Bad Data and the "Evil Empire": Interpreting Poll Data on Gun Control

Article excerpt

Abusing the National Rifle Association (NRA) is always good sport for the intelligentsia and for gun-control true believers. Weil and Hemenway (pp. 353-365 in this issue) apparently could not resist the temptation to take some cheap shots at the "evil empire," hanging their case on some dubious data from a commercial survey. Fun aside, two things are clear about this article. First, its conclusions bear little relationship to the evidence. The authors' survey data, even if taken at face value, simply do not support the authors' conclusions. Second, the data are in any case so seriously flawed that no reliable conclusions of any kind could be drawn from them on the subjects that the authors address.


Although the authors describe at length their bivariate and multivariate comparisons of NRA members with nonmember gun owners, their main conclusions have little to do with these findings. Instead, their main conclusion derives from their Table 3 and is that the official positions of the NRA deviate significantly from positions supported by its members and by gun owners in general (p. 363). One reasonable response to this conclusion would be "So what?" Surely, this is true to some degree of all advocacy and interest groups, a point that is weakly conceded by the authors themselves (p. 361). It is unlikely that any advocacy group's policy positions always match up perfectly with the majority sentiments of their members. The finding is banal and uninteresting, or would be if it were actually supported by the evidence. The authors' data, however, do not in fact support even such a minor claim as it concerns the NRA, however true or false the claim may actually be.

The authors identify three attitudes for which this deviation or discrepancy supposedly exists. First, they frequently refer to a "7-day waiting period" that is endorsed by most gun owners and self-identified "NRA members" in their sample but opposed by the NRA. Only fitfully do the authors identify the measure more fully in the text as a background check with a 7-day waiting period. Their abbreviation serves to subtly obscure the fact that the NRA, although opposing waiting periods, officially endorses background checks (NRA, 1990) and thus is in perfect accord on this matter with both its members and gun owners. Because the authors cannot know which of the two elements of this measure respondents (Rs) were endorsing, they also cannotknow if any gun owners actually supported the waiting period per se, as distinct from the background check. It is perfectly possible that most gun-owning endorsers supported only the NRA-endorsed background check while being indifferent or hostile to the waiting period element.

Supporting this interpretation, one 1991 national survey of registered voters gave Rs a direct choice between an optional background check with a waiting period and a mandatory background check without the waiting period, and it found that most Rs preferred the option without the waiting period (Lawrence, 1991). This survey also found that support for the Brady bill dropped by 12 percentage points after Rs were told of difficulties in performing background checks and by another 15 percentage points after they were told that police did not have to perform the checks. These findings directly indicate that background checks are an essential element underlying support for the Brady bill among a large share of supporters. This share might have been even larger if analysis had been confined to gun owners. Weil and Hemenway blandly note that somehow it might have been nice to see whether Rs preferred an "instant check" measure, without conceding that NRA members may support only the NRA-supported background check and not the NRA-opposed waiting period.

The authors also conclude that there is an NRA-versus-members discrepancy concerning support for "stricter gun control laws." For many gun owners and NRA members, possibly most of them, this could refer to laws with stricter or mandatory penalties, such as laws specifying mandatory minimum or add-on penalties for committing various violent felonies with a gun. …