Ross makes the case:

Hillary may not be the best choice for the Democrats, but she's definitely the safest; I think nominating her more or less guarantees the party 48 percent of the vote, since she's sufficiently tested and savvy and all the rest of it to make a Dukakis or Dole-style wipeout almost completely unimaginable. And in a year when things will (probably) be going the Democrats' way anyway, there's a lot to be said for nominating a known quantity and assuming that, in spite of what Jonah rightly calls the "irreducible core" of anti-Hillary sentiment, the political landscape alone will ensure that her guaranteed 48 percent rises to 51-53 percent by November '08. Whereas Obama and to a lesser extent Edwards both have a higher ceiling, but also a much lower floor, since neither has been through the fire already the way Hillary has (indeed, Obama has never run against significant GOP opposition of any kind), and either one could flame out disastrously in the heat of a general-election campaign.

I don't disagree. Clinton is the hardest candidate to sell but equally the hardest candidate to beat. I think the salient questions are: how dangerous do you think the world now is? And how dangerous is the polarization that Clinton - with say a brutally divisive, Bush-style 51 percent victory - will, to my mind, inevitably deepen in that context? Those are the themes of my essay coming out in the next Atlantic, so I'll shut up now until people can read and respond to the full case for Obama I try to make.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.