Holder Poll: No Real Opinions Formed Among Many

07/11/2012 11:47 am ET | Updated Sep 10, 2012

The Slate July 9 headline reads, "Most Americans Support Contempt Vote Against Holder." The subheading suggests nuance: "But while the new poll shows a majority backs the move, most still see it as a political decision."

The poll, conducted by CNN/ORC, reveals that there is no nuance, or subtlety or news story. We are creating news when there is none.

Journalists and their readers may believe that there is meaning to the 53 percent of the respondents who approve of the contempt action by the U.S. House. (For the record, 33 percent disapprove, and 13 percent have no opinion.)

More meaningful is an earlier question, which notes that 31 percent of respondents have never heard of Eric Holder, and that 13 percent have no opinion of him.

The survey also indicates that most people think that the Republicans seek to "gain political advantage" (as compared to having "real ethical concerns"). This dichotomy is also devoid of significance and should not be over-interpreted. One can gain political advantage and have real ethical concerns. It also implies that there may be manufactured or artificial ethical concerns. Why would one assume that a party bloc would not seek political advantage? Parties are political institutions; we should not be surprised that the answer tilts toward the 'political.'

George Bishop et al.'s 1980 work, "Pseudo-Opinions on Political Affairs," (Public Opinion Quarterly, Volume 44, No. 2) remains the best text for students and scholars of public opinion to be wary of survey responses as automatically substantive and meaningful. The authors asked people about whether or not the 1975 Public Affairs Act should be repealed. But the authors created the Public Affairs Act. It did not exist. They also inquired about a non-existent social service agency with the acronym ECTA. Sure enough, some people provided views about ECTA as well.

The comparison between Holder and the "Pseudo-Opinions" thesis is not a perfect fit, but there is a parallel narrative. Some people give answers to questions, including people who are unfamiliar with an issue or an individual. Given that close to a third of respondents had not heard of Holder, and that an additional 13 percent have no opinion of him, it begs the question what to make of the majority who approve of the contempt action against him.

We can speculate. Denoting the Republican (or any) party is likely to prime respondents to answer that politics is an ulterior motive. The phrase "refusing to turn over documents" is likely to garner a disapproving response, regardless of which office or officeholder is in charge. (It should also be noted that many documents were turned over, but not the complete set requested by a majority of the House committee members).

It is reasonable to ask if citizens know what a contempt action is. This is not a trick question. Legal proceedings are complicated. If we were to learn of ignorance about contempt proceedings, then pollsters may recognize the challenges of asking about political processes about which many of us know little.

In the polarized blogosphere, my fear is that I will be accused of defending Holder, or the House Democrats. Nonsense. I would write the same column if the Attorney General were John Ashcroft.

Polls are news, and are critically valuable for democracy. They often tell us about what we believe and why. But not all polls, even reputable polls with thoughtfully constructed questions, are newsworthy. We public opinion scholars have a responsibility to inform readers and journalists how and when to interpret data. When over four in ten respondents have no opinion or knowledge of an individual, one should be careful not to read into the results supporting or opposing any actions towards that individual.

Robert M. Eisinger is the dean of the school of liberal arts and the associate vice president for academic services at the Savannah College of Art and Design.