Duncan and Maltz letter to the AEI

Keiran Healy observes that the U Chicago Federalist Society acted with integrity when Lott libeled Donohue. In the comments, Michael Maltz posts the letter that he sent with Dudley Duncan to the AEI about Lott and the reply they received:

October 21, 2003

Christopher DeMuth
President
American Enterprise Institute
1150 Seventeenth Street, N.W.
Washington, DC 20036

Dear Mr. DeMuth:

As you are doubtless aware, a number of commentators have recommended that AEI initiate an inquiry into allegations of unprofessional and unethical behavior by John R. Lott, Jr., an AEI Resident Scholar. They include allegations:

• that Lott repeatedly presented erroneous information about what various polls and surveys had found about the frequency of gun brandishing, both in his book More Guns, Less Crime and in other media link 1,
  link 2,
  link 3;

• that he subsequently claimed that the figure he had been citing came from a survey he himself had done; but he is unable to provide any documentation whatever to support this claim, which itself is implausible [same references];

• that he adopted a pseudonym that he claimed he used to protect himself from scurrilous comments, but that he used to make scurrilous remarks about his critics and to puff up his own reputation, including characterizing himself as a chaired professor link 1, link 2;

• that he misrepresented himself as a professor in a law school, before a legislative committee link;

• that, after being informed of the problems with the data sets he used in the major study presented in More Guns, Less Crime (to wit, extensive errors and missing data throughout the data set, and splicing together data sets that users were warned against splicing together), he ignored or trivialized their implications with regard to his findings [Refs: Maltz & Targonski, "A Note on the Use of County-Level UCR," Journal of Quantitative Criminology, Vol. 18, No. 3, September 2002, pp. 297-318; Lott & Whitley, "Measurement Error in County-Level UCR Data," Journal of Quantitative Criminology, Vol. 19, No. 2, June 2003, pp. 185-198; Maltz & Targonski, "Measurement and Other Errors in County-Level UCR Data: A Reply to Lott and Whitley," Journal of Quantitative Criminology, Vol. 19, No. 2, June 2003, pp. 199-206];

• that when it was shown that his analysis of these error-laden data sets was, additionally, contaminated with coding errors that he had made, he continued to present his results as accurate in public forums link;

• that, in revising his analysis using the properly coded data, he changed the model so that the findings would conform to his original statements link

• that he tried to cover up this manipulation of the model he used by predating the file he posted on the web, making it seem that he changed the model prior to others finding his coding errors [Same reference].

Much information about these criticisms can be found on the weblog site maintained by Tim Lambert. Although that site does not provide a single succinct summary of the allegations, it presents an illuminating chronicle and chronology of the discoveries and arguments relevant to them as they emerged, in a voluminous and wide-ranging discussion.

The undersigned have been involved in analyses of some of those activities. Although various weblogs have carried some of our comments, they do not give sufficient detail. We stand ready to provide additional detail.

We emphasize that these allegations do not address the merits of the policy that Lott advocates, laws permitting the carrying of concealed weapons. With 44 states already having such laws, that issue is essentially moot. Rather, the issue is the credibility and reputation of John Lott as a scholar and, by association, the credibility and reputation of the American Enterprise Institute.

We do not contemplate publishing or publicizing this letter. We look forward to hearing from you.

Very truly yours,

Otis Dudley Duncan
Emeritus Professor of Sociology
University of California, Santa Barbara

Michael D. Maltz
Emeritus Professor of Criminal Justice and of Information and Decision Sciences
University of Illinois at Chicago

DeMuth's response:

Dear Professors Duncan and Maltz:
This is in response to your letter of October 21 concerning John Lott. I am familiar with the substance and merits of all of the matters on your bill of particulars, and do not need to ask anyone else to inquire into them for me. I should add that I think your accounts of many of these matters are very tendentious. In any event, the issues of research methodology and characterization that you raise have been thoroughly ventilated by Mr. Lott and his supporters and critics (including, as you note, Professor Maltz) in professional forums, and no doubt will continue m be-aided by Mr. Lott's exemplary willingness to share the large data sets he has assembled and to assist those who wish to challenge (or just study) his work and conclusions. I do not regard debates over the effects of concealed-weapon laws (or other gun regulations) a moot; I hope and expect that they will continue, and also hope (but with lower expectations) that they will proceed in a spirit of open and vigorous inquiry on the merits rather than personal vilification.

Yours truly,
Christopher DeMuth