Steve Verdon has responded to my critique of More Guns, Less Crime.
Verdon starts by claiming that Lott's argument doesn't depend on their being more guns or less crime. He argues that you just need "more people carrying (concealed) existing guns legally" and that Lott found a substitution from violent crime to property crime rather than "less crime". Verdon then accuses me of being "extremely dishonest" for stating that Lott is arguing that more guns cause less crime. I find his accusation very strange. If, in a book entitled More Guns, Less Crime Lott isn't arguing that more guns cause less crime, then the dishonest one is Lott, not me. Anyway, Verdon just seems to have misunderstood Lott's title -- one of the things Lott means by "more guns" is "more people carrying concealed guns". I specifically address this in my critique where I offer evidence that there was no significant change in the number of guns carried in public. And with "less crime", Lott is referring to violent crime, not to total crimes.
Next, in response to this passage
[Ayres and Donahue] found that, using Lott's model, in those jurisdictions carry laws were associated with more crime in all crime categories . Lott's model fails the predictive test.
Ayres and Donahue go on to examine all the states adopting carry laws using data up to 1997 and found that carry laws were associated with crime increases in more states than they were associated with decreases.
Verdon asserts:
If the model is incorrectly specified as this argument claims then relying on said model's results is not any better when the results are what you want them to be or your opponent wants them to be. The solution isn't to say, "See, now it says we're right and you're wrong, neener neener neener," but to try and correct for the mis-specification of the model.
However, he has once again misunderstood me. The first paragraph of mine above shows that Lott's model is misspecified. The second paragraph explains what happens when you correct that misspecification.
Next, Verdon falsely accuses me of mischaracterizing the findings of Plassman and Tideman. I wrote:
They also looked at the effects on each state and found a confusing pattern of results, with the effect varying from a statistically significant increase of 6.5% (Virginia) to a statistically significant decrease of 35% (Montana). While we would not expect the laws to have exactly the same effect in every state, it seems hard to see how the effects could be so radically different.
The first sentence correctly describes their findings. The second sentence is my interpretation of what their findings mean. It is true that they think that their findings mean that carry laws sometimes increase and sometimes decrease crime, but as explain above, I think that interpretation is incorrect.
Verdon then writes:
Finally, the idea that because Lott's work might be invalid (bad statistical methodologies, etc.) to conclude that the overall conclusion is wrong is an example of the fallacy fallacy. Lets suppose the model is incorrectly specified. Suppose a relevant variable (or several variables) are missing. Does it then follow that the conclusion that shall issue CCW laws result in higher crime rates? No. All that you can say is Lott's models don't work or need refining, etc.
Verdon is responding to an argument that I did not make at all. I don't understand how he could have read my critique as saying that. My final conclusion, is, I think, quite clear:
There may be good reasons for a state to introduce ``right-to-carry'' laws but reducing crime is not one of them.
Anyway, Verdon then repeats his false claim about what I did:
Re-fitting a model that Lambert et. al. are working so hard to show is flawed then claiming the result, results that Lambert et. al. now like, is bad statistics.
This once again misrepresents what I did.
Next, Verdon accuses me of "hypocrisy" for relying on Goertzel. Verdon argues that:
So the tests aren't whether or not some correlations look funky (as Lambert argues) or what the t-test statistics are (as Lambert does), but whether or not the model gives the right answer, i.e. does it predict well. Taking the new data that is available and seeing if it predicts the correct crime rates.
However, the only thing that I rely on Goertzel for is this quote:
When presented with an econometric model, consumers should insist on evidence that it can predict trends in data other than the data used to create it. Models that fail this test are junk science, no matter how complex the analysis.
Verdon says the same thing as Goertzel but somehow it is "hypocrisy" for me to quote Goertzel.