Journalismizationizing!

blather, Forward!, Hmm..., news, politics, weapons, ಠ_ಠ Add comments

Feb 172014

First, there’s the headline:

Study: Missouri murders spike after state repeals gun background check law

Then there’s the article itself, which claims that a study reports that after Missouri repealed a law that required that private gun sales (like, say, from neighbor to neighbor) have a state background check, the state has seen an increase in murders of 60 or so per year. The implication being that now that law abiding citizens don’t have to get a background check for guns, they are being inspired to murder more often.

But then… someone at Fark.com decided to actually check the numbers.

dittybopper

2014-02-17 12:26:12 PM

Magorn: Federal law would not apply to the transfers in question:
After the law was repealed, unlicensed sellers were no longer required to perform background checks before selling their guns.
I’m calling statistical shenanigans, though.
Here is the year, population, # of homicides, and rate for Missouri from 1997 to 2012 (last year I can find data for):

Year Pop. Hom. Rate/100k ---- --------- --- ---- 1997 5,481,193 387 7.06 1998 5,521,766 372 6.74 1999 5,561,950 329 5.92 2000 5,607,285 332 5.92 2001 5,641,142 399 7.07 2002 5,674,825 348 6.13 2003 5,709,403 319 5.59 2004 5,747,741 369 6.42 2005 5,790,300 417 7.20 2006 5,842,704 384 6.57 2007 5,887,612 382 6.49 2008 5,923,916 474 8.00 2009 5,961,088 402 6.74 2010 5,996,092 435 7.25 2011 6,008,984 385 6.41 2012 6,021,988 390 6.48

Data sources:
http://www.mshp.dps.missouri.gov/MSHPWeb/SAC/crime_data_violent_crim e_ 960grid.html (homicide numbers)
http://mcdc.missouri.edu/websas/estimates_by_age.shtml (population numbers)
Rate is calculated as (homicides/population)*100,000, rounded to nearest hundredth.

There seems to have been a significant jump in homicides in 2008, just after the law changed. That may or may not be related, but the subsequent 4 years after (2009-2012) don’t seem very different at all from 2004-2007, the years prior to when the law took effect on August 28th, 2007 (majority of 2007 was “need a permit”)

In fact, the average rate from 2004-2007 is 6.67 per 100k, and from 2009 to 2012 it’s 6.72, less than 1% higher. I’m not even sure if that would be a statistically significant increase.

If the homicide rate had stayed up in the 8 per 100,000 range, or even consistently about 7 per 100,000, I’d say “Yeah, looks like there might be something to this, warrants further study”. But they didn’t. They dropped right back down to near the average, and it only took me a few minutes to figure out with publicly available data that there is something funny going on statistically.