Note: this item is cross-posted from FiveThirtyEight.com

When Pew released a poll earlier this week suggesting that there had been a significant shift in public opinion on abortion in recent months, Nate Silver did a fine post exploring the long-term trends on the subject, and expressing considerable doubt that Pew had discovered anything of great moment.
Well, today Gallup released another survey that seems to parallel the Pew findings, such as they are. And since (1) Nate’s on vacation, and (2) the two polls together are sure to get tons of play in conjunction with the anti-abortion protests at Notre Dame, not to mention Supreme Court speculation, I’ll do a brief post raising a few questions to help tide us over for a while, with particular emphasis on the key questions that pollsters rarely ask on this subject.
First of all, the headline-grabbing finding by Gallup involves its efforts to split Americans into two camps self-identifying as either “pro-choice” or “pro-life.” Aside from all the issues with how these two terms are perceived, this methodology also forces asunder and thus distorts the views of the vast “mushy middle” on abortion policy, which Gallup itself measures at 53%, in a secondary question that divides respondents into three camps (“illegal in all circumstances,” “legal in any circumstances,” and “legal only in some circumstances.”)
Second of all, the purported shift that Gallup reports, showing “more Americans ‘pro-life’ than ‘pro-choice’ for the first time,” is explained in Gallup’s analysis as a phenomenon produced almost solely by Republicans and Republican-leaning independents (e.g., roughly those voters willing to actually vote Republican). Given what’s been happening to support for Republicans in recent months, this finding raises some questions about the sampling techniques, but could also reflect a shift in the perceived threat to the abortion status quo once George W. Bush left office. After eight years of constant excitement among right-to-lifers about getting that fifth vote on the Supreme Court to overturn Roe v. Wade, that prospect is now pretty distant. And instead, Americans have been exposed to a constant wailing of alarms about Obama being “the most pro-abortion president in history,” determined, somehow, to expand abortion rights. That Republicans and Republican-leaning independents might polarize on the subject isn’t terribly surprising or necessarily significant.
But more important than any of these reservations about the Pew and Gallup surveys on abortion is a perpetual problem with public opinion research generally on this topic: it rarely deals with the nuances that matter most in setting public policy or assessing the actual political impact of each party’s positioning.
The nuance that I’ve written about recently deals with the simple fact that Americans seem to care quite a bit why a woman seeks an abortion. And once they are aware of a plausible rationale, anti-abortion attitudes appear to relax.
The best example comes from 2003, at the very height of one of the congressional battles over so-called “partial-birth” abortions. The very same ABC poll that showed 62% of Americans favoring a ban on these much-demonized procedures also showed that 61% favored a “health of the mother” exception, even in these cases.
It’s an article of faith among right-to-lifers, of course, that a “health exception” makes a mockery of any abortion restrictions. And that’s why in a famous moment in one of the presidential debates last year, John McCain sneered and held up “quote marks” when referring to a “health exception.” The public reaction was not positive, indicating that abstract hostility towards abortion may well disguise a more sympathetic attitude when it comes to actual women making actual decisions about a pregnancy. To put it another way, who cares if there’s a shift towards self-reported “pro-life” sentiments, if consistent majorities basically approve of the constitutional and legal status quo?
In any event, it’s maddening that so few polls on abortion get into these sorts of questions. Until they do, we are all entitled to dismiss the big headlines, and rely on hard data like election results to determine which basic direction in abortion policy Americans tend to support. Based on that data, the anti-abortion cause is not doing very well.