On the Michael Dukes Radio Show: the renewed calls for an assault weapons ban, or why the Washington Post doesn’t have a clue on how to do research

Feb 24, 2018 | Featured

Dr. John Lott talked to Michael Dukes, whose radio show is heard in Anchorage and Fairbanks, Alaska.  They discussed an article in the Washington Post by Christopher Ingraham that pushed for an assault weapons ban after the Florida high school attack.  One’s first reaction to all this is that banning guns based on how they look rather than how they function seems very unlikely to have any effect (these semi-automatic AR-15s fire the same bullets, with the same rapidity, doing the same damage as semi-automatic hunting rifles), but it doesn’t some such as Ingraham to believe otherwise.


Ingraham makes two points:

  1. That a survey of 30 experts on gun issues by the New York Times shows: “On a scale of effectiveness ranging from 1 (not effective) to 10 (highly effective), the expert panel gave an average score of 6.8 to both an assault weapons ban and a ban on high-capacity magazines.”
  2. A book by Louis Klarevas that compares the average attacks and deaths both before, during, and after the assault weapons ban.  The claim is that the rate of attacks were lower during the 1994 to 2004 period than the other periods.

The problems with Ingraham’s claim are that the New York Times didn’t try to get a balanced representation of experts on the subject and that the way Klarevas analyzed his data didn’t make a lot of sense.

The Times’ decision on who to include in the list was pretty arbitrary.  For example, a survey of academics in Regulation who have published peer-reviewed empirical research on firearms over 14 years found 32 economists and 35 criminologists in North America.  Economists tend to be the most skeptical of gun control, but the Times somehow only managed to only include three economists.

On the second more important point, there are numerous problems, including the fact that Klarevas doesn’t even try to separate out the very few attacks where military-style assault weapons are used.  In addition, most academics would see how the rate of deaths changed in those states where the federal assault weapons ban changed the ability to own those types of weapons relative to those states that had their own bans and nothing changed.  Some studies that do this type of test properly are discussed here.  See also here for a study by Dr. John Lott that uses panel data for a period that covers both before, during and after the federal assault weapons ban and finds no impact from the ban on any type of crime rate (“More Guns, Less Crime,” University of Chicago Press, 2010, 3rd edition, Chapter 10).  Apparently, Ingraham never bothered to wonder why so many others who have looked at the impact of the assault weapon ban, including many liberal Democrats, such as Jeff Roth and Chris Koper (here and here), have failed to find results anywhere close to what Ingraham points to.

We can only think of two cases where the attackers bought their guns from other states and in neither instance was it an assault weapon.  This is important because it indicates that the impact of the assault weapons ban is a local one on the number of or lethality of these mass public shootings.  So the state level panel evidence discussed above would be appropriate.

  • Aaron Alexis, Washington, D.C., Navy Yard, 2013. Alexis legally purchased a shotgun in Virginia, where he lived. He then carried out the shooting at his workplace in D.C.
  • Jennifer San Marco, Goleta, California, 2006. San Marco purchased a handgun in New Mexico, where she lived. She then drove to her old place of employment in California where she committed the crime.

Finally, it is pretty unusual to look at 6+ killed in attacks as used here, and without a real reason for doing this, it makes one feel that different combinations were tried.  But this type of critique is completely beside the point because you have the chance to use panel data where you can compare how the number of deaths change over time for states with and without these laws.  The fact that the data isn’t tested in the way all the other studies have done, even with this variation on how to define the attacks to examine, speaks volumes.  Nor have we bothered to take the time to see what cases were included in this particular list.

UPDATE: Here is a very useful op-ed in the Los Angeles Times that builds on a point that we raised earlier:

. . . Klarevas doesn’t disaggregate his list of mass shootings by weapon type, so I had to do that myself by cross-referencing his dataset with Mother Jones’ list of U.S. mass shootings and with news reports.

What I found was that for the decade prior to the ban, only two of the 19 mass shootings in Klarevas’ dataset involved civilian versions of military rifles. Another three involved pistols banned by name under the 1994 legislation (two Uzis and one Tec-9), but these small guns use a popular handgun round, 9mm, and not the much larger 5.56 NATO rifle round that features so prominently in current arguments for why the government should ban the AR-15.

As for the decade during which the ban was in place, Klarevas concedes in a footnote that of the 12 shootings in his dataset, only three actually involved assault weapons.

All told, that’s five mass shootings that took place with “assault weapons” in the decade before the ban, and three that took place during its tenure. These numbers are far too small for any sort of statistical inference, especially if you’re trying to build a case for banning tens of millions of legally owned rifles. . . .

(Saturday, February 17, 2018)

Part 1

Part 2


1 Comment