Do early-stage candidate favorability ratings matter?

Yesterday, I questioned Nate Silver's argument that the GOP's chances of unseating President Obama in 2012 were weakened by the party's lack of high-favorability presidential candidates. Last night, he responded, showing that favorability ratings for presidential candidates are relatively stable across the primary process. He suggests this disproves my claim that early primary polls "don't matter" and are "useless":

Even at this early stage, polls tell us something — not everything, not a lot, but something — about how the candidates are liable to be perceived next year following the primaries.

In contrast, Mr. Nyhan has written that early primary polls “don’t matter” and that they are “useless” — and several other bloggers have echoed these statements. That just isn’t true. Yes, as a first approximation, the rule of thumb “don’t pay much attention to early primary polls” is probably better than “pay a lot of attention to early primary polls,” given the way that the media tends to overrate their importance. But Mr. Nyhan’s statement is hyperbolic.

There are a number of problems here. First, Silver is confusing two different arguments. My statements that early primary polls "don't matter" and are "useless" are drawn from a different post concerning the utility of primary horse-race and straw polls for predicting a party's eventual presidential nominee. The fact that candidate favorability doesn't change much during primaries in no way undermines that claim.

Second, the fact that candidate favorability is somewhat stable during the primary process doesn't mean it's ultimately a significant predictor of the general election outcome. As Silver later acknowledges, only one candidate is nominated by either party -- whether the favorability of Pat Buchanan or another losing candidate changes is irrelevant. My post yesterday focused on this point, and I stand by it.

To illustrate the burden of proof that Silver's argument must meet, consider how well presidential election outcomes can be predicted using extremely simple models. Here is a plot from Douglas Hibbs's Bread and Peace model, which explains almost 90% of the variation in the last 15 presidential election outcomes using just two variables -- real per-capita income growth and military fatalities:

The model fits the data extremely well, leaving very little residual variation for additional predictors to explain. Moreover, only one of the two biggest outliers fits Silver's hypothesis. It's possible, for instance, that George W. Bush's advantage in favorability helped him overperform against Al Gore in 2000. However, Bob Dole significantly underperformed in 1996 despite having much better early-stage favorability ratings.

The mechanism for this result is that candidates tend to converge to the support level we expect from the fundamentals over time. Consider the example of John McCain. Due to his high favorability ratings, many Beltway insiders perceived him as a formidable general election candidate. But as I predicted back in 2005, his favorability ratings declined over the course of the campaign and he ended up receiving almost exactly the vote share that forecasting models predicted.

My take on this issue is consistent with the argument that Princeton political scientist Larry Bartels made in an edited volume in 2002 (gated). In that chapter, he examined perceptions of personal traits -- a more fine-grained measure of candidate characteristics -- and concluded that "candidates' images are largely epiphenomenal and have only a modest impact on election outcomes":

The aim of this chapter is to provide a systematic test of the conventional wisdom that personality is key in contemporary American electoral politics. Using survey data from the six most recent presidential elections, the contours are examines of the candidates’ images (traits), the bases of those images in voters’ more fundamental political predispositions, and the impact of voters’ assessments of the candidates’ personal qualities on individual voting behaviour and on aggregate election outcomes. In stark contrast with the popular conception of contemporary electoral politics as candidate–centred and image–driven, it is argued that candidates’ images are largely epiphenomenal and have only a modest impact on election outcomes. This conclusion is underlined by the analysis given of the 2000 (Bush vs. Gore) presidential election, in which the estimated impact of voters’ assessments of the candidates’ personalities was even smaller than in the previous five elections considered here, although quite probably large enough to be decisive in an election decided by a few hundred votes in a single state.

As in the Bush v. Gore case, anything can matter if the election is close enough. Likewise, nominating a candidate with unprecedented negatives like Sarah Palin would likely cause the GOP to underperform in 2012. Otherwise, however, it's not clear to me that Republicans are significantly less likely to win the presidency as a result of the candidates who are running.

[Cross-posted to HuffPost Pollster]

Comments

[Carleton Wu]

Seem to me that the original question was 'can we quantifiably determine if this is a weak field for the GOP', and you've moved now to the question of whether individual candidates matter at all or if Presidential elections are driven almost entirely by other factors.
That was never the dispute- from Silver's original post "On the one hand, incumbent presidents aren’t easy to beat; on the other, the identity of the opposition candidate only matters within a fairly narrow interval (when the president’s approval rating is between roughly 40 percent and 50 percent). But unless a candidate like Mr. Clinton emerges, Republicans may well be at some risk of underachieving."

That agrees fairly well with your final paragraph. It seems to me that you've drifted away from the original question, which is odd since you grasped it perfectly well in your first response...

[Kylopod]

A couple points.

1. Last year, Ezra Klein linked to a Larry Bartels paper (in PDF form) to refute the idea that the 2000 election defied economically based predictions. The paper pointed out that there was a slowdown in income growth in 2000, which has historically been a very good predictor of presidential elections:

http://voices.washingtonpost.com/ezra-klein/2010/07/ironies_of_the_2000_election.html

2. This is a bit tangential, but I'm wondering if you can find me some information I've been looking for. You wrote in your previous piece on the subject that Reagan had 38/39% favorable/unfavorable ratings in early 1979. My question is, what were Nixon's favorables in 1967? Were such polls even taken back then?

I ask because of discussions I've had with Palin supporters, in which I've pointed to her whopping 54.3% unfavorable rating and suggested that no one with unfavorables that high has ever gone on to reach the presidency. I don't think it's impossible that she could become president, just that it would be a real first. They then tell me that Reagan and Nixon overcame high unfavorables to become president. That didn't sound right to me, and I went searching in Google News archive and my library's databases for reports of the polls around the time. I can't seem to find anything--mostly it's just comparative candidate polls, and favorability polls are reserved for people who have already reached the White House. But you were able to find a Cambridge Reports survey from 1979, so you seem to have access to databases I do not. Can you find any info on Nixon's favorability in the years leading up the '68 election?

[Deanna Grissom]

You moved away from the original argument as many Republicans are wont to do these days when they are unable to defend their position.
Comparing predicitions prior to the 2008 election of Mr. Silver? My money will be on his take and information gleaned from past facts that DO apply to the questions YOU raised.... prior to this tangent you appear to have gone off on today......

[bnyhan]

Kylopod, I can't find any general Nixon favorability polls from that period in the Roper database (the most comprehensive).

[bnyhan]

Carleton, the point of my first post was that the Republican chances in 2012 won't be affected by the composition of the current GOP field ("It's not clear, however, that the absence of such a figure will matter."). That's also the point of the second post.

[Carleton Wu]

It certainly didn't read that way to me. For example, you correctly observed that Bob Dole started the campaign with great fav/unfav, but ended up with middling numbers and lost. That supports the thesis 'early fav/unfav is not a good predictor of later fav/unfav', not 'challenger fav/unfav is irrelevant to predicting Presidential elections'.
You mention Clinton's early numbers, but not his later ones, so in retrospect it's unclear if you meant to demonstrate that Clinton became a strong candidate despite weak initial numbers or that his strength as a candidate didn't matter bc of the economy. Those would be cool #s btw if you've got them, to add to the discussion.

If you want to support the thesis that late-game challenger fav/unfav is not important in determining who wins Presidential elections, then I think you should stick to presenting data that support that point. Clinton's and Dole's early numbers are irrelevant to that argument I think, if late-game numbers are available.

But again- if your thesis is that Presidential elections are mostly determined by other factors, I think you're in agreement with Silver. His argument, as I read it, is
1)if early fav/unfav is a good predictor of later fav/unfav AND
2)if later fav/unfav is a good predictor of strength of a candidate THEN
3)The GOP doesn't have a very exciting field AND
4)if strength of candidate affects close races, this weak field could cost them if the bigger factors like the economy give us a close race

I don't agree with all of that- particularly from 1 to 2, I think that relatively unknown candidates (eg Clark 04 from his data) can grow their favorables more than their unfavorables and develop into high-favorability candidates. And there are a number of GOP candidates in that range now. So the future strength of the field is unknown (and probably always will be this early in the race). Nate's numbers omit that the field only needs one candidate to achieve this sort of growth, even if the average candidate's fav/unfav ratio doesn't change much.

But I think it's realistic to point out that this might not happen, and the GOP might end up with some compromise candidate like McCain in 08. And that this might shave a point or more off of their results, which could make all the difference in the world.

Of course, a bigger problem with this whole thing is teasing out how much fav/unfav changes are the result of the candidate and how much they come from extrinsic factors- particularly, the huge ones like the economy that we already know have great influence on election outcomes.

[Jinchi]

A minor question:

I'm a little surprised that both 1964 and 1980 (at the minimum and maximum extremes) fall pretty much exactly on the line. Is that a least squares fit?

[bnyhan]

Carleton, I omitted later-stage favorability numbers because they're highly correlated with the eventual outcome and seem to be a result of the structural factors that drive the vote. That's the point. For nominees, early-stage favorability plays a very small role; the vote is driven toward the outcome expected by the fundamentals.

[bnyhan]

Jinchi, there's a non-linear component to the model -- see http://douglas-hibbs.com/Election2008/2008Election-MainPage.htm and the associated papers.

[Carleton Wu]

"I omitted later-stage favorability numbers because they're highly correlated with the eventual outcome and seem to be a result of the structural factors that drive the vote. That's the point."

I didnt ask for October numbers; I asked for post-primary numbers.

Perhaps early fav/unfav numbers represent something about the candidate and later fav/unfav numbers merely reflect the structural factors that we're discussing. I suppose the best way to tease that out would be to see if those underlying factors correlate with the changes in fav/unfav- Id like to see that demonstrated rather than assumed though, esp since it would show us when and how this transition happens.
My guess is that, given the variability in the numbers Silver posted, that candidates fav/unfav aren't that strongly affected by national phenomena before a nominee is selected or becomes inevitable. It'd be hard to say that the extrinsic factors were simultaneously driving Edwards up in '08 but leaving Clinton and Obama at the same difference, or pushing Edwards and Clark up in '04 while sinking Dean and Kerry.
So I think the post-primary numbers would be interesting, and might show more about candidate quality than numbers from immediately before the election.

But while that's an interesting question, it doesn't get at the earlier one directly: how much can we tell about likely candidate quality today? Regardless of how much of an impact candidate quality has on the outcome, it certainly has some effect, and could tip a close election. If you think it's not interesting because the election is likely to be driven by other factors, then you're not much of a political junkie...