Hokum requires a bit of magic.

A popular con circulating in the gun control debates is that when the state of Missouri repealed their law requiring a permit to buy a handgun, that homicide shot up. Sadly, nobody told the FBI and the folks there who run the Universal Crime Reporting database.

Sadly, too, few reporters have found the online portal to this data, otherwise the meme might never have lived longer than the Bloomberg School of Public Health press release.

click for a larger, sharable version

click for a larger, sharable version

So how did this formerly illustrious school pump forth this effluvium when the raw numbers do not support the claim? It all comes down to math. In the press release for the study, they note that they too pulled data from the FBI crime database. But they also note that “the analyses controlled for changes in policing, incarceration, burglaries, unemployment, poverty, and other state laws adopted during the study period that could affect violent crime.” Indeed, the paper itself states “The association between the repeal of Missouri’s PTP handgun licensing law on homicide rates was estimated using a quasi-experimental research design …” [emphasis mine]

In blunter terms, they used assumptions and mathematical modeling to say that the actual homicide rate was not the actual homicide rate.

It gets worse. Though they were intent on modeling assumptions about the effects of other issue on homicides, they decided not to do so when the data was inconvenient. Missouri’s homicide rate dropped steeply before 1999, the start of this study’s analysis. But the authors did not include this in the data despite Missouri’s permit law being in effect since the early 20^th century. The authors said “Periods of dramatic change … are vulnerable to omitted variable bias in estimates of policy impact.” But what variables would be omitted? After all, the authors tested for data that is commonly available over extended periods of time, such as unemployment, poverty, incarceration, burglary, and the number of law enforcement officers per capita.

It is time to squash this talking point because data is data. Missouri homicides did not rise after the permit law was repealed, mathematical wizardry aside.