In the latest Econ Journal Watch, just out, Ian Ayres and John J. Donohue III have a paper, "Yet another refutation of the more guns, less crime hypothesis - with some help from Moody and Marvell" accessible (hopefully) at http://www.aier.org/ejw/archive/comments/doc_view/4018-ejw-200901?tmpi=component&format=raw.
Besides recounting how the major study by John Lott and Mustard has lots of data problems and cut off just before the post-1992 crime decline, they focus on a paper by Moody and Marvell that claims that a study of 24 states showed a reduction in crime for increasing access to guns. It turns out that in this study 23 of the 24 states had the opposite result, and the aggregate result was solely due to Florida. However, the data in Florida is all messed up and also probably caught a general crime wave decline due to a regression to the mean after the Mariel boat landings from Cuba in the early 1980s. I am not doing their paper justice, but the media discussion is often dominated by Lott and his allies who are now pushing for loosened gun laws in Virginia, and are counting on Dems laying low and not challenging their incessantly repeated claims that such gun law relaxations reduce crime. They should not lay low. The claims are baloney and lies, based on distorted date and misrepresentations of results from ones with better data.