I saw a wonderful quip on GERRYMANDERING at the American Prowler website:

Gerry-Rigged Democracy
Political Hay
We have gone from voters choosing their representatives to
representatives choosing their voters.
John H. Fund, 8/13/2003 12:02:00 AM

[more, http://php.indiana.edu/~erasmuse/w/03.08.13a.htm ]

JOHN DONOHUE'S NEW PAPER, "The Final Bullet in the Body of the More Guns, Less Crime Hypothesis," is worth reading, but disappointing. It starts with this purpose:

Hailed as heroes by the National Rifle Association (NRA) and its supporters, while derided as scoundrels by their staunchest critics, Lott and Mustard precipitated a scholarly and political odyssey that can teach us much about the techniques and limitations of sophisticated empirical research and the divergent norms of the scholarly and political realms.

The next part of the papers is a good survey of the articles testing whether less gun control results in less crime. Donahue, a participant in the literature, concludes that in the end we really have no evidence that right-to-carry gun laws reduce the amount of robbery by scaring off potential robbers. That is probably a fair conclusion (I haven't followed the literature closely), but whether it is correct or not is not what most interested me. I was looking forward to the story of how the scholarly work was used and abused in politics.

Here, though is that part of the paper (with the footnotes below):

Unfortunately, though, there is a dark side to this story. While it took about six years for the scholarly community to fully discredit the more guns, less crime hypothesis, the Lott and Mustard research had a major influence on public policy as a number of states adopted RTC laws recently with legislators touting the research of Lott as alleged proof that their action will cut violent crime.[footnote 8] Some may argue that Lott and Mustard�s academic research was only window dressing and that it did not change any legislative outcomes, but this may be too optimistic a conclusion. At the very least, Lott�s research and subsequent lobbying efforts gave cover to those who might have been reluctant to support RTC laws and emboldened their supporters to push harder for such laws. Indeed, Attorney General John Ashcroft asked the U.S. Supreme Court to adopt a more NRA-friendly interpretation of the 2d Amendment using Lott�s research as "proof" that more guns would lead to less crime. In addition, Lott has inspired an entire cadre of gun-toters to believe that they are responsible for one of the most important benign trends in crime in American history (the large crime drops of the 1990s), even though the yet larger crime drops in the states that did not adopt the RTC laws should show the folly of that belief. As pro-gun groups like Guns Save Lives.com have sprouted across the country, these excitable and engaged (and armed) supporters of Lott�s work have been highly resistant to the refutations of the more guns, less crime thesis, and have been energized to greater political activity on behalf of the NRA agenda. Of course, if RTC laws are harmful and Lott and Mustard�s now discredited work has led to their greater adoption, then Lott and Mustard have imposed serious costs on the victims of the increased crime. Conversely, if the RTC laws have virtually no effect on crime but legislators voted for them thinking that they lowered crime, then at least there would be no blood on Lott and Mustard�s hands but there would still be the harm to the democratic process of encouraging the adoption of laws on false pretenses (however innocent the erroneous findings originally were).

Thus, we have conflicting lessons from this episode. The benign lesson in the scholarly realm is that those who ask interesting and important questions may help stimulate the ultimate attainment of truth even if they themselves generate the wrong initial conclusion.[footnote 9] The lesson in the political realm, though, is far more malign. At least until the truth emerges -- and perhaps even after it has become clear to open- minded scholars ---those in the political realm who wish to push a particular agenda will do so as soon as a superficially supportive academic study hits the stands....

[footnote 8] 8 From 1996-2000, there were no adoptions of RTC laws but since then Michigan adopted such a law in 2001, and Missouri, New Mexico, Minnesota, and Colorado all followed suit in 2003, with other states actively considering adoption. Alaska, which already has a RTC law that requires those wishing to carry guns to secure a permit, has now adopted a law allowing anyone who can lawfully carry a firearm to do so without a permit.

[footnote 9] "Mr. Lott's 1997 paper on gun policy was, "to that point, the most important piece of empirical research that has ever been done in the social sciences," says Jeffrey S. Parker, a professor of law at George Mason University. "I doubt that even Ayres and Donohue would dispute that point."" David Glenn, "Scholarly Debate Over Guns and Crime Rekindles as States Debate Legalization," The Chronicle of Higher Education, April 30, 2003. I do think that Lott�s work was important in that it stimulated the ultimate understanding that while RTC laws do not reduce crime, neither do they vastly increase it (since any overall crime increases are likely to be in the range of 1 to 2 percent).

I was also thinking that one of the lessons of the episode might be that it takes empirical work to get across theoretical ideas. The theory underlying Lott and Mustard's work is simple: if more non-criminals are allowed to carry guns, then criminals, fearful of getting shot, will be more reluctant to try to rob people. (With the nice extra twist that the laws would not deter them from burglary, as opposed to robbery, so we should also see burglary rates rise as criminals substitute into that crime). What all the scholarly arguments have been about is whether this theory shows up in the data we have available. But the big contribution of that debate may have been to bring the simple theory into the public debate. It's hard to do that with a simple theory by itself, no matter how profound the theory, but the extra fanciness of empirical work (or fancy mathematics) can pull along a good idea.

Let's suppose the empirical work ends up showing no effect of right-to-carry laws. That's what I'd have expected without reading any of the empirical work, but not because I think private gun ownership fails to deter crime. I think it probably does. But detecting the effect of private gun ownership on crime in the data we have is a tough task. First, crime data and gun ownership data are both highly imperfect-- just think about how it has to be collected. Second, the effect would probably only show up with a lag, since it might take several years for people to start carrying guns and several more before criminals figure it out. Third, the laws might have no effect, since people might carry the guns illegally (or even not knowing it was illegal!) even without the law. So if Donohue is right and in the end we can't find an effect of the right-to- carry laws, that just takes us back to where we usually are in this kind of public policy debate: the data is not good enough to give us good answers to all the questions we have, and we have to rely on theory and on less objective evidence.

There is a parallel somebody should explore between gun laws and the death penalty. For a while there were lots of studies of whether the death penalty deterred murder. Edward Leamer made fun of these in his famous AER article, "Let's Take the Con Out of Econometrics." He made fun of them for the good reason that the data was not good enough to answer the question. In recent years there were just too few executions in the U.S., and too little variability across states for any amount of fancy statistics to come to a sound conclusion, given how many things affect murder rates.

So, in the end the idea that people carrying guns reduces robbery is still alive, even if its empirical support is dead. What I'd like to see Professor Donohue do is check carefully not only whether the empirical work influenced politics, but separate this from how much influence the theoretical idea had.

POSTSCRIPT. Professor Parker's remark on the vast importance of Lott and Mustard's work is, of course, silly. Right-to-carry laws just aren't important enough to be the subject of the most important empirical work ever, for good or for ill. But it does raise the question of which empirical scholarly articles have been most influential. Stigler and Friedland on public utility prices? Feldstein on savings? Friedman and Schwartz on monetary policy? That silly psychology study that was the supposed basis of Brown vs. Board of Education?

[more, http://php.indiana.edu/~erasmuse/w/03.08.12a.htm ]

To return to Eric Rasmusen's weblog, click http://php.indiana.edu/~erasmuse/w/0.rasmusen.htm.