Cold Comfort: An Interview with John R. Lott

Economist John Lott discusses the benefits of guns--and the hazards of pointing them out.


Until recently, when he bought a 9-mm Ruger after his own research impressed upon him the value of gun ownership, John Lott had no personal experience with firearms, aside from one day of riflery in summer camp when he was 12. That fact did not stop a reviewer of Lott's 1998 book, More Guns, Less Crime (University of Chicago Press), from labeling him a "gun nut." Writing in The American Prospect, Edward Cohn also identified Lott as "a leading loon of the Chicago School of economics, known for its ultra-market ideology." But that was gentle–a backhanded compliment, even–compared to the attacks from anti-gun activists, who accused Lott of producing his landmark study at the behest of the gun industry.

Lott, now a senior research scholar at Yale Law School, used to be the John M. Olin Law and Economics Fellow at the University of Chicago. That position, like similar ones at other major universities, was endowed by a foundation based on the personal fortune of the late John M. Olin, former chairman of the Olin Corporation. Among many other things, the Olin Corporation makes Winchester ammunition. These facts led Kristen Rand of the Violence Policy Center to conclude that "Lott's work was, in essence, funded by the firearms industry"–a charge that was echoed by other gun control ad-vocates, including Charles Schumer, then a Democratic representative from New York and now a senator.

Never mind that assuming the Olin Foundation takes orders from "the firearms industry" is like assuming the Ford Foundation does the bidding of automakers. Never mind that Olin fellows are chosen by faculty committees, not by the foundation (with which Lott never had any contact). Proponents of gun control were desperate to discredit Lott, because his findings contradicted their dark predictions about what would happen if states allowed law-abiding citizens to carry concealed handguns.

Analyzing 18 years of data for more than 3,000 counties, Lott found that violent crime drops significantly when states switch from discretionary permit policies, which give local officials the authority to determine who may carry a gun, to "shall issue" or "right-to-carry" laws, which require that permits be granted to everyone who meets certain objective criteria. That conclusion, first set forth in a 1997 paper that Lott co-authored with David Mustard, now an economist at the University of Georgia, heartened defenders of gun ownership and dismayed their opponents. Arguing that "shall issue" laws are beneficial, while other gun laws are ineffective at best, Lott quickly became one of the most widely cited–and reviled–scholars in the gun control debate.

Though it was the gun issue that brought Lott notoriety, it hasn't been the focus of his career. The 41-year-old economist, who earned his Ph.D. at UCLA, has published papers on a wide variety of topics, including professional licensing, criminal punishment, campaign finance, and public education. Last summer he published Are Predatory Commitments Credible? (University of Chicago Press), a skeptical look at theories of predatory pricing, and he is working on a book about the reputational penalties faced by criminals, a longstanding interest. In addition to his positions at Yale and the University of Chicago, Lott has served as chief economist at the U.S. Sentencing Commission and taught at UCLA and the University of Pennsylvania, among other schools. He lives in Swarth-more, Pennsylvania, with his wife and four children. Senior Editor Jacob Sullum and Washington Editor Michael Lynch talked to Lott at his Yale Law School office in mid-October.

Reason: How did you become interested in guns?

John R. Lott Jr.: About six years ago, I was teaching a class dealing with crime issues at the University of Pennsylvania, and it dawned on me that my students would be interested in some papers on gun control. It forced me to look at the literature systematically to decide what papers to assign to the class. I was shocked by how poorly done the existing research on guns and crime was. You had very small samples. By far the largest previous study on guns and crime had looked at just 170 cities within a single year, 1980. Most of the rest looked at, say, 24 counties or 24 cities within a single year. No one had tried to account for things like arrest rates or conviction rates or prison sentence lengths. And the studies were all very limited in the sense that they were purely cross-sectional, where you look at the crime rates across jurisdictions in one year, or [purely longitudinal], where you pick one city or one county and look at it over time.It was basically because of that class that I saw the benefit to going out and trying to do it right. So I put together what I think is by far the largest study that's ever been done on crime. The book has data on all 3,000-plus counties in the U.S. over an 18-year period. And simply having that large a data set allows you to account for hundreds of factors, thousands of factors, that you couldn't have accounted for in those smaller data sets.

Reason: What has been the most gratifying response to the book? Do you know of any criminologists whose views have been changed by your research?

Lott: Some well-known people like [University of Pennsylvania criminologist] John DiIulio and [UCLA political scientist] James Q. Wilson have said very nice things about the study. I think it's caused DiIulio to look at these issues differently, and there are other criminologists I know of who have been amazed by how strong the data are. I've done lots of empirical studies, and the regularities that you see here, in terms of the drops in violent crimes right after these laws go into effect, are very dramatic.

The intensity of the issue on both sides is something I wouldn't have expected before I got into it. I've been involved in a lot of debates, and people tell me, "You should have anticipated this before you did the study." But I've written about 80 academic articles, and the interest in this has been so outside the range of experiences I've had before. With the vast majority of articles, you're happy if you can get 10 people to read it.

Reason: The thrust of your argument in More Guns, Less Crime is easy enough to understand. But the details of the evidence you cite are hard to follow for anyone who is not trained in econometrics. Does it bother you that people who support the right to keep and bear arms are apt to accept your conclusions at face value, while those who are inclined to support gun control will tend to reject your findings, even though few people in either group are equipped to evaluate the evidence?

Lott: My guess is that [my critics] assume that the vast majority of people who hear their claims are not going to even look at the book. So they say, "Lott didn't account for poverty." Or they say, "Lott didn't account for other types of gun laws." Those are things that are easy to evaluate: Either I did, or I didn't. But I think they feel that they can get away with making those claims, because it'll be only a tiny fraction of 1 percent who will go and buy the book or get it from the library. I've never been involved in a debate like this, because in your normal academic debate, where there are 10 people involved and they've all read the paper, if somebody says, "Professor X didn't account for other gun laws," everybody else in the room would laugh, because they would know it was an absurd claim.

I don't think that most of the comments [the critics] are making are really that difficult to understand. One of the claims, for instance, is that I'm assuming that when these laws are passed there will be a one-time drop in violent crime rates, and it should be the same across all places that adopt these laws. That's absurd. I don't know how much time I spend in the book saying that the level of deterrence is related, according to the data, to the probability that people are going to be able to defend themselves, and the rate at which people get permits changes over time. When you pass these laws, not everybody who eventually is going to get a permit does it the first day. Fifteen years after these laws go into effect, you're still seeing an increasing percentage of the population getting these permits and a decreasing rate of violent crime because of the additional deterrence.

I spend lots of time in the book talking about why you don't expect the drop in crime to be the same in all places….In more urban areas [of states with discretionary permit laws], public officials were especially reluctant to issue permits. So when you change to a nondiscretionary rule, the biggest increases in permits tended to be in these urban areas, and that is where you observe the biggest drops in violent crime.

Reason: Your analysis shows that liberal carry permit policies are associated with lower crime rates even after controlling for a variety of factors that might also have an impact on crime. In the book you concede that some other variable that you did not consider could be responsible for this association. Yet at the end of the book, you write, "Will allowing law-abiding citizens to carry concealed handguns save lives? The answer is yes, it will." Do statements like that go too far?

Lott: I don't think so. That's one of the last sentences in the book, and at that point the evidence is pretty overwhelming. There are different types of information, and they're all pointing in the same direction.

After these laws are adopted, you see a drop in violent crime, and it continues over time as the percentage of the population with permits increases. If I look at neighboring counties on either side of a state border, when one state passes its right-to-carry law, I see a drop in violent crime in that county, but the other county, right across the state border, in a state without a right-to-carry law, sees an increase in its violent crime rate. You try to control for differences in the legal system, arrest and conviction rates, different types of laws, demographics, poverty, drug prices –all sorts of things. You look at something like that, and I think it's pretty hard to come up with some other explanation. I think you're seeing some criminals move [across the state line].

You find the types of people who benefit the most from these laws. The biggest drops in crime are among women and the elderly, who are physically weaker, and in the high-crime, relatively poor areas where people are most vulnerable.

There are five or six things that one could point to that confirm different parts of the theory. I haven't heard anybody come up with a story that explains all these different pieces of evidence….Since you have all these states changing their laws at different times, it becomes harder and harder to think of some left-out factor that just happened to be changing in all these different states at the same time the right-to-carry law got changed.

Reason: A review of your book in The American Prospect claims that "his results are skewed by the inclusion of data from tiny counties with trivial rates of violent crime. In fact, when you consider only large counties and exclude Florida from the sample, his case completely falls apart." How do you respond?

Lott: When you drop out counties with fewer than 100,000 people, if anything it actually increases the size of the effect. What [the reviewer is] saying is that if you not only drop out counties with fewer than 100,000 people–which is 86 percent of the counties in the sample, so it's not just a few small counties that we're talking about–but also drop out Florida, then the changes in two of the violent crime categories, when you're just looking at the simple before-and-after averages, aren't statistically significant. But the results still imply a drop, and for robberies and aggravated assaults you still get a drop that's statistically significant.

Now, I think it's somewhat misleading to look only at the simple before-and-after averages. Take the case where violent crime rates are rising right up to the point when the law goes into effect and falling afterward, and let's say it was a perfectly symmetrical inverted V. If I were to take the average crime rate before the law goes into effect and the average afterward, where the point of the V is when the law changed, they're going to be the same. Does that mean the law had no impact? When you drop Florida from the sample, [the results] look more like this inverted V than they do when Florida is in there. So I would argue that it strengthens the results, if what you care about is the change in direction.

In any case, the bottom line to me is this: I wanted all the data that were available….I didn't pick and choose, and when somebody drops out 86 percent of the counties along with Florida, you know they must have tried all sorts of combinations. This wasn't the first obvious combination that sprang to mind. And it's the only combination they report….If, after doing all these gymnastics, and recording only one type of specification, dealing with before-and-after averages that are biased against finding a benefit, they still find only benefits, and no cost, to me that strengthens the results.

Reason: University of Florida criminologist Gary Kleck recently told The Salt Lake Tribune that "Lott has convincingly demonstrated there is no substantial detriment" from "shall issue" laws. But he questioned whether these laws could have as substantial a deterrent effect as you suggest. Kleck provided a blurb for your book, and his work is often cited by opponents of gun control. Why do you think he has trouble buying your conclusions?

Lott: Gary has had a strong opinion for a long time that, on net, guns neither reduce or increase crime. He thinks it's essentially a wash. And I'm not sure I understand how he comes to that conclusion, particularly given the survey data that indicate that many more violent crimes are stopped with guns than are perpetrated with guns. But it is something that he has written and felt strongly about for a long time. Now Gary may think that there's something else that's being left out that maybe could explain these changes in crime rates. If he can tell me what that factor is, I'd be happy to try to test it.

Reason: Do you still hear the argument that you're in the pay of the gun industry, or has that been discredited?

Lott: I think the gun control people are going to continue to bring it up. I've been in debates this year with people from Handgun Control Inc. and other gun control groups in which they asserted flat-out that I've been paid by gun makers to do this study.

Reason: When they raise this charge, how successful are you in making the point that people should be able to assess evidence and arguments on their merits and that your motives don't matter?

Lott: Well, most people aren't going to look at the data. They're not going to have the data in front of them. The credibility of the data and the message depends on whether or not they believe that the person who's telling them about the data is credible. And I think the gun control groups feel that it's a win to the extent that they even divert three minutes of a show to talking about this issue. Even if it doesn't stick in people's minds, it's still three minutes that I couldn't talk about something else.

Reason: In a working paper you wrote with University of Chicago law professor William Landes [available at papers.ssrn.com/paper.taf?abstract_id=161637], you conclude that "shall issue" laws are especially effective at preventing mass public shootings. Given that the people who commit these crimes seem to be pretty unbalanced, if not suicidal, how does the deterrent work?

Lott: Most of these attacks do end in the death of the attackers themselves, frequently from suicide, but also because they're killed by others. But part of what's motivating them is the desire to harm other people, and to the extent that you can take that away from them, I think you reduce their incentive to engage in these attacks. Whether they do it just because they intrinsically value killing people or whether they do it because of the publicity, the fact that there might be a citizen there who can stop them well before the police are able to arrive takes away, in their warped minds, some of the gain from the crime, and stops a lot of them from doing it.

Reason: You often say, based on surveys, that Americans use guns to fend off criminals more than 2 million times a year. But in the book, you note that people who report incidents of armed self-defense could be mistaken or lying. How big a problem is that, and how confident can we be that the true number is more than 2 million?

Lott: Well, 2 million is the average of the various surveys. Different problems may plague different surveys, and the problems can go in both directions. You may have questions that weed out people who shouldn't be weeded out.

Reason: Like "Do you own a gun?"

Lott: Or it could be you ask them "Has a crime been committed against you?" before you ask them whether they've used a gun defensively.

Reason: And they might not consider it a crime if it wasn't completed?

Lott: Right. And so, we have errors that can exist on both sides….But that's the only type of evidence that we have on this….The most striking thing to me is the comparison between the results from these surveys and [survey data on] the number of violent crimes that are committed with guns each year. You see many more crimes that are averted by people who defend themselves with guns. I think that difference–even though both sets of numbers can be tainted for all the same reasons–is what's striking.

Reason: You say that resistance with a gun is the safest option when confronted by a criminal. What's the basis for that conclusion?

Lott: You hear claims from time to time that people should behave passively when they're confronted by a criminal. And if you push people on that, they'll refer to something called the National Crime Victimization Survey, a government project that surveys about 50,000 households each year. If you compare passive behavior to all forms of active resistance lumped together, passive behavior is indeed slightly safer than active resistance. But that's very misleading, because under the heading of active resistance you're lumping together things like using your fist, yelling and screaming, running away, using Mace, a baseball bat, a knife, or a gun. Some of those actions are indeed much more dangerous than passive behavior. But some are much safer.

For a woman, for example, by far the most dangerous course of action to take when she's confronted by a criminal is to use her fists. The reason is pretty simple: You're almost always talking about a male criminal doing the attacking, so in the case of a female victim there's a large strength differential. And for a woman to use her fists is very likely to result in a physical response from the attacker and a high probability of serious injury or death to the woman. For women, by far the safest course of action is to have a gun. A woman who behaves passively is 2.5 times as likely to end up being seriously injured as a woman who has a gun.

Reason: Why does the mainstream press seem to downplay the value of armed self-defense?

Lott: One question is, Why don't they report people using guns defensively? If I have two stories, one where there's a dead body on the ground vs. another where, say, a woman has brandished a gun and a would-be rapist or murderer has run away, with no shots fired and no dead body on the ground, it's pretty obvious to me which one of those is going to be considered more newsworthy. It doesn't require any conspiracy. Now if we care about policy, if we care about what types of actions are going to save the most lives, or prevent the most crimes, we want to look at both of those cases: not only the newsworthy bad events but the bad events that never become newsworthy because they don't occur.

But I don't think that explains everything. One example is gun deaths involving children. My guess is that if you go out and ask people, how many gun deaths involve children under age 5, or under age 10, in the United States, they're going to say thousands. When you tell them that in 1996 there were 17 gun deaths for children under age 5 in the United States and 44 for children under age 10, they're just astounded. There's a reason why they believe these deaths occur much more frequently: If you have a gun death in the home involving a child under age 5, you're going to get national news coverage. Five times more children drown in bathtubs; more than twice as many drown in five-gallon water buckets around the home. But those deaths do not get national news coverage.

This type of news coverage has consequences, because it affects people's perceptions of the benefits and costs of having guns around. Concentrating on gun deaths in the home, exaggerating the risks of that, creates a false impression. People are going to die because of that false impression. They're not going to have guns in the home, even though that's by far the safest course of action for them to take when they're confronted by a criminal. You may prevent some of the accidental deaths, but you're going to create other types of deaths because people won't be able to defend themselves.

I think the debate would be so different now if, even once in a while, some of the life-saving uses of guns got some attention in the news. A couple of the public school shootings were stopped by citizens armed with guns well before the police were able to arrive. Or take the case of the day trader shooting in Atlanta, which got huge attention. Within 10 days after that, there were three separate attacks in the Atlanta area that were stopped by citizens with guns, in two cases permitted concealed handguns. They got no attention outside of the local media market.

Reason: You've said that if Chicago Mayor Richard Daley, who filed one of the first city-sponsored gun lawsuits, really believes guns are so bad, he ought to take them away from his bodyguards. Explain that comment.

Lott: Daley has been arguing that there's no benefit from owning guns. Yet he has a whole set of full-time bodyguards following him every place he goes. He won't even think about visiting some of the more dangerous areas in Chicago without his bodyguards. But there are poor people who have to live in those areas, who live there at great risk, and he's not willing to let them own guns in order to protect themselves….I view it as very hypocritical, that Daley can understand the defensive benefits of guns when it comes to himself, but he's not willing to afford that same level of protection to the poorest, most vulnerable people in his city.

Reason: You've pointed out that somebody who gets turned down for a gun purchase after a background check may simply get the gun by other means. That's a legitimate point, but don't you also have to consider the possibility that people are deterred from even trying to get a gun because they know there's going to be a background check and they know they won't pass it?

Lott: They may just try to get it the illegal way to begin with. Personally, I don't believe the claims that the Clinton administration makes about the number of people who are stopped from buying a gun. My guess is that to the extent that people are stopped, the vast majority of them are people who may have something on their record from 30 years ago, and they don't realize that it prevents them from buying a gun. These are people who may pose no risk to anybody. In fact, that's one of the reasons why I think there's such a low prosecution rate of those people.

Reason: The feds say they don't have the resources to prosecute.

Lott: I don't think that's it at all. I think you have prosecutorial discretion. I think that you have a case where somebody who's 50 years old may have done something as an 18-year-old that was wrong. The prosecutor looks at it and says, "This guy has been an upstanding member of the community for 30 years, and he had this one run-in as a teenager. We don't really think that he intended to violate the law. We're not going to send the guy to jail for doing this."

Reason: The National Rifle Association criticizes the Justice Department for not prosecuting enough of these cases.

Lott: I think that's a mistake. They're also talking about prosecuting cases where guns were brought onto school property. My guess is that a prosecutor would bend over backwards to bring a case against a juvenile who had brought a gun onto school property. He doesn't want to not bring the case and then have something bad happen later on. That would be disastrous for his career.

But let's say a kid's gone hunting in the morning before school. He has the gun in the trunk of his car, parks it in the school parking lot, and goes into school. Somebody finds out that he has a gun there. The prosecutor looks at the case and says, "This is a good kid, never done anything wrong. He probably just didn't realize he shouldn't have done this. Do I really want to send this kid to jail for three years for this type of violation?" It's wrong to think that these prosecutors are making the types of mistakes that are being assumed.

Reason: You've criticized the NRA for doing a poor job of making its case. What should it be doing differently?

Lott: My biggest complaint with the NRA is that they're too defensive. It seems to me that some of the [mass shootings] that have occurred are a result of gun laws that are already on the books. Rather than talking about what new law should be put in place, we should ask to what extent have well-intentioned laws in the past caused us to get to point where we are right now.

It's only been since the end of 1995 that we've banned guns within 1,000 feet of schools by federal law. Right after the Columbine attack, a friend of mine dropped off his kids at a public school in Seattle, and he e-mailed me afterward, because there was a big sign in front of the school that said, "This is a gun-free zone." The question I had was, if I put a sign like that in front of my home, would I think that people who are intent on attacking my home would be less likely, or more likely, to harm my children and my wife? You may be trying to create a safe area for your family, but what you've ended up accidentally doing is creating a safe zone for [criminals], because they have less to worry about.

The thing that I'd like to see the NRA try to do is to say, when attacks occur, since we can't have the police every place all the time, why not let these people defend themselves? The people who get permits for concealed handguns tend to be extremely law-abiding. They've never done one of these attacks in the 70 years that we've had these types of permits. When these people lose their permits, and it's only a tiny fraction of 1 percent who do, it's usually for reasons that have nothing to do with posing a threat to other people. Laws [like the Gun-Free School Zones Act] are obeyed by honest, law-abiding citizens, not by people who are intent on carrying out attacks. And to the extent that you disarm the law-abiding citizens in certain areas, you increase the probability of these attacks, which perversely leads to calls for more regulations.

Another example is gun locks. If I were with the NRA, I would emphasize the cost of constantly talking about this issue. You're actually endangering people's lives, for two reasons. One, you're exaggerating in their minds the risks of having guns in the home. And two, I would say it's not in everybody's interest to have a lock on their gun. If you live in a safe area and maybe have young kids, that might be fine. But if you live in a city, even if you have kids, I don't think it's really the wisest thing to have the gun locked up, because you're not going to be able to quickly access it to defend your family. And when you compare probabilities, accidental gun deaths in the home are trivial compared to the rate at which other types of deaths occur from crimes where innocent victims are attacked and a gun would benefit them.

Reason: Some advocates of gun rights base their claims mainly on the Second Amendment, while others offer a more utilitarian argument. Which approach is more effective?

Lott: I understand the constitutional arguments, but I think for the vast majority of people the bottom line is whether the presence of guns, on net, saves lives or costs lives. They may be able to understand in the abstract that having guns owned by civilians is some type of restraint on government, but I don't think most of them view that as a problem that they're facing any time soon. For them the bottom line is, What will save lives? And so I think that's where you have to argue.