Review of Academic Journal Article

¶ … 1994 Federal Assault Weapon Ban on Gun Violence Outcomes, by Koper and Roth (2001) was published in the Journal of Quantitative Criminology.

Key Claims in the Text

The key claim of the authors is a little bit strange. They perform a study on short-term data and then claim that they need long-run data in order to be able to analyze short-run outcomes of the assault rifle ban. They claim that "the ban may have contributed to a reduction in gun homicides" but then follow that up with "any likely impact from the ban will be very difficult to detect statistically for several more years." This is waffling -- they are avoiding making any meaningful claims. In addition to this, the authors when describing the rationale for the assault weapon ban, cited themselves and an unattributed newspaper article, creating an easy strawman. They actually cite themselves throughout the paper, at an abnormal rate.

They also created a weak argument for their hypothesis. On the idea that assault rifles kill better, they decided to examine the gun homicide rate -- this after admitting that most gun crimes did not involve assault rifles. What they did not do was test how many gun crimes were committed with assault weapons before and after the law was enacted. Thus, they did not test what they put in their title. This is probably why they ended up being unable to make any meaningful claims -- they avoided answering the research question altogether.

The authors end up with an argument that there is no data that can be used. To what end this argument serves is a matter for speculation -- are there lobbyists involved? -- but they had an easy question to research and went out of their way to obfuscate the issue, to avoid actually researching the subject, and then ultimately arrived at a conclusion that contradicted itself. If there is not enough data to make a short-run analysis -- a dubious contention -- then why run the study? The authors conclude that there are methodological difficulties that prevent them from getting answer. This is false -- they deliberately opted for a convoluted nonsensical methodology when the dead obvious methodology was right there all along.

I am deeply suspicious of this paper, and would not have passed it for publication in an academic journal. The authors rely on their prior work for almost the entire paper -- and they admit such for the quantitative portion of the work. That alone begs the question of what the point of this paper is. But beyond that, the authors have avoided doing the one thing they said they would do -- to look at the rates of assault rifle use in crime before and after the law. They dance around with proxies like "multiple-victim gun homicides" -- which could just as easily be committed with a handgun -- and "multiple-gunshot wound victimizations" -- again which includes handgun crimes. When you deliberately bury the results under an avalanche of irrelevant data, which is what happens when 90+% of gun crimes are committed with handguns, naturally the data is obfuscated, and conclusions become more difficult to reach. The only question I have is why are the authors so keen to avoid reaching conclusions? Is it because they support the law and don't want to admit it's failed? Or are they against the law and don't want to admit it's working? Either way, the authors deliberately muddled their conclusions and avoided taking any stand at all with this paper.