Martin
This is a great example of a topic where really rigorous designs are important. This topic is so partisan that only the most rigorous designs can move people's prior views. One issue with cross-sectional designs for this type of policy is that places with the biggest problem are also the most likely to implement restrictive policies. A bias in the other direction is that places where gun ownership is most common (and therefore gun-related deaths most common) may also have the most difficulty getting gun ownership restrictions passed. It's also really important to be specific about what types of restrictions are effective and in what situations. I think more convincing studies are coming out of longitudinal data, when a state or municipality has changed their law, e.g., Kara Rudolph's recently published study in Am J Public Health on Connecticut's policy change. The lack of data on this issue, due to federal funding restrictions, has been incredibly limiting for the research progress.
Maria