By Randall K. O'Bannon, Ph.D., NRLC Director of Research
In the August 1998, edition of the academic journal Sociological Focus, noted Kent State University sociologist and researcher Raymond J. Adamek offers evidence of a tendency among pollsters to reflect the "pro-choice" perspective in many of the abortion polls conducted by major media outlets.
Adamek's study, "Abortion Perspectives and Pollsters' Questions," examines 1,227 abortion questions asked in 312 polls conducted by major media pollsters between 1986 and 1993,1 looking at the language and context of such questions. "The data," says Adamek, "suggest that in querying the public about abortion rights, in describing the legal and empirical realities of the abortion situation and in seeking the public's reaction to abortion politics and policy, pollsters tended to reflect the dominant pro-choice perspective."
Adamek begins by examining how the pollsters framed the issue. He says that the abortion issue, like many others that the courts are called on to settle, involves a "clash of rights" between the pregnant woman and her "right to choose" and the unborn individual and its "right to life." (Pro-lifers, of course, believe that mother and child are not competitors.)
Those favoring abortion, not surprisingly, want the matter framed exclusively in the perspective of the woman and her rights, Adamek points out. Pro-lifers, at a minimum, want the rights of the unborn child discussed as well.
Adamek examined all questions from the relevant polls during that period that explicitly mentioned the woman or the unborn and the word "right" or "rights." He found 51 questions asking about the woman's right to (choose) abortion and none asking about the unborn's right to life.
The mistake is so fundamental that the reliability of polls themselves become suspect. Adamek declares, "By asking questions about only one side of the rights issue, polls yield an incomplete and skewed picture of public opinion." Such polls often indicate that majorities may feel "women should have the right to choose to have an abortion."
Yet when a rare poll does ask about the unborn's rights, majorities agree that "unborn children should have the right to life." Thus the public is much more ambivalent on the rights issue than the results of many major media pollsters seem to suggest.
Public opinion polls fall short of completeness and accuracy in other ways as well, Adamek makes clear. In 1973, the companion cases Roe v. Wade and Doe v. Bolton made abortion legal through viability for any reason and legal after viability for the sake of "the woman's health," which Doe defined broadly to include "all factors - physical, emotional, psychological, familial, and the woman's age - relevant to the well being of the patient." This is, in essence, abortion on demand for the full nine months of pregnancy.
Nevertheless, Adamek observes, when asking people about their support for the Supreme Court's position in Roe and subsequent decisions re-affirming Roe, "Major pollsters' questions seldom refer to the fact that abortion is legal beyond the first trimester." On 64 occasions between 1986 and 1993 in the polls studied, pollsters misleadingly characterized the Court's Roe v. Wade decision as permitting abortion during "the first three months of pregnancy," though the Court's actual position was much broader. Adamek points out that except for nine questions on viability testing prompted by the 1989 Webster Supreme Court decision, only six times did poll questions acknowledge that later abortions are legal. Again, pollsters lend support to the pro-abortion side by failing to accurately convey the extremity of the "right" to abortion.
Moreover, Adamek notes that while most abortions are performed for social reasons,2 most polls asking about the circumstances under which abortions should be permitted tended to give the impression that most abortions were for physical reasons.3 Despite the fact that more than 90% of abortions are sought for primarily social reasons, Adamek notes that pollsters' list of possible reasons for abortion tended to have either an equal number of physical and social reasons or even more physical than social reasons.
Of 16 lists of possible reasons, or "indications," for abortion presented to respondents by pollsters, only four came close to reflecting the current reality of the abortion situation in the U.S. - - and they still listed one more physical indication than social indication, Adamek writes.
By implying that most abortions are for medical reasons or for "hard cases" such as rape or incest, pollsters are able to make it appear that public approval for abortion is much broader than it actually is.
Confusing the matter further, pollsters often frame the issue by discussing abortion as a decision made between "a woman and her doctor." Of 33 items focusing on the abortion decision, Adamek found that 18 queries (or 55%) mentioned a woman's doctor, despite the fact that fewer than 25% of women ever bring the doctor into the decision-making process.4 This makes a major difference in the way people respond.
Citing two separate studies, Adamek notes how that "mentioning the doctor tends to increase the pro-choice response since it implies medical indications." Thus, again, the nature of the questions tends to color the nature of the responses. Adamek concludes this section of his analysis by making the following observation:
In summary, when dealing with the abortion situation, the majority of questions asked by major pollsters does not reflect the facts that abortion is legal beyond the first trimester, that over 90 percent of abortions are done primarily for social indications and that the vast majority of women do not involve a doctor when making their abortion decision. If questions reflected these realities more often, less public support for the present situation would be found.
Adamek also finds that pollsters frequently "subject pro-life persons and policies to closer scrutiny" than they do "pro-choice" candidates. During election periods, Adamek indicates, pollsters tended to ask how a candidate's position on abortion might affect a respondent's vote much more often when the candidate was pro-life than when the candidate was "pro-choice," implying that pro-life position was more controversial. Between 8/25/91 and 10/22/92, Adamek notes, there were 13 questions in which pollsters asked people how a presidential candidate's position on abortion might affect their vote. Only four times (twice for each candidate) were people asked how Bill Clinton's and Ross Perot's "pro-choice" stances might impact their vote. Yet pollsters queried respondents nine times about the possible impact of President George Bush's pro-life position on their vote.
Despite a substantial amount of pro-abortion activity the first two years of Bill Clinton's administration (Adamek cites 13 actions that could be interpreted as favoring the pro-choice cause), Adamek points out that pollsters asked the public about abortion policy much more often during pro-life Bush's first two years of office than they did pro-abortion Clinton's.
Finally, Adamek draws attention to the tendency of pollsters to question the public about its opinion of Supreme Court decisions when Court decisions are favorable to pro-lifers, while ignoring Court decisions that sustain or expand current abortion-supportive policies. There were several notable Supreme Court decisions- Thornburgh (1986), Webster (1989), Rust (1991), Casey (1992), Madsen (1994), and Scheidler (1994) - during the study period which offered some triumphs to both sides. Adamek notes that pollsters asked the public no questions about aspects of the decisions favorable to abortion proponents (e.g., striking down abortion regulations, establishing buffer zones around abortion clinics), but asked 124 questions about elements attractive to pro-lifers (states being given permission to regulate abortion in limited instances, government being permitted to ban abortion counseling, referral at federally supported Title Ten clinics).
The tendency of pollsters to reflect a "pro-choice," rather than a neutral or pro-life, perspective influences not only the questions asked, Adamek notes in his conclusion, but also allows other questions to go unasked. Polls have been silent, Adamek says, on the public's reaction to such things as "the under-regulation of physicians and abortion clinics contributing to injuries to women."
Adamek argues, . . . when dealing with any controversial social issue, pollsters need to take two steps back and examine the perspectives framing it, being careful not to reflect only one perspective. Emphasizing neutral perspectives or alternating between competing perspectives will help them achieve a more comprehensive and balanced picture of public opinion.
FOOTNOTES
1. Adamek specifies several major media outlets such as CBS News, the New York Times, ABC News, the Washington Post, and the Los Angeles Times, in addition to noted polling firms such as Gallup, Louis Harris, and Yankelovich Clancy Shulman, who conducted or sponsored polls asking 20 or more abortion questions during the examined time frame.
2. Adamek cites a 1988 study appearing in Family Planning Perspectives in which 93% of women listed some sort of social reason, e.g., that she couldn't afford the child, that she was encountering "relationship problems," or that a child would interfere in her life, rather than any medical condition or sexual assault, as the primary reason for their abortions.
3. In the article, Adamek does not fully specify what he considers to be "social" or "physical" reasons for abortion, but presumably the social encompasses a woman's purported inability to "afford" the baby, her reluctance to have a child because of "relationship problems" with the father, the child's interference with her life or career plans, etc. Physical reasons would probably include any threat to a woman's life or health or any indication of a health problem with the baby.
4. Adamek cites eight studies supporting the fewer than 25% figure.