Chicago: 12 hours, 1 neighborhood, 7 murders

[Dill]

(04-01-2017, 06:37 PM)bfine32 Wrote: You're smart enough to know that the effectiveness can be manipulated to support the view of those posting them.
For example how many people would be surprised to find that a study by the ACLU found the practice to be ineffective?
Conversely look what Heather MacDonald has to say about the exact same data:

 According to Heather Mac Donald, murders declined almost 80 percent and major felonies by almost 75 percent from the early 1990's to 2013 thanks to "proactive policing," which includes the practice of stop-and-frisk.

In 2011, "stops yielded nearly 800 guns and over 5,000 other weapons, mostly knives," according to Mac Donald. Critics of the practice argue that this isn't enough to justify its use, and they also claim there aren't enough arrests from the practice to justify it. However, Mac Donald points out that "the possibility of getting stopped has clearly deterred many gangbangers from packing heat — which is precisely the point" as well as deterred other crimes from being committed.

Looks like Ms. Heather has folded her narrative about a New York city policy she favors into a nationwide stat--a decline in crime occurring everywhere after 1990, regardless of stop and frisk. Perhaps she believes New York City's stop and frisk policy somehow dropped crime in Michigan, California and Florida too.  

If "gangbangers" have been deterred from packing heat, and so deterred from other crimes, then this deterrence should show up as a statistically meaningful drop in the crime rate. Has it?  Can we believe that crimes are being deterred when there is no effective change in the crime rate?

Just because some people may manipulate stats to support desired policy goals doesn't suddenly make all stat-based arguments and those who use them equally credible. It just means we have to evaluate anyone and everyone's use of stats. Otherwise, all anyone has to do is present "alternative facts" to equalize policy arguments.