Scott Lemieux gets speculative:
Admittedly, this is the kind of counterfactual that’s impossible to prove, but my guess is that if she had voted against the war Clinton would be the Democratic candidate. Given the closeness of the race, her inherent advantages going in, and that the war had to be a liability it’s hard to imagine that she wouldn’t have prevailed without the Iraq albatross. Whether or not Clinton’s support was sincere — I don’t think it really matters — sometimes getting big policies wrong really is politically damaging. (See also the 2006 midterms.) This is evidently a good thing.
I think that’s right. To take the notion that good policy is good politics out of the realm of pure idealism, I’d say the point is that policy that can be seen to have turned out poorly is bad politics. With some bad policies, the costs are either hidden or deferred to the future (or both, as with excessive carbon emissions) in which case bad policy can be excellent politics. But with something like a war that’s going to have a lot of very obvious short-run consequences, it’s genuinely quite helpful politically to have sound substantive judgment.