This article describes some original research I have done on gun sales, gunshot injuries, and gun policy in America. This article is critical of current US gun policy, and gun enthusiasts may want to avoid reading this article as it presents some facts and figures that shows our current gun policy in a negative light.
I realized that I have by this time written a couple of different articles about guns in America, and interested readers can review these post if they want. On guns and the “stand your ground” laws here (http://www.dailykos.com/...); on statistics about gun injuries here (http://www.dailykos.com/...); on gun availability and gun violence (http://www.dailykos.com/...); on civil rights and guns here (http://www.dailykos.com/...), and on how corporate profits and guns here (http://www.dailykos.com/...).
In the past four weeks we have learned of public mass murders in Aurora, CO. (12 dead), Oak Creek, WI. (6 dead), College Station, TX. (3 dead), the shooting of four police officers in LaPlace LI, (2 dead), and a highly-publicized daylight mass shooting on the streets of New York City. The rash of high-profile, highly murderous public shootings has many Americans asking questions about gun violence in our society: what causes these deadly events, and how do we prevent more needless public slaughter. As with motor vehicle accidents and cigarette use, gun violence can be viewed as a problem of public health, and then the methodologies of public health can be employed to understand the problem empirically, and to suggest rational solutions. Based on a public health perspective, I undertook a longitudinal examination of gunshot injuries and gun sales. I wanted to look at overall trends in gun sales and gun injuries, and to see if there was any linkage between gun sales and injuries.
The Center for Disease Control (CDC) keeps a databank on injuries and fatalities in the US. I collected the yearly data on fatal and non-fatal gunshot injuries, both intentional and non-intentional, for both sexes and all ages. This data includes gunshot injuries resulting form criminal acts, shootings by law enforcement, unintentional shootings, and suicides. For purposes of comparison, I also collected data on injuries to American bicycle riders during the same period. I chose injuries to bicycle riders as a comparator because bicycles, like guns, are very popular and ubiquitous in the US, but unlike guns, bicycles are not sold as tools to injure others. For data on gun sales, I used data from the FBI's National Instant Criminal Background Check System (NICS). The FBI NICS was started in 1999 to provide a way for federally licensed firearms dealers to instantly determine if a prospective gun buyer is eligible to buy a gun or explosives. Anytime a federally licensed firearm dealer is preparing to sell a gun to a customer, the dealer is required by law to make an inquiry to the FBI NICS: it is important to note that the inquiry does not indicate that a gun was actually sold. So the FBI NICS data is not a direct measure of gun sales, and only tracks transactions taking place at federally licensed dealers, and not those taking place at unlicensed dealers, guns shows, private sales and the like. I have chosen to use this data as a surrogate for gun sales because: 1) the FBI is a non-partisan and nationally recognized organization with extensive experience collecting statistics; 2) the FBI NICS data covers the nation as a whole and is available every year back to 1999; and 3) I can find no better source for good reliable data on US gun sales. Because the FBI NICS data only involves prospective purchases at federally licensed firearms dealers, and not transactions made through non-licensed dealers and private sales, the NICS data is widely acknowledged to underestimate national gun sales and therefore provides a very conservative measure of actual guns sales. The data on gun injuries and NICS data were then examined for trends and correlations.
While far and away more people are injured every year using bicycles than guns, gun injuries are far more likely than bicycle injuries to be fatal: the crude (non age-adjusted) death rate for bicycles for the period 1999 – 2009 was 2.8 persons per 1 million, while the crude death rate for guns during the same period was 103.2 persons per 1 million.
The data on gunshot injuries in the US from 1999 to 2009 shows a clear and steady increase in both fatal and non-fatal gunshot injuries, rising from a low of 28,000 fatal gunshot injuries in 1999 to over 31,000 gunshot fatalities in 2009 (the last year for which data is available). The data on non-fatal gunshot injuries shows a similar pattern. This same pattern is also observed in the FBI NICS data; from a low of 8.5 million NICS inquires in 2000, steadily increasing to over 16 million inquiries in 2011, almost a doubling of transactions in a little more than ten years. Interestingly, the data on bicycle injuries shows no such trend: both the numbers for fatal and non-fatal bicycle injuries jump around a good deal, without any clear overall increase or decrease. A regression of injuries over time shows that bicycle injuries (both fatal and non-fatal) did not change significantly over time (Beta did not differ significantly from 0), while the variable for both gun injuries (fatal and non-fatal) and gun sales were all increasing significantly over time (Betas differed significantly from 0 and were all positive).
Year - - - Fatal - - - Nonfatal - - - Fatal - - - Nonfatal - - - FBII performed a correlation to look for any association of guns sales with gun injuries. The FBI NICS data and fatal gunshot injuries were highly correlated (Pearson's r = 0.741), and this measure was highly statistically significant (p = 0.009). The FBI NICS data and non-fatal gunshot injuries were also highly correlated (Pearson's r = 0.645), and this measure was also statistically significant. (For those interested in this kind of thing, the Pearson's correlation coefficient – r – is a measure of how closely two variables move together. A Pearson's r of 1.0 means that any change in one variable is perfectly mimicked by a change in the same direction of the other variable, while a Pearson's r of 0.0 means that a change in one variable is completely unrelated to changes in the other variable. Pearson's r can take negative values from zero to -1, indicating that two variable move in congruent fashion, but in opposite directions).
Gunshot Gunshot Bicycle Bicycle NICS
Injuries Injuries Injuries Injuries Data
1999 - 28,874 - 800 - 9,138,000
2000 - 28,663 - 740 - 8,543,000
2001 - 29,573 - 63,012 - 792 - 519,424 - 8,910,000
2002 - 30,242 - 58,841 - 767 - 505,233 - 8,454,000
2003 - 30,136 - 65,834 - 762 - 492,900 - 8,481,000
2004 - 29,569 - 64389 - 843 - 490,864 - 8,687,000
2005 - 30,694 - 69,825 - 927 - 481,205 - 8,952,000
2006 - 30,896 - 71,417 - 926 - 466,712 - 10,036,000
2007 - 31,224 - 69,863 - 820 - 495,500 - 11,177,000
2008 - 31,593 - 78,662 - 893 - 494,003 - 12,709,000
2009 - 31,347 - 66,769 - 785 - 519,736 - 14,033,000
2010 - 73,505 - 516,912 - 14,409,000
2011 - 16,454,000
The results of this study show unmistakeably that: 1) gunshot injuries, both fatal and non-fatal, have increased steadily over the period studied; 2) guns sales as modeled by the NICS data have likewise risen steadily since this record was initiated; and 3) there exists a tight linkage between gun sales and gunshot injuries, especially for fatal gunshot injuries. The statistics on bicycle injuries over this same period show no overall increase or decrease.
That gun sales and gunshot injuries are highly associated is supported by other studies comparing gunshot injuries in the US with gunshot injuries in other western industrialized nations where gun ownership is more tightly controlled: while the US enjoys the highest number of guns per capita of any nation studied, the US also has the highest rate of fatal gunshot injuries of any country studied (Krug EG. Intl J Epidemiology. 1998; 27:214-22).
The increasing number of gunshot injuries is also occurring at a time of decreasing incidence of violent crime. According to reports issued by the FBI, violent crime statistics have been decreasing over the past several years (http://www.fbi.gov/...). The divergence of violent crime and gunshot injury statistics occurs because gunshot injuries occur in a variety of settings, not just during violent crimes.
It is important to note that a correlation of rising guns sales with increasing gunshot injuries does not prove that rising gun sales cause more gun injuries. It only proves that the variables guns sales and gunshot injuries move in tandem. However, because we now have evidence that guns sales and gunshot injuries tend to move in similar ways, this suggests one way to reduce gunshot injuries in the US is to sell fewer guns. Sadly, the numbers do not show any steady decrease in gunshot injuries because, since its inception, the FBI NICS has not recorded any multi-year periods of decreasing gun sales.
This longitudinal study show conclusively that both gunshot injuries and gun sales in the US have in recent years shown a steady increase, and that gun sales and gunshot injuries are highly associated. This study also demonstrates the utility of examining gunshot injuries from the perspective of a public health problem. Sadly, our nation's premier public health institution, the CDC, has been denied funding specifically for the purpose of studying gunshot injuries. With more and more fatal gunshot injuries occurring every year, this kind of research is very much needed.
1) I must first apologize for the appearence of the data table. I have not yet learned how to properly format a table within the dk diary application. Any tips on how to present a data table in a dk diary is greatly appreciated.
2) I have omitted many of the details of the regression and correlation analyses to make the article friendly to non-scientists. I am happy to provide more information to any interested readers.
3) As I noted in the article, the FBI NICS statistics provide only an indirect measurement of gun sales. The NICS data has been used by other authors as a marker of gun sales, for the same reasons I listed above. I would be interested in hearing from any readers who know a better source of more complete information about yearly gun sales in the US.