I'm sorry that the charts for my data aren't included in this post. I don't know how to embed them in one of my diaries. If anyone can tell me how, it'd be much appreciated.
Afterthe latest shooting today in NYC, I started wondering if the bad economy might have anything to do with the recent number of mass shootings. What I did can hardly be called a full analysis, and as the WSJ shows the correlation-cause relationship is hard to pin. But I found two things. First, that there is a positive correlation between the US unemployment rate and the homicide rate for every year between 1948 and 2006. The r-squared is 29.5% with p < .0001. (For the non-statistically inclined, that means that 29.5% of the variation in the homicide rate is explained by variation in the unemployment rate. The p value means that the probability of this relationship being due to chance is less than .01%)
What was also interesting is that there is a stronger relationship between the unemployment rate of the previous year and a given year's homicide rate. So the unemployment rate of 2004 is a better predictor of the murder rate in 2005 than 2005's unemployment rate. In this case for the years 1949-2005, the r-squared was 36.6%.
I don't see how there could be a one-to-one connection here, because it would almost seem like the homicide rate drops before the unemployment rate. But especially in the 70's and 80's they seem to change together. This is fascinating. I'd love to know more about what is behind this relationship.