An academic review of Lott's work. This is just the conclusion, the link has so much information that proves a direct statistical correlation is impossible to draw and doesn't hold up.
CONCLUSION
Judge Richard Posner has recently criticized moral philosophy for failing
to persuade on any contentious issue.132 A similar criticism might be made of
quantitative empiricism. Readers might tend to accept only those quantitative
analyses that resonate with their prior normative beliefs. Indeed, Judge
Stephen Reinhardt famously proclaimed at a Yale seminar that social science
had never affected his judicial decisionmaking.133 And Donald Braman and
Dan Kahan have recently called upon econometricians like Lott and us to put
away our statistical packages. In a piece provocatively titled, More Statistics,
Less Persuasion, Braman and Kahan argue that rather than quantifying the
impact of gun control laws on crime, academics should dedicate themselves
to constructing a new expressive idiom that will allow citizens to debate the
cultural issues that divide them.134
We disagree. Over time, a body of empirical research can disentangle
thorny issues of causation and lead toward consensus. We view this Article as
playing a role in this process (not in ending the conversation). On net, we
believe that Lott and Mustards efforts made an important contribution to the
literature. They asked the initial question, amassed an important new panel
dataset, and then energetically and creatively analyzed it. (Indeed, their
dataset, which we know from experience was quite costly to construct, has been
used by many researchers to explore this and other questions about crime.)
Nevertheless, their results have not withstood the test of time. When we added
five years of county data and seven years of state data, allowing us to test an
additional fourteen jurisdictions that adopted shall-issue laws, the previous Lott
and Mustard findings proved not to be robust. Importantly, we showed that the
Lott and Mustard results collapse when the more complete county data is
subjected to less-constrained jurisdiction-specific specifications or when the
more-complete state data is tweaked in plausible ways. No longer can any
plausible case be made on statistical grounds that shall-issue laws are likely to
reduce crime for all or even most states. How much further one can go in
arguing that shall-issue laws likely increase crime across the board or have
heterogeneous effects across states (albeit most commonly pernicious) will be
matters about which various analysts will differ. We conclude with Learned
Hands advice that, unlike a policy advocate, an academic must keep an open
mind to every disconcerting fact, [and] an open ear to the cold voice of doubt.
Hand admonished: You may not carry a sword beneath a scholars gown.135
http://islandia.law.yale.edu/ayers/Ayres_Donohue_article.pdf 