Questionable Polling: GOP Presidential Preference Questions Vary Widely
Herman Cain has been ahead of Mitt Romney in the most recent GOP presidential candidate polling average at Real Clear Politics by a microscopic margin since late last week.
Readers might be surprised to know that the wordings of the presidential preference questions at the various polling organizations differ significantly. In my view, the same person might given a different answer depending on which organization's polling question was asked. Here are the examples, with the Cain-Romney split identified in each instance (links are to fairly large PDFs in some instances):
- AP-GfK (Romney 30%, Cain 26%) -- "Which of the following candidates would you MOST like to see win the Republican nomination for president?"
- CNN/Opinion Research (Romney 26%, Cain 25%) -- "Next, I'm going to read a list of people who may be running in the Republican primary for president in 2012. After I read all the names, please tell me which candidate you would be most likely to support for the Republican nomination for President in the year 2012, or if you would support someone else." (list of candidates then read in random order)
- Rasmussen (Cain 29%, Romney 29%) -- "If the 2012 Republican Primary for President were held today would you vote for Rick Perry, Mitt Romney, Herman Cain, Michele Bachmann, Ron Paul, Rick Santorum, Newt Gingrich or Jon Huntsman?" (half of respondents hear names in order indicated; other half hears them in another order)
- Public Policy Polling (Cain 30%, Romney 22%) -- "If the Republican candidates for President were Michele Bachmann, Herman Cain, Newt Gingrich, Jon Huntsman, Gary Johnson, Ron Paul, Rick Perry, Mitt Romney, and Rick Santorum, who would you vote for?"
- NBC/WSJ (Cain 27%, Romney 23%) -- "Let me read you a list of people who might seek the 2012 Republican nomination for president. If you were voting today in the 2012 Republican primary for president, which one of the following candidates would you favor? (READ LIST. RANDOMIZE. IF "NOT SURE," ASK:) Well, which way do you lean?"
- Reuters/Ipsos (Cain 23%, Romney 19%) -- "Several commentators have said the Republican presidential field for the 2012 presidential election is becoming clear. If the 2012 Republican presidential primaries were being held today, for whom of the following would you vote?"
- Gallup (link will probably download to desktop; NOT currently part of the RCP average; last poll showed Romney 20%, Cain 18%) -- "Next, I'm going to read a list of people who may be running in the Republican primaries for president in the 2012 election. After I read all the names, please tell me which of those candidates you would be most likely to support for the Republican nomination for President in 2012, or if you would support someone else."
Items 1 and 7 from AP-GfK and Gallup are questionable, because there may be a difference between who someone would personally prefer right now and who they would most prefer for the nomination. I would argue that Mitt Romney currently benefits from this question given his alleged perception (not so coincidentally fostered by the AP) as the "inevitable" candidate. Geez guys, why not just ask the "if the election were held today" question?
Items 2, 5, and 7 from CNN and NBC/WSJ are also weak in the sense that the candidates named at the actual questions are actually running; no one is a "may" or a "might" (i.e., Palin, Christie, and others are NOT mentioned). If a respondent isn't sure that a bottom-tier candidate they might otherwise support is really running, that might skew the answers towards the front-runners, again (since his name recognition is currently the highest) benefitting Romney. Perhaps these pollsters just haven't updated the presentations even though they're getting the questions right in actually polling conversations or robocalls. If that's the case, they're just being sloppy in reporting.
Items 3 and 4 are fine, unless you think Rammussen's omission of Gary Johnson is somehow a problem.
The introductory statement at Item 6 from Reuters/Ipsos also seems designed to mislead some voters. Some may conclude that "the field is becoming clear" means that lower-tier candidates are close to being winnowed away, again leading to a bias towards the front-runners, and currently to Romney because he has the highest name recognition. Why do that?
Unfortunately, the wide variance in the presidential polling question means that -- in addition to sampling and other problems often found -- how the question was asked has to be considered before reaching a conclusion as to how particular candidates are actually faring -- and perhaps "unfairly" benefitting.
Cross-posted at BizzyBlog.com.