Also, are you 100% certain that this doesn't reflect a change in what people are charged with, rather than their actual behavior? Have you any experience with prosecutorial discretion and how that affects what people are charged with, and hence, the "crime rate."
Do you have any evidence to suggest the definition of violent crime has changed at all? I'd be very curious if that were the case. In the absence of evidence, though, I'll continue going with what I understand to be the generally-accepted theory that crime rose dramatically throughout the 1970s and 1980s.
Whether that has any bearing whatsoever on crime stats, I have no idea. It could even be a response to crime going down, or completely unrelated.
It would be really interesting to try to figure it out.
In the time period that the incarceration rate in the USA grew 4x, the number of violent crimes in the US grew nearly 7x [1]. This was a much more dangerous country then. The incarceration _per crime_ rate is only about 25% higher than it was in 1960 [2].
None of which to say there is not something deeply, deeply flawed in our country today and in the way we deal with imprisonment. Similarly there are real fundamental issues of race and class that make crime and enforcement problematic.
That said, we seem to look at institutions more than society when we look at social ills. But just like with healthcare and education, the two are inseparable -- we must consider the priors.
[1]: Data sourced from two sources, focusing on 1960-1994: http://www.disastercenter.com/crime/uscrime.htm and http://en.wikipedia.org/wiki/United_States_incarceration_rat...
[2] http://www.crimeandconsequences.com/crimblog/documents/pictu...