Conservatives, Republican Primaries, and Governor Palin

Some people, I’m told, have ideas pop into their heads when they’re in the shower, others while driving. Personally, mine occur while running. Most of these ideas are, of course, random bits of silliness which pop into my head for a brief moment only to disappear forever, thankfully, a few moments later. However, on rare occasions, a random idea sticks and festers in my brain and, despite conscious efforts to banish and replace it with another one which, hopefully, requires less mental exertion, it remains. Such thoughts aren’t particularly profound (as anyone who knows me personally will confirm) but, that said, I occasionally feel the need to explore them. This morning, while running, I had such a thought.

Last week there were a couple Rasmussen Polls which seemed to show Governor Palin losing support among Republican voters. One showed her in third place behind Mike Huckabee and Mitt Romney. Another showed her losing in head to head matchups with Huckabee and Romney. At the time I argued, justifiably, that hypothetical polls, even scientific polls, this far from an election, are unreliable at best. I still believe that to be the case. For example, in March 2007 Rudy Giuliani was beating John McCain by 21 points in the Rasmussen Poll for the 2008 GOP nomination, yet less that none year later McCain won.

For that reason, I was not concerned by Governor Palin’s showing in those polls and that remains the case today. Romney’s numbers didn’t seem particularly noteworthy either since there are and always will be a contingent of blue-blood, establishment Republicans from which tedious candidates such as Romney draw their support.

What I did find interesting about the polls, however, is the strong showing by Huckabee. These polls, especially if they hold up for the next 2 years (very unlikely), would lead one to conclude Huckabee is indisputably the frontrunner for the GOP nomination. However, I have yet to hear anyone utter the words "Huckabee" and "frontrunner" in the same sentence. In fact, nobody I know or talk to believes Huckabee has even an outside chance at the nomination, nor do I personally know anyone who says they would vote for him. I include myself in this category as well. Even the pundits generally think it will come down to Palin as the movement conservative vs. Romney (or perhaps Pawlenty) as the Republican establishment candidate. Matt Lewis has noted this, as has Chris Cillizza, to name just two. Again, they don’t even bring up Huckabee as a viable candidate, despite his seemingly strong showing in the polls referred to above.

Further supporting this is Alex Knepper, at Race42012, who wrote the following after attending the 9/12 event in Washington (emphasis mine):

Sarah Palin was without a doubt the unsaid queen of the 9/12 March. Of the many people I interviewed, about a dozen expressed solid support for a Palin candidacy, with all of the others at least open to supporting her.

Many were caught up in the fervor of Joe Wilson’s recent outburst. At moments, it seemed as if every other person was holding a “You Lie” sign. Joe Wilson, indeed, was a name I heard with much greater frequency than Mitt Romney, Mike Huckabee (whose name crossed no one’s lips), and Tim Pawlenty. Floated names did include, however, Michele Bachmann, Mitt Romney, Sean Hannity, and Glenn Beck. But these were not serious propositions.

As someone who teaches statistics, among other things, I know this is all anecdotal. However, even allowing for that, if Huckabee is anywhere near frontrunner status among Republicans, and I know many, wouldn’t at least one of those I know support him? This question has been bugging me for some time and is what I was contemplating as I ran this morning.

Other random bits of information were also swirling in my head. One of these was the recent Rasmussen Poll which showed that an eye-popping 73% of Republican voters believe that Washington GOP leaders have lost touch with their base. Another was the recent Gallup Poll which showed that only 20% of Americans consider themselves Republicans, the smallest bloc in terms of part ID, yet another Gallup Poll indicated that 40% of Americans consider themselves conservatives, which was the largest bloc in terms of political ideology. Further complicating the picture is an August 2009 Gallup Poll, that received little attention, which showed conservatives outnumbering liberals not only at the national level, but the state level as well, even California:

Despite the Democratic Party's political strength -- seen in its majority representation in Congress and in state houses across the country -- more Americans consider themselves conservative than liberal. While Gallup polling has found this to be true at the national level over many years, and spanning recent Republican as well as Democratic presidential administrations, the present analysis confirms that the pattern also largely holds at the state level. Conservatives outnumber liberals by statistically significant margins in 47 of the 50 states, with the two groups statistically tied in Hawaii, Vermont, and Massachusetts.

As I ran with all this seemingly unrelated information coagulating in my brain, an idea, the one I referred to earlier, popped into my head. Suppose Huckabee’s inexplicably strong poll numbers among Republicans are related to the enormous disparity in the number of self-identified Republicans and self-identified conservatives found in the Gallup polls to which I referred? If so, this could help explain Huckabee’s poll numbers despite his lack of grass roots conservative support. Allow me to explain.

We keep hearing that conservatives are leaving the Republican Party in droves and that the primary reason for this is that the establishment GOP is becoming more and more Democrat-light. Rasmussen’s poll, which indicates that 73% of Republicans believe the DC elite in the party has lost touch with the base, is strong evidence of this. We also see this in Governor Palin’s endorsement of Doug Hoffman in NY-23. Even stronger evidence is the fact that even though 40% call themselves conservative, only half, or 20%, call themselves Republicans.

If I received a call from a pollster today, Scott Rasmussen for example, and he asked me to identify my party, I would not identify myself as a Republican. From the above polling data, about half of conservatives would do the same as me. This is important. I don’t know Rasmussen’s precise methodology but I suspect, when he polls for Republican primaries, he excludes Democrats and Independents, using self-identified Republicans for his sample. I further suspect that most of those conservatives who no longer call themselves Republicans (like me) are also those most likely to support Governor Palin. Further, these disaffected conservatives are least likely to support a fiscal liberal like Mike Huckabee or a plastic establishment Republican like Mitt Romney.

In short, the sample may be predisposed to exclude a greater percentage of conservatives who are disgusted with the Republican Party and thus don’t self-identify as Republicans. This would result in an under sampling of those most likely to support Governor Palin and conversely, an over sampling of those most likely to support someone else. This would explain why Huckabee over performs in these polls. Those Republicans who still identify themselves as such are far more likely to be moderate establishment types and, therefore, more likely to eschew a grass roots movement conservative like Governor Palin in favor of a “conventional” choice like Mike Huckabee or Mitt Romney. If my logic is accurate, this is bad news for Mitt Romney. These are the very Republicans he should be dominating, and yet he isn’t.

The large disparity between the number of conservatives and Republicans raises another interesting issue. If a significant amount of grass-roots conservatives have left the Republican Party, as appears likely, what are the implications for the primaries? Many have argued, myself included, that Michael Steele should close the primaries in order to avoid “another McCain”. However, with many conservatives changing their registration from Republican to Independent, we may need to rethink that position.

It’s certainly possible, given the above, that by closing the primaries, we may have a situation where many grass-root conservatives won’t be able to participate in the Republican primaries. If that’s indeed the case, the nominee will almost certainly be determined by the same voters who are pre-disposed to vote for a more conventional candidate than Governor Palin such as Huckabee or Romney. In other words, by closing the primaries in order to avoid "another McCain", we could, ironically, be helping to achieve just that. This is something that should be explored further, possibly in a future post.

To sum this up, it is my hypothesis that, given the fact that many disaffected conservatives are leaving the Republican Party, any polls which attempt to sample those that still identify as Republicans will have a built-in bias against those candidates most likely to be supported by disaffected grass-roots conservatives. Moreover, such polls will tend to favor establishment candidates who are supported by Republicans who think the current status of the party is just fine. Pollsters will certainly attempt to mitigate the sampling error arising from this, but it will be exceedingly difficult to accurately quantify. Time, as they say, will tell.