James Carville and Mark Penn, who is Hillary Clinton’s pollster, have written an op-ed in the Washington Post arguing that “if [Hillary] runs, she can win.”

But as Matthew Yglesias argues, that sets “a mighty low bar”:

She can win? Sure. Is she more likely to win than are plausible alternatives? The piece doesn’t say. Would she be a good President? Would she be better than the plausible Democratic alternatives? There’s nothing on that at all in the op-ed. Presumably the authors think the answer to at least one of those questions is “yes” but I’d like to hear the argument.

In fact, the evidence is clear that Clinton would be a weak candidate. To the extent that the personal characteristics of the candidate matter (an open question in political science), Clinton is likely to underperform relative to your average Democratic candidate. Let’s consider the evidence.

Carville and Penn tout Hillary’s success in New York, writing that Clinton won her Senate seat in New York in 2000 “by 12 percentage points” despite naysaying from pundits and $60 million being spent against her. But as I wrote before, Clinton matched Chuck Schumer’s 1998 margin while underperforming Al Gore by five points:

[H]er 55%-43% win was not exactly a landslide. As the Almanac of American Politics 2002 points out, Chuck Schumer beat Al D’Amato by an almost identical margin of 55%-44% in the 1998 race for New York’s other Senate seat, and Hillary was riding the coattails of Al Gore, who won the state 60%-35%. According to Barone and company, when you break it down by region, she won New York City 74%-25%, lost in the suburbs 53%-45%, and lost upstate 51%-47%. The latter two numbers are pretty good, but again, compare her to Schumer — he won New York City 76%-23%, lost the suburbs 51%-49% and lost upstate 53%-45%. The figures are almost identical.

The obvious conclusion is that Hillary did about as well as your average Democrat in a Democrat-leaning state. While things could have gone much worse given how polarizing she was, it proves almost nothing about her ability to win over voters in the the battleground states of the industrial Midwest, let alone the South.

Carville and Penn also try to spin Hillary’s weak poll ratings:

Hillary is the only nationally known Democrat (other than her husband) who has weathered the Republican assaults and emerged with a favorable rating above 50 percent (54 percent positive in the latest Post-ABC poll).

Yes, she has a 42 percent negative rating, as do other nationally known Democrats. All the nationally unknown Democrats would likely wind up with high negative ratings, too, once they’d been through the Republican attack machine.

The difference with Hillary is the intensity of her support.

Pundits and fundraisers and activists may be unsure of whether Hillary can get elected president, but Democratic voters, particularly Democratic women and even independent women, are thrilled with the idea.

The Post-ABC numbers that Carville and Penn tout are more grim than they suggest. The headline of the poll is “Clinton Does Well on Attributes, Lacks Crossover Appeal.” It shows that “[a]s many Americans strongly dislike as strongly like her (three in 10 in each case)” and that “42 percent wouldn’t even consider voting for her.” In addition, Clinton has high negatives with the independents Carville and Penn think she can win over — “Clinton’s unfavorable rating among this crucial group is 46 percent.”

It’s also not true that all nationally known Democrats have negatives as high as Hillary. John Edwards, for instance, has a much better favorability profile, with a recent Pew poll putting him at 47 percent favorable, 29 percent unfavorable.

And while Carville and Penn are right that the negatives of other candidates will rise as they come under attack, those candidates still have an important advantage over Hillary — they have an opportunity to avoid becoming such a polarizing figure. As I wrote previously, negative ads are uniquely effective if they draw on pre-existing stereotypes of a candidate:

Negative stereotypes of Hillary have deep roots, and many voters are likely to revert to them once she comes under serious attack. I worked for a Nevada Senate candidate in 2000 (Ed Bernstein) who had similar image problems to Hillary. He was well-known to most Nevadans and had a highly defined, polarizing personality. Over the course of the campaign, we built up his favorable/unfavorable ratings from 21/33 in late 1999 to 44/36 in Sept. 2000, and pulled within four points of our opponent in a DSCC poll. But when the Republicans unloaded a million dollars in negative ads on us, all that went out the window. Voters snapped back to their initial perceptions of Bernstein, his unfavorables spiked over 50 percent, the DSCC dropped us, and the race was over. Hillary is a better politician than Bernstein, but I think the dynamics are likely to be similar. As I’ve said before, a bad economy could put her over the top, but the combination of a polarizing persona and a liberal track record is likely to be devastating to her chances.

Remember, Hillary’s unfavorables are over 40 percent before she has come under serious attack by Republicans. What will happen when battleground states are saturated with negative advertisements? The sky is the limit. Independents will dislike her, and the GOP base will be energized (at a time when it might otherwise be demoralized).

More broadly, the X-factor of the race is not female voters, as Carville and Penn claim, but voter fatigue with the extreme partisanship of contemporary Washington. Bill Clinton won in 1992 and George W. Bush won in 2000 promising to put old conflicts behind us and bring Democrats and Republicans together. We should not underestimate the power of that sort of rhetoric, disingenuous as it may sometimes be. Clinton’s election would seem to guarantee four more years of viciousness to voters, who know that the GOP will be gunning for her from day one.

For all of these reasons, Hillary Clinton is a weak candidate. She can still win in a favorable environment, but compared to a generic Democrat, it’s hard to imagine how she can do anything but hurt her party’s prospects.

Update 7/3 2:07 PM: To put Hillary’s negatives in comparative perspective, let’s see where she stands relative to Al Gore and John Kerry, the two previous Democratic presidential candidates. With 28 months to go before the 2008 election, her favorable/unfavorable rating is 54 percent favorable, 42 percent unfavorable according to the latest Post-ABC poll — a ratio of 1.3:1. By contrast, the Post-ABC poll from July 1998 — the comparable period for Gore — shows that his favorable/unfavorable rating was 54 percent to 26 percent even though he was the sitting vice president. That is a ratio of 2.1:1. And two polls from late 2002 show that John Kerry’s favorable/unfavorable ratings were 31 percent favorable, 7 percent unfavorable and 31 percent favorable, 13 percent unfavorable — ratios of 4.4:1 and 2.3:1, respectively.

To sum up, Hillary Clinton is far more polarizing today than Al Gore was in 1998. And look what happened to Gore.

(Postscript: Another problem with the op-ed, as Greg Sargent points out on his American Prospect blog, is that the Post failed to disclose that Penn is Clinton’s current pollster. Also, see The Atlantic’s insider’s poll from last July for more perspectives on Hillary’s candidacy.

Update 7/3 10:52 AM: Sargent tells me that Atrios caught Penn’s lack of disclosure first.)