Dormant, for all intents and purposes, for the entirety of Campaign 2008, abortion has made its way back into the forefront of the American political conversation over the last few weeks. Part of this, to be sure, is owed to the fact that Barack Obama, a Democratic President supportive of abortion rights, gave a commencement address at Notre Dame, arguably the most prominent Catholic university in America.
But a large part of it is also owed to an incredibly well-publicized public opinion poll by Gallup, which stated that for the first time in their nearly two decades of polling on the abortion question, more Americans identified themselves as "pro-life" (51%) than "pro-choice" (42%). The poll became big news on the internet, and crept into the traditional media, as well.
A lot of the early criticism of the survey was predicated on the sample of the survey. Gallup’s abortion poll was based on a sample that had equal parts Republican and Democratic respondents. That seemed to strain credulity, since it was directly contradicted by other contemporaneous surveys by Fox News, Pew, and...wait for it...Gallup.
But as Pollster’s Charles Franklin so astutely pointed out, the only true apples-to-apples comparison was to look at this poll in comparison solely to previous Gallup surveys. Even by that standard, the abortion poll was an outlier.
Surely, by over-representing Republicans and under-representing Democrats, Gallup managed to skew the numbers to some extent.
But as responsible, if not more so, for the Gallup numbers was the wording of the question. When you are doing something as sensitive as trying to divine the will of the people, words matter a very great deal.
In an example from the first Persian Gulf War covered in David Moore’s "The Superpollsters", contemporaneous polls from ABC and CBS on public support for the proposed conflict in Iraq were quite disparate—an almost 20-point spread between the two polls. Upon looking at the question wording, however, the discrepancy became a little easier to understand. The ABC poll had simply asked whether the U.S. should go to war at some point after January 15th or not. The CBS poll gave an actual alternative, asking voters whether they supported military action in Iraq, or if they preferred to see the U.S. wait to see if economic sanctions would work instead. By providing a concrete alternative to going to war, the respondents to the CBS poll were considerably less likely to support the war. The SENTIMENT of the public did not change between the two polls.
Only the wording.
Fast forward to 2009. The commonly cited question from Gallup’s poll, on the self-identification of respondents as "pro-life" or "pro-choice", can be criticized rather easily for being far too vague. This simplified dichotomy is next to useless in defining someone’s position on an issue as complicated and fraught with qualifiers as abortion is. As ABC polling unit leader Gary Langer pointed out:
The reality is that most people are both "pro-life" and "pro-choice" (both highly charged terms) at once. Public opinion on abortion is complicated, even conflicted, and heavily dependent on circumstances. Most people think it's between a women and her doctor, but most also object to it on moral grounds; many accept it when it's needed, but not as a casual matter. This has been so for many years.
To show the potential bias inherent in the simplified dichotomy, Fox News also polled on the abortion issue. With a substantially different sample (42% Dem, 30% GOP), they got nearly the same results as Gallup, with 49% identifying as "pro-life" and 43% identifying as "pro-choice."
Most pollsters take a more precise approach to gauging support for abortion rights. Most choose a four-option measurement of abortion rights. On the ends of the spectrum we find complete legality and complete illegality. In the middle of the spectrum we find support with some reservations, or opposition with some reservations.
By that metric, we get very different conclusions that were drawn by Gallup in their simple pro-choice/pro-life dichotomy. In seven studies done in the past twelve months (by four different pollsters), we find that the average level of support for abortion rights (either in total or with some qualifications) stood at just below 54%, with the average level of opposition for abortion rights (either in total or with some qualifications) stood at just below 42%. Not an overwhelming margin, but a clear statement that the public has at least a modest level of support for abortion rights.
(Source for the above data HERE)
What is somewhat surprising, but also instructive, is that Gallup themselves used this metric in the same poll in which they trumpeted the "pro-life" vs. "pro-choice" results. They were not included in the seven-poll study above, because the question wording by Gallup varied slightly from the others. But the bottom line is this--by expanding the options, far fewer people took the position most often presumed to be the "pro-life" position (abortion illegal, no exceptions).
Of course, some would even argue that the four-option metric on abortion rights is a fairly meaningless statistic. What matters most, in the realm of government, is whether there is any public appetite for a change in the existing legal standing first cemented by the 1973 Roe vs. Wade case.
Just days after Gallup published the results of their survey, CNN was in the field with an abortion question, but with substantially different verbiage. This question went to the heart of the public’s desire to overturn Roe. The question posed by CNN was as follows:
The 1973 Roe versus Wade decision established a woman's constitutional right to an abortion, at least in the first three months of pregnancy. Would you like to see the Supreme Court completely overturn its Roe versus Wade decision, or not?
The results? Contrary to the other polls would not even begin to describe it. Only 30% of the respondents took the ostensibly "pro-life" position of wanting to overturn Roe, while 68% were opposed to a Roe reversal.
Conservative critics will point out, with some justification, that the CNN version of the question has some inherent wording biases as well. By adding the "at least in the first three months" qualifier, CNN might change the minds of those who generally oppose abortion rights based on the grounds of the much discussed (but decidedly rare) phenomena of "late-term" (or as the anti-abortion community framed it: "partial birth") abortions. Also, the word "completely" in the question is a pretty charged word, as well.
And, indeed, that is precisely the point. Words DO matter. Is there a two-to-one national majority in favor of abortion rights as framed by current law? Seems doubtful, but conversely, it seems just as unlikely that there is a gathering national majority for the dissolution of Roe.
