, which confiscated about 20 percent of civilian-owned guns almost immediately following the passage of a 1996 law. This policy may have reduced suicide rates somewhat, but even the analysis of the program most often cited by liberals (from Australian economists Andrew Leigh and Christine Neill) is inconclusive as to its effect on firearm homicides, to say nothing of total homicides — and the basic homicide trend lines provided by the Australian government certainly don’t reveal anything obvious. Similarly, in the U.K., the homicide rate actually rose shortly after the country enacted stricter gun controls, though it has since come back down.
It is very possible, then, that disarming the citizenry would do nothing at all to decrease homicide. Disarmament is also the policy least likely to garner political support in America, not to mention pass muster in any court honestly interpreting the Second Amendment.
A much more promising idea is to focus on keeping guns out of the hands of criminals. Under current federal law, it is illegal for anyone to sell a gun knowing that the buyer is prohibited from having it. (Prohibited categories include felons, those convicted of certain domestic-violence misdemeanors, someone who is “an unlawful user of or addicted to a controlled substance,” and those who’ve been involuntarily committed to a mental institution.) On top of that, licensed gun dealers are required to run background checks before making sales, and to keep records of sales for at least 20 years, but there are no such requirements for a private citizen selling a gun to a stranger. This is often called the “gun-show loophole,” though it is not a loophole and has nothing to do with gun shows: It’s simply what the law says, and the distinction between dealers and non-dealers applies whether the sale takes place at a gun show or not.
There are many proposals to reform this system. The most commonsensical ones focus on domestic violence and mental illness. Hillary Clinton, for example, would expand the list of misdemeanors that legally disqualify people from gun ownership to include domestic violence against a non-cohabiting partner, as well as stalking. A related proposal, heavily promoted by the Center for American Progress and enacted in some states, would strip gun rights from those under a temporary restraining order, which raises some due-process concerns because such orders are issued without a hearing. Mental-health records, meanwhile, are a disaster, with the names of too many dangerous people left out of the system that background checks rely on, a problem that President Obama has begun to address with an executive order. Mental-health reforms are especially promising when it comes to preventing suicides and mass shootings.
These ideas, while important, do not address the core of the gun-violence problem: male-on-male acts of aggression, largely by men with criminal records, sometimes quite extensive ones. Few career criminals get their guns from licensed dealers, where a background check would be run and records kept. Inmate surveys indicate that criminals overwhelmingly get their guns through social connections instead.
A recent study of 99 criminals facing weapons charges in the Chicago area revealed why they were more likely to have acquired guns through social connections than in private deals with strangers. Illicit gun sellers avoid doing business with strangers because they worry about being caught in a sting operation; buyers, meanwhile, are fearful of being sold a gun that has been used in a crime. Most of the guns that criminals use are old and change hands repeatedly, sometimes being borrowed rather than sold. However, legally purchased guns often quickly make their way to criminals; several inmates reported that gun suppliers bought weapons in stores, reported them stolen, and resold them.
These findings are a reality check. There are many, many guns already in circulation in the United States, and it would be incredibly difficult to impose background checks on criminal social networks. Our best hope is to stop guns from being diverted to the criminal market to begin with. And while background checks cannot stop people from reporting guns stolen that actually haven’t been, they do seem to have an effect in the handful of states that have enacted them: Gangs are often forced to bring in guns from states without such restrictions. Criminals will not simply give up if background checks go nationwide, of course, but this does suggest that the policy affects behavior on the margins.
Polls indicate overwhelming support for universal background checks — with a majority in favor even among Republicans, gun owners, and NRA members. But the devil is in the details, and proposals face staunch opposition from NRA leaders. In the wake of the massacre at Sandy Hook Elementary School, even a relatively modest bill from Senators Pat Toomey and Joe Manchin fell flat on its face. (The authors are reportedly looking to bring the legislation back.) Here are the major tradeoffs such a scheme poses:
Which gun transfers require checks? Manchin-Toomey would have been limited to sales that took place at gun shows or were advertised online or in a publication — precisely the sales that criminals are already avoiding. A requirement of background checks in a broader swath of transfers, while difficult to enforce in real time and harder to sell politically, would aid the prosecution of those whose guns wind up in the hands of criminals. Under this approach, the simple act of transferring a gun without a check would be a crime; by contrast, today, prosecutors need to prove that someone deliberately sold to a prohibited person or served as a “straw purchaser,” passing a background check and buying a gun on someone else’s behalf. Other tricky issues include how to permit inheritances, gifts, and the temporary borrowing of guns among family and friends without creating loopholes or making criminals of innocent people. (For a detailed discussion of how this might be achieved, search online for David Kopel’s new law-review article “Background Checks for Firearms Sales and Loans: Law, History, and Policy.”)
Are records kept? Under Manchin-Toomey, gun dealers conducting checks on behalf of private sellers would have kept records the same way they do when selling their own inventory — but this was a major point of contention during the debate in Congress. Without records of these sales, police would have less of a paper trail to follow. The objection to keeping such records is that they constitute a de facto gun registry, albeit a scattered one, that the government could use to track down and confiscate legally purchased firearms. This seems rather unlikely, but, in fairness, gun confiscation has happened in other modern Anglosphere nations: Australia, the U.K., and Canada.
Who keeps the records? Currently, records are spread through countless gun stores, and getting to them involves tracing the gun through manufacturers, importers, and wholesalers. When stores go out of business, their records are transferred to the federal government, which must not organize them into a searchable database because doing so would create a registry (which would be illegal under current federal law). All this would become even more complicated with private sales in the mix — different sales of a gun would be documented at different stores, and the gun’s various owners would need to point police to each new dealer. One solution would be to have gun manufacturers, rather than stores, keep track of who has purchased their guns: If the gun is a Ruger, Ruger has its records. Another would be for stores to report the serial numbers of the guns they’ve sold, but not the buyers’ information, to a searchable federal database. This would quickly point investigators to the correct gun store without compromising the privacy of those not under investigation. (It’s similar to a requirement already in place regarding used guns sold by dealers who are considered high-risk.) Of course, these approaches would heighten concerns about a registry.
Who pays? Manchin-Toomey would have allowed those conducting the checks to charge fees. If the government decides to require these checks of people exercising a constitutional right, perhaps it should pick up the tab.
In my opinion, a background check — provided it’s free and convenient — is not too much to ask of someone buying or selling a gun, and I am not particularly troubled by the prospect of a “registry” held by private businesses. But background checks will not be the ultimate solution to our homicide problem. They may not even be worth the considerable expense.
America has a high rate of homicide relative to similar countries, and reducing that rate is a laudable and pressing goal. But preventing violence is a far more difficult and complicated task than many would have us think.