Gun Control Laws and Homicide Rates -- Lies, Damned Lies, and Statistics
In most discussions of gun control, the question arises, "Do gun control laws make us safer or not?" The argument usually goes that, where there are fewer firearms, there are few homicide by firearm, and fewer suicides by firearms. The assumption is then made that gun control laws result in fewer guns in the population, ergo, there will be fewer homicides and suicides committed by firearms. (Since the point is to look for negative correlations between gun control laws and homicide and suicide rates, the studies usually neglect to look at the rates of homicides and suicides committed by other means.)
In my research, I came across a paper published in the Journal of Community Health 29.4 (August 2004) by James H. Price, Amy J. Thompson and Joseph A. Dake entitled "Factors associated with state variations in homicide, suicide, and unintentional firearm deaths."
This paper indicates that, when all other factors are held constant, the presence of gun control laws actually has a positive correlation with firearms homicide rates.
Firearm Homicide Death Rates
Prior to adjusting, several state-level variables were significantly associated with age-adjusted firearm homicide death rates (Table 1). Percent of state population that was African American (r = .752), the violent crime rate (other than homicide) (r = .679), and firearm prevalence (r = .422) were the most strongly positively correlated with firearm homicide death rates. Neither the combination of all gun laws nor any of the gun law categories were significantly associated with homicide death rates.
In order to investigate the individual relationship between each of these variables and firearm homicides, partial correlation coefficients were calculated. The partial correlations were adjusted for each of the other variables in the table. After adjusting for the other variables, the percent African American variable was still significantly related (r = .491) although much less strongly. The prevalence of firearms in the state also remained significant with a stronger relationship found (r = .516). In fact, over 50% of the variation of firearm homicides from state to state was associated with the prevalence of firearms (27%) and the proportion of the population that was African American (24%). After controlling for potentially confounding variables, it was found that the combination of gun laws was significantly positively associated with homicide death rates (r = .311). The gun control laws were associated with 10% of the variation in firearm homicide rates. None of the individual categories of gun laws were found to be significantly related.
(COPYRIGHT 2004 Plenum Publishing Corporation)
This result, that the presence of gun control laws correlates to higher homicide rates, may seem surprising to some people. There are several explanations. The first is that the r=.311 figure is a partial correlation coefficient, with means all other factors are held constant. In other words, if the rate of firearms ownership were the same, the states with gun control laws would have higher firearms homicide rates. The contention, however, is that gun control laws reduce the prevalence of firearms. (Anecdotal evidence indicates that this has not been the case in D.C.) The unadjusted coefficient is -0.062, which, while negative, is statistically insignificant. So even if the prevalence of firearms is reduced, the homicide-by-firearms rate is not significantly reduced by the existence of gun control laws.
The second possible explanation is that cause and effect are reversed, that more firearms homicides spur lawmakers to enact gun control laws, but those laws have not been in force long enough to take effect. Again, however, the evidence is against that theory. When the draconian gun control laws went into effect in D.C., the rate of homicides by firearms did decrease immediately and markedly. Unfortunately, the black markets quickly filled the void, and the criminals became well armed while the populace was disarmed. Furthermore, we have seen murder rates (and the rates of all other violent crimes) go down markedly with the passage of concealed-carry laws.
A third explanation is that gun control laws are ignored by criminals, so the bad guys have guns and the good guys don't.
In the final analysis, the existence of gun control laws seems to do more harm than good.
|Potential Risk Factors||Pearson|
|Firearm Prevalence||0.422 ***||0.516 ****|
|Number of Firearm Dealers||-0.198||-0.085|
|....Percent African American||0.752 ****||0.491 ****|
|Presence of Gun Laws (combined)||-0.062||0.311 **|
|....Crime Deterrence Laws||-0.040||0.239|
|....Government Control Laws||-0.192||-0.123|
|....Sales Restrictions Laws||0.056||0.061|
|Per Capita Alcohol Consumption||-0.080||0.033|
|Level of Urbanization||0.079||0.186|
|Violent Crime Rate (other than homicide)||0.679 ****||0.267 *|
|....Poverty Rate||0.389 ***||-0.016|
|....Unemployment Rate||0.236 *||0.085|
|....Percent College Graduate||-0.311 **||-0.032|
Note: Partial correlation coefficients are controlled for each of the other variables identified. Firearm homicide death rates are age adjusted rates.
* p [less than or equal to] 0.10;
** p [less than or equal to] 0.05;
*** p < 0.01;
**** p < 0.001.
Listed below are links to blogs that reference this entry: Gun Control Laws and Homicide Rates -- Lies, Damned Lies, and Statistics.
TrackBack URL for this entry: