As Neil points out, one of the main arguments Edwards' supporters have made for his candidacy is that Edwards is a Southern white male with a distinct drawl. "Edwards had the superficial features designed to convince a uninformed general election swing voter, particularly in the white working class, that he was just a good-hearted regular guy from your town. He could take on some pretty solid left-wing policy positions without anybody regarding him as a fringe candidate." And this appears to have worked. A plurality of New Hampshire voters who consider themselves conservative voted for Edwards. But the idea here was always that there would be a massive asymmetry of information, and Edwards would keep fooling these voters while racking up the liberal voters who were in on the secret. Problem is, liberal voters appear to have examined the same information as conservative voters and most decided that Edwards wasn't for them. Where he won 42 percent of self-described conservatives, he took only 16 percent of the "very liberal." So the theory was right: Voters do work off cues rather than platforms. The problem is it applies broadly, not only to a certain subsegment of voters.
