The initial take on the allegedly central role of "values voters" in the 2004 election had a shaky empirical foundation: the slight plurality of voters in the NEP exit poll who selected "moral values" as their most important voting issue and who voted heavily (80 percent) for Bush. This line of analysis has come under increasing fire in the recent weeks, as many observers have noted that moral values did not really belong in a list of voting issues like the economy and the war in Iraq and that the NEP exit poll question has not been asked with a values choice before and hence provides no information on any change in this election in the level of values voting.

The latter point is where Christopher Muste picks up the story in his excellent article, "Hidden in Plain Sight: Polling Data Show Moral Values Aren't a New Factor" in Sunday's Washington Post. Muste notes, to begin with:

[T]here's another exit poll that has asked voters about moral values in the past four elections. The Los Angeles Times conducts its own national exit poll. Since 1992, it has asked voters which two issues they considered most important in deciding how they would vote. This year, 40 percent of voters the newspaper surveyed cited "moral/ethical values" as one of their two most important issues. Guess what? That's about the same proportion as in the previous two elections: 35 percent named moral values in 2000 and 40 percent did so in 1996, up from 24 percent in 1992. So this year didn't see an unprecedented surge in values voters rushing to the polls.

I've seen these particular findings from the LA Times polls cited in other articles, but Muste goes on to cite some addtional and very interesting findings from these polls that I have not seen before:

And while Bush strategist Karl Rove must be gratified that the 2000 dip in the turnout of values voters was reversed in 2004, he can't be entirely thrilled by how they cast their votes. The L.A. Times survey showed that moral values voters gave 70 percent of their votes to Bush this year. But that's a drop from 2000, when he won 74 percent. Put another way, 54 percent of Bush voters this year cited moral values -- a decline from the Republican high-water mark in 1996, when 67 percent of Bob Dole's voters named moral values. For Democratic nominees, by contrast, the trend has been up, not down, steadily rising from a scant 9 percent of Bill Clinton supporters naming moral values in the "it's the economy" election of 1992 to 24 percent of John Kerry's voters this year.

Muste goes on to cite other data from the NEP poll and data from a post-election survey by the Pew Research Center that suggest the dominant role of values voters in the 2004 election has been exaggerated and that values voting, in general, should not be narrowly defined by reference to issues like gay marriage and abortion. He concludes:

A large and fairly stable group of moral values voters, whose numbers have been largely consistent over the past three elections, who vote Republican in roughly the same or smaller proportions year after year, who provided no clear winning boost to Bush, and whose idea of what constitutes moral values is hardly uniform. This is a poor fit for the reigning image of a crucial swing vote -- animated single-mindedly by cultural wedge issues -- that turned out in unprecedented numbers to push Bush over the top in 2004. It's time to reel the moral values myth back down to earth.

Amen. I might add, though, that even if values voters weren't important in the way election mythology has indicated, it doesn't mean values, broadly defined, weren't important to voters. Questions of presidential character and of America's role in the world, especially vis a vis the war on terror, are very much bound up with values and affected voters' decisions. But that broad conception of values and voting should't be collapsed to the image of swing voters "animated single-mindedly by cultural wedge issues", as Muste correctly points out.