Have you noticed how each side in the current argument about guns has a favorite statistic that they like to throw around? For the NRA it's the 2.5 million crimes that didn't occur last year because more and more people use a gun to deter a crime. For the president and his supporters it's the 40 percent of all gun transfers that take place without a background check. The NRA claims that their number "proves" that we would all be safer if everyone walked around with a gun. The gun-control crowd says that their number "proves" that we need to expand background checks so that guns won't end up in the wrong hands. Does anyone know where these numbers come from? Here's the answer.
The NRA number comes from a telephone survey conducted by the criminologist Gary Kleck in 1994. As for Obama, his number also comes from a 1994 telephone survey conducted by the sociologist Philip Cook. Note that 1994 was the last time we had a national debate on guns that ended with passage of both Brady and the assault weapons ban. Twenty years later we have a new debate but we're using the same old numbers. But it's not that the numbers are old; they are flawed.
Let's take Kleck's numbers first. No respondent was asked to prove that the incident could actually be independently verified. Although 60 percent said the incident had "come to the attention of the police," they were not asked whether it had actually been reported to the police or, for that matter, to anyone else. Kleck's explanation for this extraordinary lapse in methodology was that he assumed that many of the respondents might have been walking around either with guns they weren't supposed to be carrying or were carrying guns in places where such behavior was prohibited. But the survey didn't seek to determine that issue either. Did the people who claimed they used a gun to deter a crime really know what they were talking about? Can we trust anyone to accurately describe an event without having some way of independently verifying the truth of what they said? It's a no-brainer to verify the results of a political poll. Just wait for the election and then count the votes. But how do you verify something like whether someone really knew that a crime was going to take place? Especially when the whole point is that the crime didn't take place.
The methodology of Cook's survey is not only as flawed as Kleck's, but might even be worse. Cook asked his respondents if they knew how they acquired their weapon, and 60 percent said they "believed" they got it from a "licensed" dealer. They "believed." Then the Department of Justice took this number and assumed that the other 40 percent who admitted to acquiring guns in this survey must have gotten them from someone else. And this is the 40 percent who, 20 years later, still get their hands on guns without undergoing a background check.
By the way, there were no background checks in 1994. The NICS system only became operational in 1995. So nobody underwent a background check in 1994 and if Obama, Bloomberg and Manchin want to base their campaign for expanded background checks on the DOJ survey, they should be consistent and say that the rate of gun transfers without background checks today is 100 percent. Because that was the rate in 1994.
Last but not least: Kleck's survey was based on interviews with 225 people; Cook contacted 248. A national political argument that has consumed the attention of the government, the media and God knows how many advocacy groups is based on discussions with less than 500 people. Should we be at all surprised that we never found any WMDs in Iraq?