Komen Coverage Makes Ross Douthat Sad

As you might expect, Ross Douthat is unhappy about the backlash against the Susan G. Komen for the Cure Foundation's decision to defund Planned Parenthood. His argument rests upon assertions of media bias that are shaky since, as Sarah Kilff notes, it's likely that media bias wouldn't have been a factor in Komen coverage precisely because of the political leanings of the average journalist. While it's plausible to assume the typical journalist is more socially liberal (as well as more economically conservative) than the median public opinion, I would argue that this is less true with respect to abortion than with other social issues. Punditry dismissing the importance of Roe v. Wade and reproductive rights, in particular, is so common as to be banal.  

In addition to this argument about media bias, Douthat also cites public opinion data, focusing on "as many Americans described themselves as pro-life as called themselves pro-choice" and that a "combined 58 percent of Americans stated that abortion should either be “illegal in all circumstances” or “legal in only a few circumstances.” John Sides objects to Douthat's cherry-picking:
One cannot divide the public into “pro-life” and “pro-choice” camps based on the kinds of survey questions he cites. These questions fail to capture the true complexity and the ambivalence in most Americans’ attitudes toward abortion.  Most Americans approve of abortion in certain cases and oppose it in others.  Juxtapose, for example, abortion in the case of rape with abortion for the purpose of sex selection.  At best, a small minority—20 percent at most—would approve of or oppose abortion in every case.
While I agree that Douthat's use of public opinion is tendentious, I think the problems are different and worse than the ones that John cites. The most obvious problem is that Douthat combines two survey response categories to create what looks like an anti-choice majority, adding the 20 percent who want abortion banned to the larger number who believe that abortion should only be legal under "a few circumstances."  Since these "circumstances" aren't specified and presumably mean many different things to different people, to combine the two numbers is fundamentally misleading.
I agree with John that many people have an intuitive sense that abortion should be legal for the "right reasons" but not for the "wrong reasons," which is reflected in the public opinion data that shows a great deal of support for abortion only being legal in certain unspecified circumstances. The problem is that these distinctions are completely irrelevant to public policy. There's no way of crafting abortion laws that only makes abortions that women obtain for certain reasons illegal. "Centrist" abortion regulations such as waiting periods or requiring the approval of panels of doctors don't ensure that women will get abortion for the "right reasons"; they just produce contexts in which affluent women can obtain abortions for any reason and poor women—especially those outside major urban centers—find it difficult or impossible to obtain abortions.   
I don't think "women should only be able to obtain abortions if their reasons are good enough" is a normatively attractive basis for abortion policy, but whatever one thinks of the argument it's irrelevant to crafting policies. Getting selective moral judgments mixed up with abortion policy confuses matters in ways that work to the benefit supporters of abortion criminalization. A fair fight between the actual policy alternatives would strongly favor pro-choicers, as the public's overwhelming support for Roe v. Wade reflects.