Morning Jay: A Primer on the 2012 Polls
6:00 AM, Mar 18, 2011 • By JAY COST
The invisible phase of the presidential campaign is upon us, as prospective GOP nominees are travelling to Iowa, New Hampshire, and South Carolina, meeting with donors, and of course making appearances on Sunday news programs to deny that they have any interest in the party nomination. And with all this nonsense (for which we can pin much of the blame on George McGovern!), we are also now getting a stream of election polling. Right now, it is but a stream, but soon it will turn into a raging river -- with multiple polls coming out every day telling us which Republican is up, which is down, and which has been left for dead.
Before we get to that point, we'd better learn how to navigate those treacherous waters, and so today's Morning Jay is meant as an initial primer on the 2012 polls.
To get us started, let's flash back in time, to the Spring of 1967. The Monkees are at the top of the charts, Gilligan's Island is ending its run (spoiler alert: they don't make it home), and New York governor Nelson Rockefeller has quietly thrown his support (and his money) to Michigan governor George Romney, who has already announced the exploratory phase of a campaign to seek the GOP nomination. In May of that year, Gallup dutifully conducted a poll asking Americans whom they would support -- Romney or incumbent Lyndon Johnson. Romney had a slight edge -- 46 percent to 42 percent.
Can you imagine how that result would have set all the tongues wagging back then if 1967 had the kind of political class that we have today? My goodness! The chitter chatter of the talking heads on the cable news would have been nonstop! As it turned out, neither Romney nor LBJ ended up as nominees (or, for that matter, contenders for the nomination), so all of that analysis would have been totally forgotten in a year.
The more prudent approach, of course, is not to be put in the kind of position where all those old "insights" have to be abandoned, which is why we need to be really careful with the early polling on 2012. It can lead us down a lot of blind alleys. The mass public at this point is paying virtually no attention to the campaign, which is mostly happening behind the scenes anyway, so polling conducted at this early stage is really not worth very much.
With that in mind, here are three basic ground rules for watching the polls over the next 6 months or so:
(1) Ignore the head-to-head GOP primary matchups. After the experience of 2008, one would think that we wouldn't have to remind the pundit class of this. Remember how wildly the polls swung that year? Here's the Democratic nomination polls over the course of 2007 and 2008:
Through all of 2007, Hillary Clinton had a very substantial lead in the national polling, so much so that MSMers reported that Obama supporters were fretting that he had been misallocating his resources. The Republican nomination polling had a similarly strange arc, with McCain falling to as low as 10% in the RCP average around Labor Day. Rudy Giuliani enjoyed a national lead all through the year. We all know how well that worked out for him.
There are two reasons these numbers were problematic last cycle, and might be once again. The first, as mentioned above, is that voters aren't paying a heck of a lot attention through most of the pre-primary season. This is a big reason why historically we have seen a phenomenon called "momentum," wherein a victory in one early contest helps a candidate win another, then another, and so on. Second, there is no national primary -- so these polls aren't predicting any actual event. What matters is how candidates perform in each contest.
(2) Be wary of any polling on candidates with low name recognition. Consider, for instance, this untenable analysis from Public Policy Polling on Tim Pawlenty's standing:
This is not "bad news" for Pawlenty whatsoever. When a candidate's name recognition is so low (less than half of voters have an opinion), it's often the case that just the core partisans are registering any opinion at all. And, sure enough, Pawlenty's under-water numbers are largely due to the fact that, for whatever reason, more Democrats given an opinion on him (55%) than Republicans (39%) or Independents (48%).
With this in mind, reconsider the second quoted paragraph. Pawlenty is losing among voters who have an opinion about him because they are largely Democrats at this point. How PPP can look at this data and suggest that the numbers don't "bode well for Pawlenty" is beyond me. Especially when later on, PPP comments:
Then how can they draw any inferences about where Pawlenty stands as a candidate?
All in all, this should serve as a good example of why we need to be careful in how we digest the polling information we're fed.
(3) The number to watch on Obama is his "deserve reelection" number. Later on in the cycle, when the primary campaign heats up and voters are paying at least a little bit of attention, the head-to-head matchups (e.g. Obama v. Romney, Obama v. Daniels) will give us an early sense of which Republican nominee will perform the strongest. But those head to head numbers are not worth very much yet; right now, they're mostly based on partisan affiliations and whatever spare bits and pieces of information Independent voters have put together about the various Republican nominees. Going back to that PPP poll, this should help explain why Romney does better against Obama (42-47) than Pawlenty does (33-47). It's not necessarily that Romney is the stronger candidate, but that he is better known (and, on that point, note that Obama's numbers are the same in both polls).
Right now, the metric to monitor is "Does Obama deserve reelection?" Here's Republican pollster David Hill on that one:
Exactly. And, given the dearth of polling on this question, the best proxy is probably Obama's job approval rating. If you don't think the president is doing a good job, you probably don't think he deserves reelection (although you might still have an open mind about it).
As of this writing, Obama's job approval rating in the RCP average is at 48.9 percent, and it has been under 50 percent for most of the last 18 months. That's not good for the president.