In a June 28 Los Angeles Times op-ed, American Enterprise Institute resident scholar John R. Lott Jr. presented false statements and misleading comparisons to assert a supposed link between falling crime rates and the September 2004 expiration of the federal assault weapons ban. In spite of Lott's long record of duplicity and misinformation in his published scholarly research, the Los Angeles Times editorial page continues to publish his opinion columns.

The federal assault weapons ban passed as part of the Violent Crime Control and Law Enforcement Act of 1994. It prohibited the manufacture, importation, and sale of 19 specific weapons (such as the AK-47), and prohibited firearm manufacturers from producing weapons with specific “assault weapon” features (such as mounts for bayonets or grenade launchers). It was set to expire after 10 years and did so on September 13, 2004.

In his June 28 op-ed, Lott wrote: “Last week, the FBI announced that the number of murders nationwide fell by 3.6% last year, the first drop since 1999.” Lott referred to preliminary data released for the 2004 Uniform Crime Report (UCR). Published annually by the FBI, UCRs provide detailed national, state, county, and metropolitan crime statistics. Lott, however, went on to write:

The trend was consistent; murders kept on declining after the assault weapons ban ended.

Even more interesting, the seven states that have their own assault weapons bans saw a smaller drop in murders than the 43 states without such laws, suggesting that doing away with the ban actually reduced crime. (States with bans averaged a 2.4% decline in murders; in three states with bans, the number of murders rose. States without bans saw murders fall by more than 4%.)

It is unclear how Lott could have obtained the data for these claims. The FBI will not release state-by-state and month-by-month UCR data for 2004 until fall 2005, as part of its final report “Crime in the United States, 2004” (the bureau has released previous years' reports in late October). The 2004 UCR preliminary data are not broken down this way, so they cannot justify Lott's conclusions.

Additionally, Lott committed a basic statistical fallacy by suggesting a link between the end of the assault weapons ban and lower murder rates: Correlation does not imply causation. One could similarly note, for example, that ice cream sales and crime rates both statistically increase during the summer, but that is no indication that higher crime and ice cream are linked. That's why the FBI website explicitly warns against comparing data the way Lott does. “Because of the many variables that influence crime in a particular town, city, county, state, or region, the UCR Program does not encourage comparisons of this nature,” the FBI states. Lott certainly knows this, since he has made a name for himself analyzing crime and other data using complex (and, as it turns out, fraudulent) regression analyses, a statistical method specifically designed to isolate the “predictive relationship” between a single dependent variable (e.g. crime) and one or more independent variables (e.g. gun ownership, poverty, population density). But Lott's op-ed gives no indication that he attempted any such analysis concerning the assault weapons ban.

Lott then indignantly noted that in spite of the widespread news coverage of the expiration of the assault weapons ban, “news that murder and other violent crime declined last year produced just one very brief paragraph in an insider political newsletter, the Hotline.” This statement is wrong on its face, however, since National Journal's “The Hotline” merely excerpts news that originally appeared in other sources. In this case, the June 9 “Hotline” referred to a weblog post that in turn quoted a June 8 Cybercast News Service (CNS) article. But beyond CNS, the Los Angeles Times, USA Today, the Baltimore Sun, and the Associated Press also reported on the national statistics.

Lott has been caught using fraudulent data, lying about it to cover his tracks, and using a fake Internet persona to hype the falsified work. Lott claims he conducted a 1997 survey on defensive gun usage, but strong evidence suggests he never conducted such a survey. A February 11, 2003, Washington Post article noted that “his critics are asking: What national survey? Lott has been unable to produce the poll data, which he says were lost when his computer crashed.” He also misrepresented the findings of the U.S. Commission on Civil Rights on voter disenfranchisement in Florida during the 2000 presidential election. A useful summary of Lott's various scholarly misdeeds is available here from blogger Tim Lambert. In November 2004, Media Matters for America endorsed a plea to Los Angeles Times editorial page editor Michael Kinsley by Kevin Drum, author of Washington Monthly's "Political Animal" weblog, to stop publishing Lott's columns.