Articles and Analysis


Gore Revealed as Frontrunner?

Topics: 2008 , The 2008 Race

Earlier today, Zogby International released what it calls a national "blind bio" telephone poll of 527 "likely Democratic primary voters" that was sponsored by AlGore.org, an independent group dedicated to convincing former Vice President Gore to join the race for president.

Here's the gist of the results from the Zogby release:

When Democratic likely voters were given brief biographical descriptions of the top three Democratic candidates - New York Sen. Hillary Clinton, Illinois Sen. Barack Obama, and former North Carolina Sen. John Edwards - along with the biography of Gore, the former Vice President won 35% support, while Clinton won 24%, Obama won 22% and Edwards trailed with 10% support.

The key twist here is that the question provided "brief biographical descriptions" of candidates, but not the names of the candidates (something less than obvious in the spiffy animated graphic of the poll results produced by AlGore.org). That practice is not unheard of, but is typically used by internal campaign polls to test campaign messages. This sort of test can be a powerful tool, though the results are very sensitive to the descriptions tested. Were these fair? Fortunately, Zogby provided the full text:

  • 35% - Candidate A (Gore) is an experienced candidate from the South who has been Vice President of the United States and a US Senator. This person has won several awards, including an Oscar, a Grammy, and an Emmy for his documentary about global climate change. This person has won the Nobel Peace prize and is recognized as an international authority on foreign policy, energy, the environment, and technology. This candidate has opposed the Iraq war from the beginning.
  • 24% - Candidate B (Clinton) is a candidate with roots in the South and the Midwest, but is currently a US Senator from a Northeastern State. This candidate is well known for work on many domestic issues, including education, children's issues, and health care. As a US Senator, this candidate voted to authorize the Iraq war. This candidate is critical of how the war has been handled by the current administration.
  • 22% - Candidate C (Obama)is a first-term US Senator from the Midwest who has emphasized efforts to reach out to include in the political process many people who are disaffected and unused to involvement in politics. This candidate brings a fresh face to Washington and draws huge crowds to campaign rallies. This candidate has opposed the Iraq war from the beginning.
  • 10% - Candidate D (Edwards)is a former US Senator from a southern state. This candidate also has run as a Vice Presidential candidate in the past. This candidate champions health care and education for the poor, and has experience running a national political campaign. As a US Senator, this candidate voted to authorize the Iraq war but has since said it was wrong to vote for authorization.
  • 10% Not sure

Again, keep in mind that respondents were not told that Candidate A is Gore, Candidate B is Clinton and so on, although some presumably made that connection on their own. However, one could argue that a few pertinent facts that are missing. The gender and race of the candidates for one. Another is that the "work on domestic issues" for which Candidate B is "well known for" occurred when she was First Lady of the United States. The question also assumes that the only issue differences among the candidates worth noting are their positions on the Iraq war authorization. Readers will undoubtedly spot other issues.

What is really unusual about this survey is that campaign pollsters typically use descriptive paragraphs like this to test the potential of unknown candidates with the resources to become much better known (such as Chris Dodd, Joe Biden, Mike Huckabee or even Ron Paul in Iowa or New Hampshire). Even then, I know of few campaign pollsters that go to the trouble of testing truly "blind" bios. Most include the real names to make the test as realistic as possible.

Of course, Hillary Clinton and Al Gore are two of the best known Democrats in the United States. Even on a recent CBS News survey that prompts respondents to say when they "have not heard enough to have an opinion," 92% of Democratic primary voters can rate Gore and 99% can rate Clinton. So the need for a "blind bio" serves little obvious purpose here, other than allowing the pollster to try to conceal the identity of "Candidate B."

A better test might involve a question like the following:

Suppose Al Gore decided to run for President. If the race for the Democratic party's nomination in 2008 comes down to a choice among Hillary Clinton, John Edwards, Al Gore, and Barack Obama, who would you like to see nominated -- Clinton, Edwards, Gore, Obama, or someone else?

The good news is, that question actually appeared at the end of the same recent CBS News survey. And the results? "Were he to enter the race," as the CBS release puts it, "Al Gore could be a serious contender," running just five points behind Clinton (an advantage that is not quite statistically significant assuming the usual 95% confidence level).


These results suggest that a Gore candidacy would shake up the race and cut deeply into Clinton's current support. But "reveal" Gore "as frontrunner," as the AlGore.org poll claims about their poll? No. To do that, Gore might have to convince Senator Clinton to change her name to "Candidate B."

PS: The filing deadline for the New Hampshire primary is tomorrow. Opportunities to file for primaries in other states will remain open for some time, but as a practical matter, the window of opportunity for an actual Gore candidacy is closing fast.

Typos corrected.



We commissioned a "blind bio" poll for one simple reason: to minimize response bias against Gore. Most voters are going to name a declared candidate when polled, and some may feel awkward naming Gore who hasn't shown much interest in running thus far. Our contention is that you have a more honest peek at what the field would look like if Gore was in the race with a "blind bio" than by using their names due to this bias. Obviously, if Gore enters the race, this particular issue dissolves overnight, and you can poll with names attached expecting to get a true read.

While the CBS poll with Gore trailing Hillary by 5 points without a campaign was impressive, the new Zogby poll shows that response bias had obscured the true strength of his support. He is the stealth frontrunner.


Mark Lindeman:

"...To do that, Gore might have to convince Senator Clinton to change her name to 'Candidate B.'"

Nailed it!

Wow, that animated graphic could do double-duty in Tufte's next book about how to lie with bar charts. By distorting the scale of the vertical axis, it manages to make Gore's advantage over Clinton and others appear even larger than it supposedly is. For instance, 35%/24% is about 1.46, but I estimate that the ratio of the 'leading edges' of the Gore and Clinton bars is more like 1.64.

I don't know that Gore really needs this sort of 'help.'


Mark Warschauer:

I general admire this site, but these kinds of postings don't help your credibility. The descriptions provided by "AlGore.org" are hardly unbiased. As for the second supposedly more objective poll you described, it highlights Al Gore's name right from the start and thus uses his name one more time than that of the other candidates. How about a poll as follows:

If the following people were all declared candidates for the Democratic nomination, which one would you support?


Mark Lindeman:

Mark W., Mark B. surely didn't say that the AlGore.org descriptions were unbiased, so I'm not sure what your beef is there. The main point of the post is to point out why AlGore.org's claim doesn't hold up.

As for the second question, well, I sure think it's "more objective" than the first. Don't you? You're right that it gives more prominence and play to Gore than to the other candidates, but in fairness to Mark B., he didn't write it. He was looking for a question that had actually been asked. (And if an extra mention of Gore's name is sufficient to detract from Clinton's vote share, that is not without interest in itself.) It would be great if someone had fielded your version so we could compare.


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.