Last week, researchers at the Johns Hopkins Bloomberg School of Public Health released a “study” claiming that Connecticut’s 1995 permit-to-purchase (PTP) handgun law reduced gun-related killings in the state by 40%. Published in the American Journal of Public Health, the study is similar to research regarding a similar law in Missouri that the school published last year. While firearms and crime researcher, John Lott, has already picked the study apart over at the Crime Prevention Research Center, I would like to take some time to discuss some of the shortcomings posed by this Bloomberg-funded effort.
The Johns Hopkins study makes a few major, but related claims. The first of these is that Connecticut’s firearm-related homicide rate dropped by 40% after passing the PTP law. The study also asserts that the decrease in gun-related killings in the state outpaced the nationwide fall in firearm homicides and that the results cannot be discounted by an overall drop in violent crime throughout the United States during the 1990s and 2000s.
According to the study, the above graph tells the story. While the composition of the control line remains somewhat of a mystery, the “Synthetic Connecticut” homicide rate is a combination of data from Rhode Island, Maryland, Nevada, California, and New Hampshire that has been matched to historical data for real homicide rates in Connecticut. As you can see, the fit for the synthetic line as it compares to real Connecticut is impressive when we look at the pre-1995 data.
This is where the problems begin. Creating the Synthetic Connecticut line relied on combining historical data from a collection of other states in effort to match homicide figures from Connecticut. This allowed them to create a sort of homebrewed formula to take partial crime rates from other states and match them to crime in Constitution State. The Johns Hopkins researchers assumed that historical statistics would ensure a suitable match moving forward. They did not account for variations in state law or the myriad of factors that might also contribute to crime rates. The researchers over fit the synthetic data to a very specific set of real figures from Connecticut and did so over a highly limited timeline. Understandably, it is tempting to match a model to real data as closely as possible, but over fitting that model often results in poor forecasts. This fact is aggravated by the fact that the researchers failed to adequately consider the individual factors that might contribute to crime rates, including (but not limited to) state laws, state economies, and demographics. This is the sort of mistake that we would expect from a statistics novice or (more likely) someone with a clear agenda.
Another issue is that the researchers claimed that firearms-related homicide rates fell further and faster than the national average after the passage of the law. By and large, this affect appears to be exaggerated, but the Johns Hopkins data does show Connecticut faring somewhat better than the control group. The problem with this is that the researchers chose such a limited timeframe to examine the post-PTP data. As Lott calls out in his article, if the data had been extended to 2010, the reduction in gun-related homicides within Connecticut would have been outpaced by the rest of the nation. It seems just a bit convenient that the researchers chose to ignore another 5+ years of data.
So did the PTP law make Connecticut a safer place to live? The study certainly seems to indicate that it did. However, when we look at the years beyond 2005 and compare them to the overall downward trend in violent crime throughout the nation, the affects of the law seem much less clear. From 1995 to 2010, the firearms-related homicide rate in Connecticut actually stayed reasonably consistent as compared to the rest of the nation. The past 30 years have seen a consistent decline in violent and firearms related crime rates throughout the US and more complete data from Connecticut indicates that the state has not been uniquely successful in combating firearms and non-firearms related violent crime. If anything, the state has lagged behind the rest of the nation’s downward crime trend.
When assessing this study, we really must consider the source. In 2013, former New York City mayor and gun prohibitionist, Michael Bloomberg, made a $1.1 billion donation to Johns Hopkins to form the Bloomberg School of Public Health. Part of this school is the Johns Hopkins Center for Gun Policy and Research. $250 million of the donation was tagged to hire 50 new researchers for this school. Viewed in this light, it is practically impossible to conclude that the study comes from a policy-neutral source.
Furthermore, one of the lead researchers for the study, Daniel Webster, has published a book titled Reducing Gun Violence in America. This anti-gun publication even features a foreword from former mayor Bloomberg himself. When it comes to anti-gun research, Daniel Webster is one of Michael Bloomberg’s most reliable “scholars”.
That the release of this study coincides so well with a push from Congressional Democrats to implement a federal PTP law should not be taken as mere coincidence. Bloomberg-funded anti-gun groups have shown a remarkable ability to fixate on and rally around specific issues before. Impressively, most media outlets have completely ignored the link between the anti-gun former mayor and the source of this study or those who are pushing for similar laws in Congress. With this in mind, it is difficult to imagine why Bloomberg would waste further money starting his own anti-gun media outlet when mainstream news is so willing to buy in.