[traditionalist]

I suspect very much that is the case here as well.

I did too, at first, but unfortunately it is quite sound. The good news is that the study does not find that CCW increases crime. The only valid conclusion one can draw from it is that CCW doesn't affect crime one way or another.

[sailor4321]

Evidently Lott disagrees --- the article says the May issue of the Stanford Law Review will have a piece on this study plus an update by Lott of his work and a rejection by Lott of the findings of the Brookings "study".

When Lott's work was first published, all sorts of people came out of the woodwork with critiques. Lott patiently demolished these, and I suspect that will be the case here as well....

[Washingtonian]

This paper is garbage. Look at the most generally results that break down the impact of the law on a year-by-year basis. Graph out the coefficients and you clearly see that violent crime is falling immediately after the law. This is the most general specification, much more general than the "hybrid" model. It is also pretty clear what is happening with the intercept shift and straight line in the hybrid model. The data is nonlinear. Crime rates are falling at an increasing rate after the law is in effect. Fitting a straight line to that with an intercept shift overpredicts the crime rate in the early years. So much for their claim about a small initial increase. If you doubt me, draw a verticle line and then a quarter of a circle that starts at that line. Now fit a straight line through the middle of that curved line and you will see that it is above the curve line in the beginning. This is the same thing that is happening here.

[Washingtonian]

This paper is garbage. Look at the most generally results that break down the impact of the law on a year-by-year basis. Graph out the coefficients and you clearly see that violent crime is falling immediately after the law. This is the most general specification, much more general than the "hybrid" model. It is also pretty clear what is happening with the intercept shift and straight line in the hybrid model. The data is nonlinear. Crime rates are falling at an increasing rate after the law is in effect. Fitting a straight line to that with an intercept shift overpredicts the crime rate in the early years. So much for their claim about a small initial increase. If you doubt me, draw a verticle line and then a quarter of a circle that starts at that line. Now fit a straight line through the middle of that curved line and you will see that it is above the curve line in the beginning. This is the same thing that is happening here.