reasoned thought for an age of uncertainty
In my previous post on gun violence, I examined national gun violence statistics in order to assess the scope of gun violence in the US. The US, however, is a large country and gun violence can vary widely from one state to the next. So what do US state gun violence statistics tell us about guns in America? Well, they tell us where the killers are, of course.
Based on FBI Uniform Crime Reporting data (I have generally used 2010 state gun violence statistics throughout this article for consistency), the average gun murder rate per US state in 2010 was about 2.4 homicides per 100,000 persons (FBI figures are on average about 0.5 lower than CDC estimates, and no data was available for Florida). Louisiana was by far the state with the greatest number of gun homicides, with an average close to eight gun murders per 100,000 persons. Other states with unusually high murder rates include Delaware, Georgia, Maryland, Michigan, Mississippi, Missouri and South Carolina. The states with the lowest gun homicide rate are Hawaii, Idaho, Iowa, North Dakota, Maine, New Hampshire and Vermont. In addition, the states with the greatest number of gun murders also generally have a higher rate of gun homicides (correlation is 60%), though there are a few outliers, such as Connecticut and Wyoming.
There are a lot of interesting ways to slice these state gun violence statistics to try to garner further information. For example, how does the murder rate compare among red and blue states? Or north and south? As it turns out, there is not a very significant distinction between red and blue states, though red states do have a slightly higher gun homicide rate than blue states by a margin of about 0.5. The category that does show a significant distinction is the difference in the gun murder rate between states in the north and states in the south. The gun murder rate in southern states was about double that of northern states, while states in the middle averaged right in between these extremes. In looking for national policies that reduce gun violence, it therefore would be helpful to examine policy differences between southern and northern states that explain the gap in state gun violence statistics.
While an examination of US state gun murders may inform our understanding of the nation’s gun violence problem, the number of gun murders is a narrow way of viewing gun violence, because this does not take into account other forms of gun violence, including gun deaths that have not been classified as homicides (which occur at 3 times the rate of gun homicides), as well as non-fatal gun injuries. The Violence Policy Center recently analyzed the total number of gun deaths per US state, comparing this to level of gun ownership and gun regulation in each state. What they found was that the states with the weakest gun laws and high levels of gun ownership had the highest number of gun deaths per 100,000 persons.
VPC based their data on CDC death rate statistics (see Table 19, “Injury by Firearms”) and a 2002 study of household gun ownership conducted by Pediatrics, a the official journal of the American Academy of Pediatrics. The analysis and weighting of gun-law leniency was done by the VPC. The VPC study, while by no means definitive, reinforces the idea that policy can really make a difference in addressing gun violence–we just have to make it.
One issue with the VPC study, however, is that it confounds a number of variables and uses its own weighting system for the level of gun regulation. This introduces subjectivity and complicates results with multiple variables. In order to parse this study to its basic data elements, I compared the 2010 FBI homicide data on gun ownership rates to gun ownership data from the Pediatrics study that VPC relied on, running some basic regressions. What the data shows is a moderate correlation between gun deaths and gun ownership (about 50%) and no correlation (less than 1%) between gun murders and gun ownership.
Based on this regression, the states with the lowest gun deaths relative to the gun ownership were South Dakota, North Dakota, Minnesota, Iowa and Nebraska. Given that these states are all contiguous, it is highly likely that there is a reason why their state gun violence statistics shown low gun deaths relative to their gun ownership, and we would be smart to examine why. (My best guess is that cultural and regulatory factors have resulted in the lack of a black market for guns in these states, but then again, it could just be the winter.) The states with the highest level of gun murders relative to gun ownership were Louisiana, Maryland, Illinois, New Jersey and Nevada. Likewise, we would be well served to examine the reason why these states are at the bottom of the list. (My best guess? Black market dealing–think of mob states.)
The correlation here between gun deaths and gun ownership supports the idea that reducing gun ownership would be associated with a significant reduction in the number of gun deaths. It does not, however, go far in addressing our gun murder problem. There are many reasons why this could be the case, including that the gun ownership data from the Pediatrics study is over ten years old.
The more likely reason, however, is that household gun ownership is not an accurate indicator of the number of guns that exist in a state. This is because gun ownership studies generally ask an up or down question on whether you own a gun (“Do you have a firearm in your household?”), but they do not examine how many guns might exist within a household, and they do not account for black market guns. As it turns out, according to a study by the Harvard School of Public Health, gun ownership is in fact becoming increasingly concentrated, with almost 50% of gun owners (and 41% of households) owning 4 or more guns. The prospect of individuals amassing personal arsenals raises serious problems for law enforcement, because it makes guns harder to track and may be fostering the black market in guns. Because many homicides are committed with illegal weapons (up to 50% according to a late 1990s Department of Justice study), federal law enforcement officials should focus enforcement efforts for violations of current gun laws and black market sales. This is relatively easy to determine from the state gun violence statistics: those states with the highest gun murder rates (southern states) are most likely to be the states in which illegal black market gun dealing is most active.
From a policy perspective, there are a number of ways to reduce illegal gun sales, including closing the loophole on background checks for private gun sales. But in all likelihood, new measures will be required in order to reduce black market dealing. One approach to tracking weapons would be to require gun owners with over a certain number of weapons to be subject to oversight, including gun auditing and reporting. Gun dealers are currently subject to federal oversight, but there is little to no oversight of individuals who amass large personal arsenals and engage in underground dealing.
While Congress will certainly argue over which policies to implement and implement fewer policies than necessary, given the severity of gun violence in America, let’s hope they can at least agree on one thing: we are in dire need of more sophisticated gun ownership and crime data. If we can get that, we can find where the murderers and black market dealers lie and hopefully prevent future tragedies.