Thursday, February 10, 2011

Ignore Those Polls! (Palin Edition)

Every time someone writes the story about how badly Sarah Palin would be clobbered in the general election based on current polling, I'm going to do an item pointing out that it's based on a major fallacy.

This time it's WaPo's Rachel Weiner, telling us that "A series of state polls released over the last few weeks suggest that if Republicans nominate former Alaska Gov. Sarah Palin in 2012 it could hand President Obama a considerable electoral vote margin next November."

Look, this is really simple. Sarah Palin has receive essentially zero positive nationwide publicity for over two years. Consequently, she's very unpopular -- including among GOP primary voters. Therefore, she will not be nominated for president unless something changes.

If, however, something changes, and she becomes popular enough to win the nomination, then she will also be more popular overall. What's more, if that happens, she will receive all sorts of positive publicity -- that's what happens to candidates who win key primaries! And, in turn, that will stand to make her even more popular overall.

Of course, that's no guarantee that she would be a competitive nominee, or even that she would be basically comparable to any other potential nominee. She could certainly also win the nomination and then stumble; what's more, it is at least theoretically possible that she could win the nomination by narrowly targeting a slim majority of Republicans while alienating everyone else.

But those are possibilities, only. It's much more likely that if she wins the nomination -- and, granted, that looks less likely every day she fails to do the things one must do to win presidential nominations -- she'll have a very different national image than she does right now.

This is true, by the way, for every possible nominee. Which means that looking at horse race polling between Barack Obama and potential GOP nominees is utterly useless at this point. Whatever Mitt Romney, or Tim Pawlenty, or Rick Perry, or Sarah Palin looks like by August 2012 just doesn't have all that much to do with what they look like now.


  1. Those are mostly Public Policy Polling surveys and even Tom Jensen over there has said that she remains the most favorable among Republicans in the same polls where she underperforms relative to Obama in head-to-heads. That jibes well with the CNN numbers from earlier in the week showing Republicans wanting to win in 2012 over picking a standard bearer who matches them (the voters) perfectly on all issues.

    You're right. Unless her favorable numbers U-turn, she's not going to win the Republican nomination.

  2. Oh, please. Look at her nationwide unfavorables:

    54.3% of Americans have a negative view of her, and it has been in the majority since mid-2009. You can crow all about how it doesn't matter and how positive coverage would change all of that, but this is completely unprecedented. I have never heard of a candidate with such high unfavorables going on to later win the presidency. I have heard of incumbent presidents overcoming low approval ratings to win a second term (Reagan, for example), but never someone who wasn't in the White House to begin with.

    Does that make it impossible? I never said that. Politics is like quantum mechanics, nothing is impossible, just highly improbable. If she's nominated and unemployment shoots to 12%, I honestly have no idea what would happen then. It would be like the old philosophical conundrum of the immovable object versus the irresistible force, except it would be the doomed incumbent versus the unelectable challenger. We got a taste of that sort of situation in the Reid-Angle race last year (the doomed incumbent won), but that doesn't mean it was at all predictable.

    The larger truth, of course, is that she may be toast in the primaries due to these numbers, but the failure of the party so far to agree on any alternative suggests the possibility she may somehow squeak by, courtesy of her core of devoted supporters, who are as delusional a group as I've ever seen in politics.

  3. Kylopod,

    The point is that there's no way she can win the nomination with those numbers. She can't "just squeak by" if she's that unpopular. But if she starts, say, showing up at debates and doing really well, then both her GOP numbers and her overall numbers will rise.

    It's possible that her negatives are so firm that she can't turn them around -- but if that's the case, then she can't win the nomination, either.

  4. "Consequently, she's very unpopular -- including among GOP primary voters."

    If she's "unpopular" with GOP primary voters (meaning 20% of GOP primary voters), then everyone is "unpopular" with GOP primary voters.

    Both Hillary and Obama had 20% negatives with Democrat Party primary voters during the primaries.

  5. That same CNN poll shows her the top first and second choice among Republican and Republican-leaning independents. Everyone says that's a poll of only Republicans when it's not. It includes all the independents that would conceivably vote in a GOP primary (GOP-leaning indies).

    The key stat is that she's the top "second-choice" on that list. Deductive reasoning would lead one to believe that she's the heavy second-choice among Huckabee supporters. Huckabee leads her 21-19 at the moment. So as long as she remains what appears to be a 2.5-1 favorite among Huckabee supporters over Romney, then she's probably a strong favorite against Romney without Huckabee in the race.

    The other stat about electability isn't that vital becuase most primary voters (even when you include all GOP-leaning indies) think she's electable according to some polls. Also, it's a misleading question too. CNN presented it as an either-or-choice between electability and agreement on the issues. Most people would choose the former if you frame the question in that way and that's what the CNN poll reflected. But it's not bad news for Palin considering that most Republican and Republican-leaning indies believe she's electable.

    Again, the CNN poll included all the indies that could vote in a GOP primary.

  6. Also, here's the biggest problem with Bernstein's analysis. He argues that Palin should be higher in the polls than 19% given her competition. That would only be true if Romney and Huckabee weren't there. Both Romney and Huckabee have a near equivalent amount of name recognition that she does. So it's not like she's tied with a bunch of unknowns.

    Bernstein also argues that she doesn't appeal to enough of the party. Around 70% of the party likes her. Of course, not 70% of the party will vote for her. But if you remember correctly, Obama and Clinton both didn't break 38% in either Iowa and New Hampshire. Does that mean 62% of the party didn't like Clinton or Obama? Nope, it's just hard to get over 38% in a multi-candidate field where there are other well-liked Republicans.

    The reality is too that Pawlenty and the lesser knowns aren't completely unknown. Around 50% of GOP and GOP-leaning indies know who they are according to the CNN poll. It's bad news for them that the 50% who do know who they aren't indicating that they are voting for them. For example, Pawlenty is at 3% with 50% name rec. That means, he wins less than 1 in every 15 people who know who he is. If this trend holds for the 50% who do not know Pawlenty (and there's no reason to believe it wouldn't), then he'll be at 6%.

    So the key number for Palin in my opinion is how much of Huckabee's support can she win when compared to how much of Huckabee's support will go to Romney. As long as this ratio is 2.5-1, she'll be in roughly the same percentage of Hillary, Obama, Huckabee himself, and McCain were in Iowa and New Hampshire when they won in 2008 (it's unreasonable to believe that all the undecideds will break against her as such a scenario rarely occurs in primary races).

  7. Gotta disagree with you here.

    Those numbers are real. She isn't unknown. Those unfavorables aren't mild, either. Palin would end up doing significantly worse than almost any other GOPher. That doesn't mean she can't win; if the economy gets much worse, Hitler would win. But, I'd say Palin ends up doing 2-3 points worse than a replacement level candidate. Yes, parties matter. A pitcher can perform better with a better defense behind them, but that doesn't necessarily make a bad pitcher into a league average one.

    Yes, she would need to be doing better amongst Reps to get the nomination, and that would raise her overall numbers. But, look at how polarized her numbers are already. She could easily make gains amongst Reps that don't translate into gains amongst Indies or Dems.

  8. JB:

    >But if she starts, say, showing up at debates and doing really well, then both her GOP numbers and her overall numbers will rise.

    I agree. What I disagree with is your implication that if her numbers rise enough for her to win the nomination, that will be enough for her to win the general election. It's quite possible she could win the nomination and still be practically unelectable.

  9. K,

    No, I'm not saying it will be enough. I'm just saying that there's a huge unknown there. It's certainly possible she could win the nomination but be, say, 15 points worse in the general than Romney or whoever. But it's also possible that she could win the nomination and be as strong as anyone the GOP could put up. The point is we have to imagine what it would take for her to win the nomination, and whatever that is would have to be something that shows she's different than what she's shown so far.


Note: Only a member of this blog may post a comment.

Who links to my website?