Monday, January 9, 2012

The trend Josh Horwitz doesn't want you to see

Josh Horwitz, executive director of some third-tier gun control group called the Coalition to Stop Gun Violence, gives us a rather odd interpretation of the levels of NICS checks over the past thirteen years:

These are rates of (potential) gun sales per capita utilizing annual U.S. population data:
2011: 10,037,110 (3,217 per 100,000)
2010: 8,753,555 (2,835 per 100,000)
2009: 8,927,138 (2,907 per 100,000)
2008: 8,426,245 (2,771 per 100,000)
2007: 7,530,727 (2,499 per 100,000)
2006: 7,361,033 (2,467 per 100,000)
2005: 6,935,952 (2,346 per 100,000)
2004: 6,599,292 (2,253 per 100,000)
2003: 6,333,371 (2,182 per 100,000)
2002: 6,347,492 (2,206 per 100,000)
2001: 7,207,720 (2,528 per 100,000)
2000: 7,067,634 (2,504 per 100,000)
1999: 7,857,932 (2,816 per 100,000)
As you can see, over the past 13 years, the per capita "sales" figure has fluctuated between a high of 3,217 per 100,000 in 2011 and a low of 2,182 per 100,000 in 2003. But there have been no "dramatic" spikes in either direction dating back to the final two years of the Clinton administration.
The context here is that Joshy-boy wants us to believe that there is a continued decline of gun ownership in America, based on data collected in the General Social Survey, in which respondents are asked if they have guns during a telephone call from a stranger working for the government.  
In a wonderful Baghdad Bob-esque fashion, he implores his followers to ignore the steady increase in NICS checks.  One NICS check is performed every time an individual buys, or attempts to buy, one or more firearms from an FFL.  People generally don't like to waste time, so people will not attempt to buy a gun from a federally licensed dealer unless they will pass the instant background check.  Most transactions involve one gun at a time.  Therefore NICS checks track well with new gun purchases.  But this gun banner doesn't want you to think that.  
So let's vizualize  Horwitz's time series, and see if his claim - that  per capita "sales" figure has fluctuated between a high of 3,217 per 100,000 in 2011 and a low of 2,182 per 100,000 in 2003, is true.  The only correct use of "fluctuating" here would be that there is random variation around a constant mean.  Let's look:
Clearly, there has been a steady increase in NICS checks since 2002 after four years of decline from 1999 to 2001.  There's no argument about these numbers and no white-noise fluctuation as Josh Horwitz wants you to believe.  He intentionally avoided presenting the data in the clear manner I have done.  Look at his data presentation, and consider why he used reverse chronological order, along with levels and rates in the same row.  His intention was to hide meaning in the data, not to enlighten.  
You can also look at the annual percent change of the same series:
Again, the trend is clear.  For the past decade we're seeing consistent, positive growth rates in NICS checks, which closely tracks with gun purchases whether or not the Horowitzes and Sugarmanns want to believe it. 
Maybe the Joshes would believe it if they just visualized the data honestly.  

1 comment:

  1. They don't want to believe anything that doesn't fit with their lies.