I have previously observed that the National Opinion Research Center’s General Social Survey (GSS) underestimates the rate of gun ownership in the United States. This is a product both of non-response (people who do not answer the “do you own a gun question”) and denial (false negatives) among people who are actually gun owners.

Iowa State University political scientist Robert Urbatsch recently published a peer-reviewed article in The Social Science Journal addressing the non-response issue in the GSS data.

Among other findings, Urbatsch documents that Republicans, who are particularly likely to own firearms relative to Democrats and Independents, are also particularly likely to refuse to respond to the GSS gun ownership question.

I do not take this to be a blanket indictment of the General Social Survey. I have used GSS data for 25 years. No data is perfect. But this systematic under-representation of gun ownership in the GSS is significant because their (in my opinion low) rate of ownership is often used politically to suggest that gun owners are becoming a smaller and hence more marginal demographic in American society.

And Urbatsch’s study does not speak to the issue of false negatives — those who own guns who claim that they do not — which is very hard to document empirically. It is also possible that there are false positives — those who do not own guns but claim they do — but I do not think people are as strongly motivated to falsely claim ownership as to falsely deny it.