The left’s Rasmussen problem isn’t with polling bias – it’s with American democracy
posted at 4:25 pm on January 4, 2010 by CK MacLeod
By now, several thoughtful observers – including Ed Morrissey, Karl at the Green Room, and William Kristol – have responded to a Politico front-page story describing a leftwing attack on pollster Scott Rasmussen. Their replies have explored familiar “shoot the messenger”/”sour grapes” rhetorical territory, have pointed to Rasmussen’s excellent record, and have revealed little inclination to give up on Rasmussen polls just because they’ve lately seemed both to favor and to be marginally helpful to the conservative side. The shared sentiment, if not the shared geographical reality, was summed up by Kristol: “[S]erious people in Washington pay attention to Rasmussen’s polls.”
The most charitable intellectual assumption about the attack is that Daily Kos, Media Matters, and other usual suspects on the left are seeing their political numbers worsen, and, fearing the other side’s momentum, have been feigning ignorance about the differences between Rasmussen’s likely voter-based, sample-controlled models and the models used by other pollsters. Given those differences, it would truly be remarkable if Rasmussen polls didn’t yield consistently different results. Perhaps aware of this fact, and having in the past spoken favorably of Rasmussen’s work, respected leftwing poll-analyst Nate Silver declines to join the frontal assault, and instead argues for a subtle distinction (bold face and italics in the original):
If you’re running a news organization and you tend to cite Rasmussen’s polls disproportionately, it probably means that you are biased — it does not necessarily mean that Rasmussen is biased.
Silver may be closer to a useful truth on this otherwise rather trivial topic, but his use of the word “bias,” it seems to me, either misses the point or simply reflects his own predispositions.
To criticize Rasmussen’s polls as somehow unrepresentative is senseless. They represent what they purport to represent, nothing more and nothing less. Likewise, neither Rasmussen’s nor other respectable pollsters’ approaches are intrinsically ideological – polling results are statistics, not arguments. They do, however, correspond to ideological positions. In this sense, what the leftists are really taking issue with isn’t Rasmussen or even his influence, but rather a mode of self-governance – contemporary American electoral democracy – that his polls successfully reflect in concept, and therefore in their results.
If there is any meaning left in the names we give our two political parties, it may be in the way each embraces one of two broadly complementary perspectives – egalitarian democracy and civic republicanism. In this sense, Rasmussen does do “conservative republican” polls – polls whose design very roughly approximates a conservative view of small-r republican governance – while Gallup and many of the media-sponsored polls of “all adults” or “all registered voters” or “whoever happens to pick up the phone” roughly qualify as “liberal democratic.”
Expansion of the franchise has, after all, long been a liberal and progressive cause. Those on the liberal left who haven’t yet gotten around to seeking the vote for non-citizen residents, felons, and other excluded groups – with longer term designs on representation for animals, plants, artificial intelligences, and the dead (where not already effectively enfranchised) – often believe that a lack of interest or knowledge among voters generally coincides with lack of privilege, and that the only way that underprivileged or alienated sectors of society can gain “fair” representation is to increase their relative turnout. It’s hardly an accident in our own day that the main voter registration and encouragement organizations like ACORN and Rock the Vote are left-liberal through and through, and sooner or later drop all pretenses of non-partisanship.
A further underlying assumption on the liberal left (at least among those who for ideal or tactical reasons haven’t rejected “bourgeois democracy”) is that, if low interest voters were more motivated, they would vote like higher interest representatives of their demographic blocs. In theory, it’s the evil of our current system that these voters don’t recognize a greater stake in elections. Once they did so, they would inform themselves adequately – and naturally recognize their greater affinity to the political left. Indeed, from this perspective, those voters don’t really need to be informed at all except as to the left’s endorsements, since leading leftist intellectuals are always happy to do all of the necessary thinking on any matter. Furthermore, even and especially between elections, a typical leftist view of the American polity is that, regardless of who actually shows up to vote, the opinions of all Americans, regardless of level of engagement or interest, should be taken into account equally.
The underlying conservative view, on the other hand, is that, as any parent knows, handing something precious to someone who doesn’t care about or understand it is often a good way to lose or destroy it. Many conservatives have been and would remain open to stricter voter standards and controls – stronger proofs of identity and residency, demonstrated understanding of the American system of government, fluency in English, respect for the laws, etc. – without going back to property requirements or taking up more radical, conservative utopian proposals.
Our friends on the left will call this perspective paternalistic, with some justification, but conservatives don’t recognize any obligation to reject a position merely because it doesn’t suit a leftist preconception. Conservatives will also point out that truly paternalistic governments have often presided over near-100% turnouts in elections that merely demonstrate and ratify state power alongside the complete absence of political freedom. Even more fundamentally, the conservative view on our national life does not equate it with whoever happens to be in government or what the government does or doesn’t do – but that’s another subject.
For now, conservatives remain more comfortable with a self-selecting electorate, already a fairly “liberal” position in historical terms. The respondents to Rasmussen’s polls of identified likely voters are effectively self-selecting, just like actual voters – that’s the idea. Until and unless the country chooses a way to run its affairs that is more progressive in the way that Daily Kos and Media Matters and possibly Nate Silver typically use the term – by embracing mandated mass participation, for instance – we can say that Rasmussen’s methodological bias is conservative in relation to real existing republican democracy in America. And more power to him, conservatives will add. As to the left: Rasmussen is not your problem; Americanism is.
cross-posted at Zombie Contentions
Recently in the Green Room: