General Discussion
In reply to the discussion: Gun Homicide Rates Up 31 Percent Since Stand Your Ground(FL) [View all]Igel
(35,197 posts)Just the press release. The headline says "rate," but the article entails "number." That's a egregious bit of confusion.
Meanwhile, there was a jump in the homicide rate from 2005 to 2006. http://www.disastercenter.com/crime/flcrime.htm
It ended a multi-year low-homicide period, and was the third lowest year (looking at the absolute numbers) since '77. In other words, the year taken as the baseline, as the standard for comparison, is an outlier. It's hard to not pick that year in some way, but "easy" isn't always "correct." Strictly speaking 2004 would have been a better year. The law too effect on 10/1/2005, so 25% of the baseline year is also in the dataset that's being evaluated. But if they'd picked 2004 they'd have not gotten the results they did. Again, picking an outlier skews the stats.
A few years after the jump in 2006, the rate declined again, presumably for reasons also not well understood since SYG didn't go away. The absolute numbers also declined, but not nearly as much. Comparing 2005 to 2015, there's an increase of 160 homicides. But the rate increased only by 0.1. What accounts for the difference?
The population increased during that time. Factor out the population increase and take a multi-year window for averaging the rates, and the increase vanishes. (Just eyeballing it, it looks like the rate declined. But that's because 2005 was anomalously low.) I'm going to go out on a limb and suggest that whatever else happened perhaps the size of the male 18-24 year-old cohort in Florida increased.
Like I said, can't get to the original article. Perhaps they did note that the homicide rate went from 5.0 to 5.1 percent, after getting down into the upper 4.x range, and count that as a 30% (instead of a 2%) increase. Or perhaps they found other deaths to include. Dunno. But press-release conclusions devoid of statistical clarity are bad premises for much of anything.