The Slippery Art Of Abortion Polling

A U.S. senator recently told journalist/author Melinda Henneberger: “Make no mistake: This is a pro-choice country, period.” 

Many in the media believe that and cite polls to back it up. But Henneberger disagrees. She offers as evidence a March 2007 CBS News/New York Times poll in which 41% of respondents favored “stricter limits” on abortion and an additional 23% said it should not be permitted at all.

Henneberger goes even further, contending that “Pro-Choice Is a Bad Choice for Democrats” (Op-Ed, The New York Times, June 22, 2007, A21) after having talked with women in 20 states over a period of 18 months. She wanted to learn women’s opinions on major political issues and how their views affect their votes. If They Only Listened to Us (New York: Simon & Schuster, 2007) presents her findings. 

As a former reporter for The Times and contributor to the online journal Slate, which favors abortion rights, Henneberger cannot be dismissed as a right-wing anti-choice ideologue. Her book shows her to be an objective journalist, as well as a very good and inquisitive listener. These traits led her to discover much that her colleagues have missed.

Recall, for example, the so-called “Catholic swing vote” that favored Democrat Al Gore in 2000 (50%-47%), but swung to Republican George Bush in 2004 (53%-47%). Conventional wisdom attributed the 6-point shift to “security moms” who believed the GOP would better protect the country from terrorist attack.

Henneberger found that “what first-time defectors [Democrats voting for the GOP presidential candidate] mentioned most often was abortion.” Time and again, she encountered self-described Democratic women who (reluctantly) voted for President Bush in 2004 because of their own opposition to abortion.

Poll results showing pro-abortion majorities often suffer from a variety of errors in design. A common error is to over-sample demographic groups who lean pro-abortion. An otherwise excellent poll by Ayres McHenry & Associates (April 26-May 2, 2007) may have unintentionally understated pro-life strength by having a much larger percentage of Democrats (33%) than Republicans (28%) among respondents. Always read the demographic breakdowns in the “fine print” of polling data to see if political parties were fairly represented among the respondents.

Another common problem is the inclusion of misleading information in the question. For example, the Supreme Court’s Roe v. Wade decision appears more moderate if we say that it “established a woman’s constitutional right to an abortion, at least in the first three months” (Pew Research Center, July 13-15, 2005; see AP/Ipsos polls and others for a similar description). In fact, Roe also made abortion legal in the second three months and even the third three months of pregnancy! 

Many polls (e.g., CNN/USA Today/Gallup, ABC News/Washington Post) ask vague questions — subject to a variety of interpretations — which prompt middle-of-the-road responses. They ask, for example, whether abortion should be legal under most [or a few] “circumstances” or “legal in most [some, all] cases” vs. “illegal in most [some, all] cases.”

In the abortion context, words like cases, circumstances, most, some, and few can mean very different things to different people. For example, do we mean most abortions, or most of the different justifications one might imagine (even if most of them might be implicated in a miniscule percentage of abortions)? And given the pervasive moral relativism in American culture, respondents are apt not to admit to “abortion extremism.”

Instead, large majorities respond some, most or generally, rather than never or always. A February 2007 Washington Post poll, for example, found 70% of registered voters answering “legal in most cases” or “illegal in most cases,” with only 16% preferring abortion to be legal in all cases and 12% preferring abortion to be illegal in all cases. These results tell us very little about support for Roe or what kind of abortion policy people prefer. 

A favorite question asked by some pollsters is: Do you agree that “the choice of abortion should be left up to the woman and her doctor”? If the issue is framed this way, most Americans will side with the individual against a government-imposed intrusion into “private” life. Between a woman and Congress, a woman will win every time. But this does not mean Americans support sucking out the brains of a developed baby trying to be born. 

Another favorite is: “Would you like to see the Supreme Court overturn its 1973 Roe v. Wade decision concerning abortion, or not?” This question assumes (wrongly) that respondents know what Roe mandated and what the impact of its being overturned would be. Which is more likely: that the 66% who answered “no, do not overturn” (CNN/USA Today/Gallup, January, 2006) support abortion on request throughout pregnancy? Or that most of them do not know Roe is that extreme? And how many wrongly assume that overturning Roe would mean outlawing all abortions throughout the United States?

Advocacy groups exploit such misunderstandings to portray a future in which women are sent off in droves to die from “coat hanger” or “back alley” abortions, as the only alternative to the status quo. 

Here’s how we know that Roe is not supported by 66% of Americans: Polls with carefully-worded, neutral questions about allowing abortion in identified circumstances show minority and waning support for the policy of Roe. These are better measures of public opinion on whether abortion law should change. 

An April 2005 poll by the polling company inc., offering respondents six choices, found only 10% support for what Roe actually does.

The Ayres McHenry poll, mentioned earlier, tested the hypothesis that when people are informed about Roe’s extremism, many change their opinion from “uphold Roe” to “overturn Roe.” Pollsters first questioned registered voters’ opinions of Roe in “red states” (where a majority voted for Bush in 2004) and “blue states” (where a majority voted for Kerry in 2004). Red state voters opposed overturning Roe by a 50 to 39% margin; blue state voters opposed overturning Roe by a margin of 61 to 27%. Overall, 55% did not want Roe overturned and 34% did. Respondents were then asked if abortion should be legal or illegal under 12 specific cases: the life of mother is at risk, the pregnancy is a threat to the mother’s physical health, it resulted from rape/incest or fetus has a severe physical or mental deformity.

Support for abortion in these cases ranged from 55% to 75%. In the next eight cases, however, an often overwhelming majority wanted abortion illegal: sex-selection (79%); child would interfere with education/career plans (72%); child has a correctable physical abnormality (66%); woman cannot afford to raise child (65%); has all the children she wants (64%); not ready to raise a child (63%); she is not married (62%); pregnancy could cause depression or other mental health problem (51%). When told by the interviewer that all those cases are legal under Roe, support for Roe dropped 7 points and opposition to Roe grew by 9 points (48% support, 43% overturn). A plurality of red state voters (47% to 44%) would now like to see Roe overturned (a 14-point shift against Roe); a smaller majority of blue state voters still supported Roe (53% to 39%), but that represents a 20-point change.

Similarly, an April 2004 poll by Zogby showed 56% of Americans taking a strongly pro-life position (18% never legal; 15% legal for mother’s life only; 23% legal only for mother’s life/rape/incest). Younger Americans were even more pro-life than older Americans: among 18-29-year-olds, 60% took a pro-life position, including 26% who said “never legal.”

What of polls that ask simply whether the respondent is “pro-life” or “pro-choice”? Unfortunately, answers may be tainted by respondents’ personal perception of pro-life individuals (people who care for the well-being of both mother and child, or self-righteous, judgmental nut cases who bomb clinics?) and their personal perception of Planned Parenthood and abortion doctors (professionals who help women out of unfortunate circumstances, or butchers who kill children for money?).

In the 2004 Zogby poll, for example, while 56% took strongly pro-life positions, only 49% identified themselves as pro-life (vs. 45% pro-choice). And yet 61% (and two thirds of adults 18-29) would outlaw abortion after an unborn child’s heart has begun to beat (22 days’ gestation), and 65% would outlaw abortion after fetal brainwaves are detected (40 days’ gestation). The number of respondents identifying themselves as pro-life has increased substantially in the past decade. Overbrook Research reviewed data from statewide polls in Missouri taken between 1992 and 2006. Authors Christopher Blunt and Fred Steeper published their findings in an article called “Turnaround on Abortion” (May 2007).

They attribute the large increase in self-identified “pro-life” voters and the corresponding decrease in “pro-choice” voters to two developments in particular: Images of violent clinic protesters have all but disappeared from the news, and the appalling reality of abortion practice has been dramatized by both ultrasound imaging and the debate on partial-birth abortion.

Conclusion: We have come a long way in the last 15 years. Don’t fall for the phony polls and abortion lobby claims that abortion is the choice of women and highly educated people, that only men and high school drop-outs are pro-life. With gratitude for our children, we recognize that those who survived the Roe regime will be the ones leading us to a pro-life future!

Susan Wills is associate director

 for education for the U.S. bishops'

secretariat for Pro-Life Activities.