I have been critical of Arthur Kellermann’s gun research in the past, and I don’t want to beat a dead horse, but in reading for my post on the validity of gun ownership statistics came across another reason Kellermann’s name is so often taken in vain by those defending the gun culture in America.
In their 1995 article on the validity of gun ownership surveys, Rafferty and colleagues write:
Kellermann and coworkers conducted a landmark study validating responses to the question, “Are guns of any kind kept in your household?” in a face-to-face survey in Memphis, TN, and Seattle, WA.
Well, “Kellermann” and “landmark” in the same sentence merits a closer look. Unlike the Rafferty study, the Kellermann study is not freely available on-line, but everything important can be gotten from the abstract.
Beginning with 75 homes of owners of recently registered handguns, they were actually able to make contact with 55 households (73%).
Of those 55 households, 36% refused to participate in the survey. The 35 participating households, therefore, represent just under half of the households selected for the study (47%).
Of the participating households, 88.6% admitted to having one or more guns in the household. Note that this proportion is quite similar to the 87.3% of handgun licensees in Michigan that Rafferty and her colleagues found.
What about the other 12.7% of households that presumably had guns because they were recent handgun licensees? Unlike the study in Michigan, none of the 35 participants refused to answer the gun ownership question, but 8.6% said they HAD guns before, but no longer do. Um, yeah. OK.
One respondent of the 35 (2.9%) “denied categorically” that guns of any kind were kept in the home. Denial – not just a river in Egypt.
According to the abstract,
The authors conclude that, at least among registered gun owners, respondent answers to questions about gun ownership are generally valid and that survey data of this type can be utilized with confidence.
In this case, the face-to-face survey yielded a gun ownership rate that is probably at least 12% too low, BUT what about those 53% of households who were selected for inclusion in the study but were not in the final sample?
If those households would be more likely to be provide “false negatives” than the households that agreed to participate – and it is not unreasonable to think they would – then the 12% under-reporting would need to be increased accordingly.
This data was also collected from June to August 1987, pre-Clinton, pre-Brady Bill, pre-Assault Weapon Ban, pre-Obama. So it does not include any increase in under-reporting that is happening since then.
Although it is not his conclusion, this study by Kellerman suggests that under-reporting of gun ownership in survey studies is in the double digits, at least 12%, and probably much more than that.