To clarify I was commenting on murder rates specifically in light of how trauma medicine has reduced the fraction of violent assaults that cause death. The factors you describe seem more along the lines of avoiding violent assault in the first place. Controlling for improvements in trauma medicine, today's murder rate would be three times that of the 1960s, but the numbers would be better than the controlled for medicine 1990s numbers, which where five times 1960s levels. In other words yes in the past 20 years Americans seem to be getting assaulted less and I think all of what you describe played a role. There is also the unfortunate problem of police sometimes having nasty incentives to misclassify crimes so some of the drop might be fictional.