Much of the DLC/centrist strategy is informed by a single fact of public opinion: far more Americans self-identify as "conservative" than "liberal." For example,
see this much-discussed new study.
To see why reliance on this factoid is overly simplistic and likely counterproductive, follow me across the jump into the realm of political science. . .
Imagine a pollster asking you the following question: "In general, do you consider yourself a ButtWanker or a SuperEagle?"
Although you have no idea what the terms actually mean, your gut-level, affective reaction to them causes you to answer "SuperEagle."
Later, in conversation with your wonky uncle, you find out that ButtWanker is a term, used mostly among opinion leaders and political elites, that describes someone who believes in expanding (or maintaining) the social safety net, and who believes in keeping the government out of the bedroom. Whereas, a SuperHawk believes in drastically cutting the federal government and legislating morality.
The Democratic pundit class fails to understand that the above scenario plays out for a majority of poll respondents every time they are asked whether they consider themselves to be "liberal" or "conservative." Here's some background on what we know about ideological self-ID:
Since the early '60s, scholars (beginning with Phil Converse) have shown consistently that only a small proportion of the population understands that liberal and conservative are two ends of a spectrum, or that they are meant to constrain a diverse array of issues into a common organizing principle. So how does the rest of the public understand the labels? Pam Conover and Stanley Feldman demonstrated that most Americans react symbolically/affectively to the labels. In other words, they pick up a positive or negative valence from them, despite having little to no idea of what they actually mean.
Add to this an important fact of contemporary politics: The labels are discussed asymmetrically in elite discourse, meaning only conservatives self-identify with a label. For a couple of decades, Republicans have made a concerted effort to stigmatize the word "liberal." Instead of fighting back, liberals/Dems ceded the label to them, with their rallying cry becoming "I'm not THAT liberal" (see esp. Dukakis, 1988). So, in typical campaigns from coast to coast, the Repub says "I'm conservative, my opponent is liberal," while the Dem either says nothing, or says "I'm not liberal, I have Ohio values," or whatever.
The net result of this asymmetry is the ButtWanker/SuperEagle effect: liberal is a bad word, while conservative carries positive affect. But, to tie it all together, the crucial point is that the conservative advantage in ideological self-identification is NOT necessarily a reflection of actual conservatism in the electorate.
So how liberal or conservative is the electorate in reality? This is harder to measure, and subject to dispute. Jim Stimson, one of the most accomplished contemporary political scientists, has accumulated a wealth of evidence that a majority of the electorate is "operationally liberal," at least on issues of government spending and priorities. Many critiques can be leveled at this, including his reliance on spending issues, but his analysis is very impressive and shouldn't be pushed aside arbitrarily.
At the very least, we know this: the public MUST be more liberal than the ideological labels indicate. This inference is a necessary logical outgrowth of the ButtWanker/SuperEagle principle -- that is, 1) most people don't know what the labels mean, but 2) they know that liberal is bad.
As noted above, the DLC/centrist class hitches its wagon to the notion that conservatives dominate liberals in ideological self-identification, usually with the implication that the party needs to embrace conservative principles, or at least find a way to appeal to people who take that label.
But I would argue that this logic is flawed, in that ideological self-identification is of very limited use in understanding the electorate's actual ideology. So unless we plan to make a serious attempt to reclaim the label or to move toward being known as the "conservative" party, it would behoove strategists, both professional and amateur, to stop crafting plans of action around a meaningless poll result.
Cross-posted long ago at TPMCafe. No link -- I'd rather it just fade away over there.