Pollster.com

Articles and Analysis

 

More on Those Romney Numbers

Topics: 2008 , The 2008 Race

I want to follow-up briefly on that ABC News/Washington Post poll question on Mitt Romney that I wrote about yesterday, the one showing 54% of adults (and "a third" of Republicans) saying they would "definitely not support" Romney. Alert reader Brent emailed to point out that the most recent Time/SRBI survey included a similar batter of "would-would not consider" questions that produced a very different response for Romney. Let's take a look.

First, revisit the ABC News/Washington Post question and the results for both Romney and other candidates:

ABC News/Washington Post (n=1,141 adults, 4/12-15): If (NAME) wins the (Democratic/Republican) nomination for president would you definitely vote for (him/her) in the general election for president in 2008, would you consider voting for (him/her) or would you definitely not vote for (him/her)?

04-20%20Washington%20Post.png

Now consider the Time/SRBI question and results:

Time/SRBI (n=1,102 registered voters, 4/5-9/07): If the following candidate were to run for president and the election was being held today how much would you support him/her...definitely support, probably support, probably not support, definitely not support?

04-20%20Time-SRBI.png

Notice a fairly consistent pattern for the candidates included on both surveys: On the one hand, the "definitely support" results differ by no more than four percentage points for any candidate - falling well within the range of random sampling error. On the other, the "definitely not" category gets a consistently bigger response on the ABC/Post survey, ranging from 9 points higher for Hillary Clinton to the whopping 38 point difference for Mitt Romney. Interestingly, the "not support" results are much closer (within single digits) if you compare the "definitely not" response on the ABC/Post survey to the total of "definitely not" and "probably not" on the Time poll.

While these questions are obviously very similar, there are two big differences that explain the general pattern.** First, the ABC question identifies the political party of each candidate. This information presumably makes it easier for partisans to offer an opinion about unfamiliar candidates. Second, the ABC/Post question offers three answer categories (definitely for, consider and definitely not for), while the Time question offers four (definitely support, probably support, probably not, definitely not).

That difference may seem subtle, but consider that attitude that many political partisans will hold toward a candidate of the opposite party they know in name only. Consider a Democrat asked to evaluate Mitt Romney, for example. It easy to imagine such a person saying they would "probably not" support Romney, but having a hard time choosing between "someone I'd actively consider" and "someone I could never, ever support," since they know so little about him. Yet the ABC/Post question forces them to choose between "definitely not support" and "would consider voting for." I'd guess most of those people end up in the "definitely not support" category, which comes closer to their opinion.

Of course, that theory is not an obvious explanation for why "a third" of Republicans would say they would "definitely not support" Romney. Although, again, if their first impression of Romney is poor - based either on the hunting story or (as some commenters suggested yesterday) negative views of the Mormon religion - the same phenomenon may occur. Many Republicans may have opted for "definitely not" when their true attitudes were closer to "probably not."

For what it's worth, the "definitely not" response for Romney has increased only 2 points since February on the Time/SRBI survey (from 14% to 16%), but the total of "definitely not" and "probably not" has increased by 12 points (from 33% to 45%).

So which question is more valid? Neither is "bad," in my view, but I think the Time/SRBI question provides a more interpretable snapshot of current opinions of most of these candidates. Few are well known enough at this stage to inspire truly strong support or opposition. When asked about individual candidates, the overwhelming majority of voters are either unfamiliar with candidates or are only willing to say which way they will "probably" vote (Hillary Clinton being the notable exception). Opposition to Mitt Romney is probably increasing, but not the dramatically high levels suggested by the ABC/Post poll.

**And yes, a third difference is that the ABC/Post poll asked these questions of adults and the Time/SRBI asked them only of registered voters. If anything, however, the difference in populations should have produced more intensity of feeling on the Time survey, since registered voters tend to be better informed and more opinionated about political candidates than non-registrants. Obviously, the numbers above show the opposite pattern.

UPDATE: While I was obsessing about Romney's numbers, Greg Sargent at TPMCafe and, in turn, Ana Marie Cox at Time were noticing the apparent similarity in the Clinton and McCain results on the ABC/Post poll: McCain's "definitely not support" number (47%) is two points higher than Clinton's (45%). So, asks Sargent appropriately, why not more "unelectable" stories about McCain?

Well, here too, the Time/SRBI version above shows a different result: Clinton's "definitely not" number is 36% to McCain's 17%, while the total "not support" numbers are 50% and 46% respectively. Again, the same pattern: The softer "probably not" answers on the Time poll seem to become "definitely not" on the ABC/Post survey.

On the subject of Clinton's electability, "Gallup Guru" Frank Newport has posted some thoughts worth reading about whether Clinton can win with a 45% favorable rating. Maybe, he concludes, after looking at past Gallup numbers. Keep in mind that favorable rating questions, like the "would you consider" item discussed here, are not created equal.

 

Comments
Watcher:

The news media exist to be political branding agents, and once they attach the brand they doggedly stick with it because it's a convenient substitute for putting any thought into their work. All coverage is essentially of the brand identity.

So, any story about McCain will be along the lines of, "Is he still a maverick?" Anything about Hillary will be, "Is she still a bitch?" Guiliani: "Still a hero?" Edwards: "Still too pretty to be president?" Gore: "That old fool again?"

What passes for drama in political coverage is the media's fixing of a new brand. Meantime, with each new election, another slice of the public looks at this circus and drops out.

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR