Are there any swing voters anymore?

The answer over the past few years has seemed to be no. Both the Bush and Kerry campaigns in 2004 were run on the assumption that there were few swingable voters and that the prime task was to increase the turnout of your own supporters. Both succeeded, one more than the other: John Kerry's popular vote was 16 percent larger than Al Gore's, and George W. Bush's 2004 popular vote was 23 percent larger than his popular vote in 2000.

Now comes a well-known pollster, Mark Penn, to assert in the Washington Post that there really are a lot of swing voters.

It should be noted that Penn has been Hillary Rodham Clinton's pollster and was Bill Clinton's pollster in 1996. He's an interested party. There has been talk that Senator Clinton is unelectable in 2008, and some have cited a poll that shows that 51 percent of respondents wouldn't vote for her. Penn has a vested interest in promoting the idea that Clinton is electable, and that there are enough moveable voters to make a Clinton '08 victory possible.

Still, I take his point seriously. Penn, in my experience, has been a shrewd and intellectually honest analyst of public opinion. And while his analysis in my view doesn't hold up for 2000, 2002, and 2004, it may be valid for 2008. One of the reasons we have had such polarized politics and such steady voting patterns—Republicans won between 49 and 51 percent of the popular vote in House races between 1996 and 2004, Democrats won between 46 and 48.5 percent—is that the major political figures of these years, Bill Clinton and George W. Bush, have been intensely polarizing. Both of them happen to have personal characteristics that people on the other side of the cultural divide absolutely loathe.

But that may not be true in 2008. In his column, Penn points to the last time that opinion changed sharply, 1995-96. In 1995, a majority said they wouldn't vote for Clinton; in 1996 a significant plurality did (though not a majority: Clinton beat Bob Dole 49-41 percent). Since then opinion has seemed to remain frozen. But 2008 will give us a different cast of characters. On the Republican side, John McCain and Rudolph Giuliani currently lead in the polls. Neither polarizes the electorate in the way that George W. Bush does. On the Democratic side, Penn's client Hillary Rodham Clinton currently leads in the polls. She tends to polarize voters a lot, perhaps as much as her husband.

But that may not be permanent. She has compiled a voting record that can plausibly be characterized as moderate, at least in comparison to her party's flaming liberals, and she has staked out issues to emphasize, which can moderate her image. It's pretty clear that that's the course Penn has urged on her, as he did with notable success on her husband in 1995-96. It's not inevitable, though it's certainly possible, that this effort to make her appear more moderate will fail. So it's possible that the electorate will be more fluid, with a much larger percentage of moveable voters, in 2008 than it was in 2000 or 2004.

Some of Penn's evidence is not convincing. He points to higher percentages identifying themselves as independents but doesn't note that most of them, their self-labeling notwithstanding, have been voting almost entirely for one party or the other in recent years. But politics doesn't stay frozen forever. When one of the architects of the last significant movement in the partisan balance, the Clinton recovery of 1995-96, says that another may be in store, he should be given respectful attention. Current polling showing John McCain far ahead of Hillary Clinton suggests that a change of alignments would work against his candidate. But maybe change in the other direction is possible, too.