A more effective way to look at how guns have impacted murder and violent crime rates is to look at the various states that have changed sgnificantly within the past however many years, and then run regression analyses against them. Even better, because states like MN have had major changes in the conceal-carry laws, is to even look at a county-by-county basis. The reason for that is, like in MN, the change in the law was to make it a uniform process throughout the state where before it had been at the county sheriff's discretion and now it's not. County-based statistics will also help control for population density issues. You would also do well to even add in consdierations for changes in laws surrounding violent crimes, especially the punishment.
Simply put, you have certain states with loose gun control laws because they don't have much in the way of law enforcement agents, like Montana, where there's a county in the western half of the state bigger in size than many of the New England states, yet it only has a single sheriff and two deputies for all that land. Its gun policies are loose, but there's also not much in terms of people around to shoot at each other. Meanwhile, you have states like Illinois, where gun policy at the state level might not reflect what's happening in the urban areas like Chicago - which then has its own laws.
My senior thesis paper was on right-to-carry/concealed carry laws and their impact on murder rates. I learned at the end of my semester, after running numerous regression analyses, that murder rates were difficult to analyse at the state levels. I controlled for things like unemployment rates, labor force participation rates, population density, % change in GDP, and a slew of other factors, and even trying to control for ethnicity and race when the data would allow (figures obtained from the BLS). At the state level, for which I gathered all of my data, the only statistically significant impact on murder rates was the female labor force participation rates. This indicated that single-income families were on the rise, which led to worse or reduced parenting, which in turn led to higher crime rates. Beyond that - at the state level - all data was inconclusive. My data sets were from the years 1986 and 1996, the most recent data available at that time, and in between those years there were a number of states that changed their gun control laws; all data sets were taken from federal government resources.
