It sometimes seems that political observers are divided between those who treat polling data, no matter how early or ephemeral, as Gospel Truth, and those who want to dismiss polling data, or even all data, out of hand, either for some period prior to real votes or forever.
At Washington Monthly this week, I discussed the issue in response to a sound but in my opinion misleading post from an observer whose work I admire:
At the Upshot today, Nate Cohn has a good primer on what you should ignore in all the early GOP nomination contest horse-race polls, but goes over the brink into one of those general injunctions to ignore early polls, presumably because he thinks readers are sure to misinterpret them. But then he makes a questionable assertion about how we should view the field:
Some might say that Mr. Rubio and Mr. Cruz’s support is enough to put them alongside Mr. Bush or Mr. Walker, the two candidates who have led the polls and have often been described as front-runners for the nomination. But Mr. Bush and Mr. Walker are front-runners in spite of their standing in the polls, not because of it.
They’re front-runners because the other candidates do not appear to have enough support from party elites to sustain a national campaign. Those other candidates do not have natural factional bases — like moderates for Mr. Bush, and conservatives for Mr. Walker — that give them clear opportunities to win early contests, or do not have the potential to build broad enough coalitions to win the nomination.
But how do we know Bush and Walker have these “natural factional bases.” You could say we know this about Jebbie because of rumors of fabulous fundraising numbers and all the blind quotes from Establishment types expressing their adoration for him. But until all this turns into reported contributions or public endorsements, it remains speculative, doesn’t it? I’d say a big reason for the Jeb the Frontrunner assumption is that his putative rival for that “factional base,” Chris Christie, is drawing terrible numbers in the early polls. And by that I don’t necessarily mean his horse-race standing, but his favorable/unfavorable ratios and the distribution of what little support he has. Similarly, we know Scott Walker is formidable not because of money or endorsements (he has little of either so far) but because early polls consistently show him with decisively strong support among conservative ideologues, and clear potential for growth in the rest of the primary electorate. And we know Marco Rubio has the potential to become a top-tier candidate because of his consistently strong approval ratios–again, in the early polls.
So I would amend Nate’s advice by saying it’s wise to ignore the order of candidates in early horse-race polling, which, as he points out, changes constantly (as it did in 2012 when even Michele Bachmann and Herman Cain led such polls at one point in the cycle), but do pay attention to the internals. And to stress a point where I may be just about alone in the punditocracy: even early general election trial-heat numbers may matter for candidates whose appeal in their own party is attributable to their claims of electability.
This is already a real problem for Jeb Bush, since the Establishment’s reported belief that he’s the strongest candidate to send up against Hillary Clinton isn’t born about in polling of these two extremely well-known polls; and it’s a potential asset for Rand Paul, whose otherwise unlikely candidacy has been strengthened by consistently stronger showings than anyone else in trial heats against HRC.
There’s a couple of other things about Nate Cohn’s take that give me pause. He concludes:
At some point, Mr. Walker, Mr. Bush and Mr. Rubio will need to take the lead in the polls, particularly in Iowa and New Hampshire. But now, it’s better to focus on the fundamentals — whether the candidates appear to hold the support from party elites necessary to win the nomination, whether they are broadly appealing throughout the party, and whether they seem capable of building support in the early states.
First of all, I just have to groan when I see yet another meaning assigned to the term “fundamentals,” by which some people mean GDP numbers alone, while others would add other economic statistics, presidential approval ratings, characteristics of the cycle, and landscape. And second of all, where do we find these fundamentals this early in the contest, particularly such criteria as “whether they are broadly appealing throughout the party, and whether they seem capable of building support in the early states”? That’s right: early polls, properly interpreted.
So: turns out there is gold to glean from early polls, so long as you know where to spot the fool’s gold.