It's been a while since I had to deal with multivarate time series analysis, and Glen, I can sympathize with you (my wife graduated as a UCR Primate Psyhologist - which could explain why she was attracted to me
) and I have a bit more than a passing acquaintance with SAS, SPSS, and a bunch of other stat pkgs, but I was just sharing a brain fart. It was a moment when I had a dazzling command of the obvious, and it made me happy.
It was just one of those what is, IS, moments. There a tons of factors that influence crime rates, not the least of which is a good economy, lots of young males in the slammer, etc. But there are many interesting correlations, including the fact that crime is lower in 'shall-issues' states than it is in dicretionary issue or non-ccw states. At some point, all these positive correlations DO start to substantiate the premise that armed citizens do deter crime. The hard part is articulating it so the 'average' Jane or Joe can get it and get it right.
What tickled me, though, was that Lott's premise, when backed up with gun sales and crime stats IS prima facie evidence that the anti's are flat-assed wrong when they say that the simple availability of guns causes crime. The thing that said it all to me was a sign at the SAA-AIIM march in DC - "Self Control, not Gun Control". Of course that's hard & requires self-discipline , and the MMM whiners can't be bothered to expect people to be accountable/responsible, so they'd hobble the competent to accommodate the idiots, lest someone's 'self-esteem' suffer. But then again, facts have never meant much to that bunch, and I'm losing the buzz. Merde. M2