Previously linked data, in chart form, and massaged a bit.
Uses 1960 crime rates as baseline.
Violent crimes are still 2.75X more common per capita than in 1960. Down from a high of 4.75X in 1990.
Murder rates near 1960 levels.
Even though the data was given at non-standard intervals, I processed it to get annual deltas (slope, derivative) of the violent crime rate.
This shows that the the annual delta was falling in 1990 or 1991, before the AWB. If the AWB caused subsequent violent crime rates to fall, you'd expect a steeper downward slope after 1994. The data shows that the delta actually lessens (negative slope gets less steep, derivative of the derivative). So, if folks claim that the crime rate deltas after the 1994 AWB are due to the AWB, they need to own the fact that it decreased the precipitous slide in violent crime rates after 1994.
Here is a handy analogy:
* Absolute number of crimes ~ Distance [miles, m] ~ x
* Crime rate ~ Velocity [miles per hour, m/h] ~ dx/t
* Crime rate delta ~ Acceleration [miles per hour per hour, (m/h)/h] ~ second derivative
HTH