I do disagree with Bartow that the five votes to overturn Roe “are already there.” In particular, I don’t agree with her claim that Kennedy “has been moving against abortion over time.” I don’t see how his position has changed at all. The plurality opinion in Casey created (as Devins notes) a regime of legal-but-regulated abortion; Carhart II isn’t inconsistent with that. And while bans on “partial-birth” abortion are idiotic, they also have less impact on access to abortion than the waiting periods and parental involvement requirements upheld in Casey. Particularly when you consider his very strong endorsement of the right to privacy in Lawrence, I think the odds that Kennedy would be the fifth vote to overrule Roe are nil.
In addition, I also disagree with the essentially functionalist account of Casey advanced by both Bartow and Devins. Both see Casey as a product of social and political forces that perhaps caused the median justices to vote against their true preferences. But the upholding of Roe was very much contingent; with exactly the same political and cultural context it could well have been overruled. Had Reagan just nominated Scalia and Bork in reverse order, or Bush I had nominated Ken Starr rather than Souter, Roe would be gone. And I think this mattered a little more than Devins assumes. It’s true that majorities favor abortion rights, but a number of state legislatures would have almost certainly passed abortion bans had the Court permitted them.
At any rate, I do agree with Devins that Roe is probably safe in the short term, and certainly isn’t immediately threatened should Obama win. On the other hand, I don’t agree with him that a court with a more conservative median vote would reject abortion regulations that push the envelope. Roberts and Alito might not want an opinion overruling Roe explicitly, but I don’t think they will ever vote to find an abortion regulation unconstitutional, and as Carhart II proves the current “minimalist” court will go to ridiculous lengths to pretend it’s not overruling precedents it clearly is. Moreover, politics can change quickly, and given the relative ages of the pro- and anti- Roe forces on the Court there’s unlikely to be much margin for error for quite a while. The 2008 election really does matter, and a substantive right to abortion will not be on safe ground for quite a while after that. Casey did mirror (for better or worse) national median opinion quite well, but the Court could have plausibly have gone against it before and could do it again.