The Clinton campaign sure picked an unfortunate day to wage the electability argument, with two new polls out yesterday both showing Obama running 5 points better than Clinton against McCain. Those numbers are in line with a host of other recent polls.

The Clinton campaign's argument, set forth by Chief Strategist Mark Penn on a conference call yesterday afternoon, has a few basic points that can be summarized by the following headings in his new memo to "interested parties":

*The GOP Attack Machine Will Redefine the Democratic Candidate; Hillary Has Withstood That Process.

* Sen. McCain Will Run on National Security; Hillary Wins That Argument.

* Sen. Obama's Negatives Will Rise; Hillary's Are Already Factored In.

* The Resiliency of Sen. Obama's Coalition Will Be Tested; Hillary's Coalition Is Stronger.

*Current Poll Numbers Don't Tell the Story of What Will Happen: Sen. Obama Routinely Underperforms While Hillary Overperforms.

All these points are debatable, but let me tackle the last two, because they seem especially specious.

In discussing the relative strength of their respective coalitions, Penn writes:

The grind of a general election will erase the freshness and excitement of the primary season and the success that Sen. Obama has earned in states he has little chance of winning in November will erode. It may even crumble. Sen. Obama will have to fall back on core Democratic voters to stay competitive with McCain. But this is where Hillary has already built a powerful base, with overwhelming support among women, Latino voters, and other stalwarts of the Democratic Party. Hillary's coalition, which has carried her to victory across the country, is a winning coalition against Sen. McCain since it draws from the voters Sen. McCain will need to win.

It's true that Clinton has run better among Democrats than Obama throughout the primary, but it seems unlikely that Democratic voters are going to either defect to John McCain or stay home for lack of enthusiasm for Obama in the fall.

The flip side to this argument, of course, is that Obama has consistently run much better among Independents than Clinton. Even in a state like California where Obama lost to Clinton statewide by 10 points, he won Independents by 24 points.

Given that the current polls show the general election could be a very close race, and given that McCain has such a strong appeal among Independents, it seems clear that Obama's strength with Independents (and even some Republicans) would make him the stronger general election candidate.

As to Penn's final point about performance at the polls, he writes:

After winning the Democratic nomination in 2004, John Kerry vaulted to a 17 point lead over George Bush. Even on Election Day, virtually every pollster said John Kerry would win. It did not happen. Today, commentators are touting a Time poll that shows Sen. Obama faring slightly better than Hillary Clinton against John McCain. Last week, the pundits were using a handful of polls to argue that Hillary Clinton would lose NJ, CA and MA. None of that happened. Instead, Hillary Clinton has repeatedly confounded pollsters and the chattering class by doing better on Election Day than the polls suggested she would -- in NH, MA, NY, NJ, CA, and AZ. Her 2000 election was no different - exit polls suggested she might lose a close race but she ended up winning a landslide victory when the actual votes were counted.

Looking back through the polling data in 2004, I can't find any public polling available at the time showing that John Kerry ever led the race by anything more than single digits (except for one CNN/USA Today/Gallup poll from Feb 16-17 showing him up 12 points) - even after the Democratic National Convention in August. Furthermore, Penn's claim that "Even on Election Day, virtually every pollster said John Kerry would win" seems off the mark as well. Only two of the final round of 14 polls in 2004 showed John Kerry leading (by just 2 points and 1 point, respectively), two showed the race a dead tie, and the remaining 10 had George W. Bush with leads ranging from 1-6%.

More to the point, however, is that while Clinton did over perform the polls in New Hampshire, Massaschusetts, California, etc., Obama has also out performed the polls in a number of contests including Iowa, South Carolina, Georgia, Nebraska, Connecticut, and, most recently, Maine. The Clinton camp will argue that with the exception of Connecticut (and Illinois), Obama's wins have all come in either caucuses or in states driven by heavy African-American turnout. Still, I'm not sure that has any relevance to the point Penn was trying to make, which is that Hillary Clinton's strength is not always captured by the polls. Neither is Obama's.