More Guns, Less Crime: Understanding Crime and Gun Control Laws.

It is almost impossible to separate the problem of crime in America from questions about how guns should be regulated. There is typically some uncertainty about how a change to our gun laws will impact crime, since guns in principle have both helpful and harmful aspects.

Given this uncertainty, it is noteworthy that more than 30 states have now passed "right-to-carry" laws that allow private citizens to carry concealed guns. Proponents of these laws hope that increased gun carrying will increase the "costs" of crime and thereby deter some criminals. Yet it's also possible that some criminals will respond to an increase in gun carrying among the citizenry by using a gun more frequently themselves, which may increase the number of shootings that result from street crimes. The other potential drawback of these laws is that the daily annoyances of modern life, such as rude driving or Yankees fans, can lead to fights that would be more lethal if one or both of the parties is armed.

Ideally, we would measure the effects of right-to-carry laws by comparing the crime rate that a state like Idaho experienced after enacting a right-to-carry law in 1990 with the crime rate that Idaho would have experienced had that state not enacted a right-to-carry law in 1990. Because we cannot actually observe what would have happened had this law not passed, we are forced to compare Idaho with other states without such laws, such as California. The challenge is to statistically control for all of the relevant factors that will cause crime to be different between places like California and Idaho, is not a trivial exercise, since reliable measures for many of these factors are simply not available.)

The first attempt at a national study of the effects of right-to-carry laws is found in a new book by John Lott Jr., whose conclusions are hinted at in the book's title: More Guns, Less Crime. The book is a dizzying collection of 27 graphs, 27 figures, five appendices, 30 pages of footnotes, and even a few qualifications about the research. Most readers are unlikely to appreciate the importance of these qualifications, and will walk away with just the message of the concluding sentences: "Will allowing law-abiding citizens to carry concealed handguns save lives? The answer is yes, it will" This is unfortunate, since Lott's analysis ultimately cannot support his conclusion.

John Lott first called me in August 1996, as I was enjoying a cup of coffee in my windowless office at Georgetown University, to invite me to discuss his paper at a forum sponsored by the Cato Institute. I had never heard of Lott and barely knew where the institute was, but he asked nicely and I thought there might be a free lunch involved. (As it turns out, there was) My job was to determine whether Lott's article had solved the relevant statistical problems that plague this type of inquiry.

Several aspects of Lott's statistical approach struck me as troublesome. For example, he cannot adequately control for the fact that states with right-to-carry laws and those without are different with respect to poverty, gangs, drugs, and police practices. Inadequate controls for these factors will cause Lott's analysis to reveal that California has a different crime rate from Idaho even when this difference is unrelated to right-to-carry legislation. The concern that his statistical "fixes" might not fully correct for these problems received some support from his results: Several findings don't line up with previous research in criminology or common sense, and some of his findings were grossly inconsistent with each other. (For example, the implied effects of right-to-carry laws on homicide are 8 percent in Table 3 and 67 percent in Table 11) There seemed to be enough of these troublesome points to decide that his conclusions about the benefits of such laws were premature at best.

Shortly thereafter, Lott's data were reanalyzed by Professors Dan Black of the University of Kentucky and Daniel Nagin of Carnegie-Mellon. Their reanalysis was published this January in the same journal that carried Lott's original article, and provided more support for my fears about the Lott study: A well-known statistical test revealed that Lott's methods produce misleading estimates. When Black and Nagin address this problem using Lott's own data, they find no evidence that right-to-carry laws reduce crime. Other reanalyses have also cast doubt on Lott's findings, including my own study, currently in press at U.C. Berkeley.

The results from Lott's original article form the heart of this book. While he includes a chapter that contains replies to his critics, unfortunately he doesn't directly respond to the key Black and Nagin finding that formal statistical tests reject his methods. The closest he gets to addressing this point is to acknowledge "the more serious possibility that some other factor may have caused both the reduction in crime rates and the passage of the law to occur at the same time," but then goes on to say that he has "presented over a thousand [statistical model] specifications" that reveal "an extremely consistent pattern," that right-to-carry laws reduce crime. Another view would be that a thousand versions of a demonstrably invalid analytical approach produce boxes full of invalid results.

If there are now well-documented problems with Lott's research, what can account for the attention that journalists, politicians, and others continue to devote to this work? The answer can be found in part in some of the responses to his original article, which, regrettably, included attacks on his integrity as an academic researcher. Lott has thus been able to paint himself as the victim of those who are out to enforce the "politically correct, anti-gun orthodoxy." This distraction has apparently caused some observers to discount the growing body of evidence on the central limitations of Lott's analysis. To dispel the notion that Lott is simply being victimized by the "PC crowd," it may be helpful to mention the reaction of Gary Kleck, a Florida State criminologist known for his generally "pro-gun" views, who has launched a nasty attack on the public health community for what he perceives to be its "anti-gun" stance. Kleck argues in his recent book that it is "more likely (that] the declines in crime coinciding with relaxation of carry laws were largely attributable to other factors not controlled in the Lott and Mustard analysis."

Most of us would be inclined to support right-to-carry or other changes to our gun laws if they would save lives. While More Guns, Less Crime cannot answer this question, the discussion surrounding the book may nonetheless be helpful by highlighting how much more we still need to know about the role of gun laws in addressing our nation's crime problem.

Jens Ludwig is a visiting scholar at the North western University/University of Chicago joint Poverty Center.