A Terrible Misunderstanding: Why the Polls Distort Roe v. Wade

In June 2000, the Los Angeles Times released a poll that found Americans “evenly split” regarding Roe v. Wade, one of the two 1973 Supreme Court rulings that created a constitutional right to abortion. Forty-three percent of respondents indicated that they approved of the decision, while 42 percent disapproved. At the time, the Times’s poll was not alone in claiming that Americans were narrowly divided over Roe. In January 1998, an Associated Press survey revealed that 47 percent of the public favored the ruling, while 43 percent rejected it.

Yet by mid-2005, when two vacancies on the Supreme Court opened up, media polls found a much different set of results. Far from being closely divided, the public was said practically to embrace Roe. As an October Pew survey concluded, “The public has long been strongly supportive of the landmark Roe v. Wade decision….” Indeed, from May to August, CBS, Pew, Gallup, and NBC/Wall Street Journal found that 59 to 65 percent of respondents approved of Roe.

Both sets of polls—besides containing adequate sample sizes—used language that pollsters consider neutral. Nevertheless, the results varied significantly. This apparent contradiction naturally raises a couple of questions: What does the public think of the abortion standards laid out by the Supreme Court? And, if the earlier polls are correct, how can later polls seem to show that most Americans want to preserve Roe?

The answers matter in ways you might not expect. In recent decades even the Supreme Court has misread poll results. In Becoming Justice Blackmun, a 2005 biography of the author of the Roe decision, New York Times reporter Linda Greenhouse revealed that Blackmun—while doing his research in the summer of 1972 consulted Gallup polls about public attitudes on abortion.


If the justices do scrutinize the polls, they might be in for a shock. Most Americans oppose the lax abortion standards laid out not just in Roe v. Wade, but in its companion case, Doe v. Bolton, as well as the 1992 decision that modified Roe, Casey v. Planned Parenthood. (This isn’t to say that most Americans favor overturning those decisions—the public appears evenly split on the matter.)

The fact is that the vast majority of media polling on Roe v. Wade is deeply flawed in its confusion over what the Supreme Court actually permitted under Roe and Doe (and later Casey). And that’s why the two sets of polls recorded dramatically different results.

Baby, You’re on Your Own

To understand why most Americans reject the Supreme Court’s abortion standards, it’s necessary first to describe what each of the rulings sanctioned. Roe’s standards are fairly well known: In the first trimester, no restrictions; in the second trimester, the state can regulate abortion “in ways that are reasonably related to maternal health” but not ban it; and in the third trimester, the state can ban abortion except when there’s a threat to the mother’s life or health. At first blush, Roe sounds as if it balances the interests of the mother and the pre-born human; indeed the ruling, as section VIII asserts, rejected the claim that “one has an unlimited right to do with one’s body as one pleases.”

But unknown to the public is Doe v. Bolton, the companion case to Roe that radically defined what the Supreme Court meant by “maternal health.” At issue was a 1968 Georgia statute that, among other things, defined health as encompassing a woman’s physical and mental states; it also allowed abortion in cases of rape and fetal defect. The court, asserting that the health provision served neither the interests of the doctor or woman, struck the law down. It then decreed that health must encompass “all factors—physical, emotional, psychological, familial, and the woman’s age—relevant to the well-being of the patient.”

The implication of this standard—let’s call it PEPFA (physical, emotional, psychological, familial, and age)—is hard to overstate. It allowed a pregnant woman to abort her pre-born child for practically any reason remotely tied to health. Even a woman’s age and family situation were considered sufficient grounds for abortion. For as section VIII of Roe elaborates on the instances in which a woman might decide to abort:

Maternity, or additional offspring, may force upon the woman a distressful life and future. Psychological harm may be imminent. Mental and physical health may be taxed by childcare. There is also the distress, for all concerned, associated with the unwanted child, and there is the problem of bringing a child into a family already unable, psychologically and otherwise, to care for it [sic]. In other cases, as in this one, the additional difficulties and continuing stigma of unwed motherhood may be involved.

When Roe and Doe are read together—as Roe specifies they are to be in section XI—the Supreme Court demolishes any claim of having established reasonable abortion standards. The rulings allowed women to abort a pre-born human at any stage of pregnancy, often for economic and social reasons; indeed, one struggles to imagine instances in which a woman could not abort. As Supreme Court reporter David Savage of the Los Angeles Times pointed out in a September 14, 2005, story, the twin rulings “made virtually all abortions legal as a matter of constitutional right.”

Nineteen years later, in its 1992 Casey decision, the Supreme Court changed its abortion standards only somewhat. States were allowed to enact a handful of regulations—chiefly, that a woman seeking an abortion give her informed consent and that she be well-versed about the procedure 24 hours beforehand. Also, Roe’s trimester framework was ditched in favor of fetal viability (which typically occurs between the 21st and 24th weeks). But the Court let stand the radical PEPFA standards of Doe. This was a crucial decision, as later became apparent. In 2000, the Court in Stenberg v. Carhart struck down laws in 31 states that enacted bans on partial-birth abortion because they failed to contain “adequate” (i.e., extremely broad) health exceptions. Thus, the Court now permits abortion at nearly any time for virtually any reason.

Women First, Children Second

Most Americans—without quite knowing it—have always rejected the abortion standards in Roe, Doe, and Casey. To be sure, the vast majority of the public doesn’t want to ban abortion altogether. Americans do put the woman before the baby, as it were. But unlike the Supreme Court, they don’t abandon the baby altogether.

This “women first, babies second” attitude has defined public opinion for decades, as even pro-choice observers once acknowledged. In the February 12, 1971, issue of Science, the esteemed LIC Berkeley sociologist Judith Blake, a self-described foe of “repressive…pronatalist policies,” wrote a comprehensive survey about polls regarding American attitudes toward abortion during the 1960s. Based on those surveys, she concluded, “If the Supreme Court became progressively involved in ruling on constitutionality of state legislation concerning abortion, for example, a decision on its part favoring elective abortion would not accord with the view of over 80 percent of the population.” (Interestingly, Blake found that American attitudes toward abortion changed only somewhat after the advent of the sexual revolution in the mid to late 1960s).

Granted, “elective abortion” is different from abortion-on-request; it’s incorrect to conclude, therefore, that 80 percent of the public opposed Roe and Doe. Yet no honest observer claims that the public supported making virtually all abortions legal, especially those for economic and social reasons. In the 1960s, not a single state in the country repealed its abortion laws. In 1970, four states traveled down the road of repeal; Hawaii and Alaska permitted abortion on “non-viable” pre-born humans (which, at the time, was the 24th week of pregnancy), while New York allowed abortion up through the 24th week and Washington the 17th week of pregnancy.

No other states, however, followed that path. In November 1972, Michigan and North Dakota voters defeated, by 61 percent and 75 percent, respectively, measures that would have legalized abortion up to the 20th week of pregnancy. In 33 other states, as Greenhouse notes in her book, officials “continued to outlaw nearly all abortions; in many of these states, there was little prospect for change.”

What some of the public did support was a middle ground on the abortion issue. In 13 states, such as Georgia, a woman could get an abortion under three or four conditions—the exemptions were typically for life of the mother, rape, incest, and a serious threat to the mother’s health. In other words, the state banned abortion for economic and social reasons the reasons given by the majority of women today who decide to abort. If the Supreme Court had nationalized those standards, Americans by and large would have supported them; the abortion controversy would have been, though not silenced, reduced to a murmur or whisper.

But on Monday, January 22, 1973, the Court radicalized our laws. Besides legalizing virtually all abortions, it invalidated the laws of all 50 states. No wonder, then, that abortion-rights activists exulted after the twin rulings were handed down. Author Lawrence Lader said the decisions “went beyond what anyone had predicted,” while NARAL executive director Lee Giddings wrote that they represented “a staggering victory.”

After Roe and Doe were announced, American attitudes toward abortion, reflecting a bandwagon mindset on the part of some, did liberalize. Even so, most of the public has never supported two key features of the rulings.

First, Americans have never backed most of the PEPFA health standards under Doe. To be sure, for decades the public has strongly approved of abortion if continuing the pregnancy endangers the mother’s physical health. But it’s hard to find evidence for the notion that Americans back abortion for the other four PEPFA standards. For example, consider polls that ask whether abortion should be legal in cases where the parents believe they cannot afford another child; this seems to fit the definition of familial health in which a “family may already be unable, psychologically and otherwise, to care for it [sic].” According to Gallup, 71 percent of respondents disapproved of abortion for that reason in 1962, while 61 percent did so in 2003. Or consider polls that ask whether a woman should be able to get an abortion if she and her partner simply do not want another child, the couple does not want to marry, or a pregnancy would interfere with a woman’s career. According to an in-depth Gallup review of polls in 2002, the average rate of support for legal abortion for those reasons was 39, 35, 29, and 25 percent, respectively.

Just as Americans don’t support abortion for any and all reasons, they don’t support legal abortion at any and all times. Since at least 1975, polls have consistently found that about two-thirds of respondents oppose legal abortion after the first trimester. According to a CNN/USA Today poll from 2003, 68 percent of Americans opposed legal abortion “in the second three months of pregnancy,” and 84 percent opposed it “in the last three months of pregnancy.”

The implication of Americans’ opposition to Doe’s lax definition of health and Roe’s trimester framework is enormous. As conservative writers sometimes point out, the vast majority of the public opposes the vast majority of abortions that are performed. (In 2003, nearly 1.3 million pre-born humans were aborted; since 1973, more than 45 million have been.) In September 2005, the Alan Guttmacher Institute, the research arm of Planned Parenthood, for the first time since 1987 released a study on why most American women decide to abort. Their conclusions underlined the importance of Doe’s Silly-Putty health standards: “While a small proportion of women who have abortions do so because of health concerns or fetal anomalies, the large majority choose termination in response to an unintended pregnancy” (emphasis added). It went on: “Among the structured survey respondents, the two most common reasons were ‘having a baby would dramatically change my life’ and ‘I can’t afford a baby now’ (cited by 74% and 73%, respectively). A large proportion cited relationship problems or a desire to avoid single motherhood (48%).” To put it another way, roughly 95 percent of women decide to abort for reasons that most Americans oppose.

A Terrible Misunderstanding

So how do pollsters understand Roe v. Wade?

The Los Angeles Times appears to be alone in grasping that the decision’s abortion standards are radical. In its June 2000 poll, it asked respondents, “Generally speaking, are you in favor of the Supreme Court decision which permits a woman to get an abortion from a doctor at any time, or are you opposed to that….” (Admittedly, the question overstates Roe’s standards, but only marginally.) Perhaps not surprisingly, the results-43 percent of respondents favored the decision, while 42 percent opposed it—represented, as the 2002 Gallup survey noted, “the lowest level of support for Roe v. Wade recorded in recent years….”

It was the lowest level of support recorded because the rest of the polls misinterpret Roe and Doe. They view Roe v. Wade as a decision that legalized abortion but restricted the procedure, not one that made virtually all abortions legal. Specifically, they fail to recognize Doe’s radical definition of health.

For example, Lydia Saad, senior Gallup poll editor, described Roe this way in the 2002 in-depth survey:

The 5-4 decision [sic] favoring the plaintiff in this case defined a constitutional right to privacy on the basis of which the Court said abortion should be allowed in certain cases. Rather than offering a sweeping decision rooted in the importance of the right to privacy or the value of unborn life, the Court employed a continuum of fetal development along which the conflicting interests of the mother and the unborn child must be balanced at various stages of pregnancy.

No wonder, then, that Gallup’s poll, conducted for CNN and USA Today in June 2005, found overwhelming support (65 percent) for preserving Roe. It described Roe as “the decision that legalized abortion” and asked polltakers if they wanted “the new Supreme Court justice to be someone who would vote to overturn” or uphold it. The implication is that if Roe were overturned, abortion would be completely outlawed.

The Pew Research Center also misunderstands Roe v. Wade. One June 2005 poll described Roe as having “established a woman’s constitutional right to an abortion, at least in the first three months of pregnancy” (emphasis added). It isn’t the only one that described Roe this way. So did an Associated Press—Ipsos poll in November 2004 and a Quinnipiac University poll in December 2004. Again, the misleading question led to strong support (63 percent) for Roe. For as the in-depth Gallup survey noted, “If Roe v. Wade is presented only as legalizing abortion in the first three months, support for the decision is much higher than if it is characterized as making abortion legal throughout pregnancy or for any reason.”

Why Pew described Roe as having legalized abortion “at least” in the first three months is hard to say. On two separate occasions I asked Scott Keeter, director of survey research at Pew, why the organization continues to provide this information. Initially, Keeter acknowledged plainly, “I don’t know the logic of that, other than that the first trimester was an absolute right. I believe that was when the state might not intervene….” During the second interview, Keeter offered two explanations. “What we found is that there are [sic] a wide array of polls, and all of them found the same results that we do.” I then asked why Pew doesn’t describe Roe as it is. At this point, Keeter grew testy: “We can’t incorporate all that [information] into the survey question.”

Other polls misunderstand Roe v. Wade in a different way: They think of it simply as the decision that “legalized abortion.” CBS’s poll, taken from July 13-14, 2005, described Roe as having “established a constitutional right for women to obtain legal abortions in this country,” while CNN and USA Today characterized it as “the decision that legalized abortion.” Not surprisingly, both sets of polls recorded strong support for upholding it. The respective figures were 65 and 59 percent.

The problem with describing Roe v. Wade in this way is not just that it’s technically inaccurate (before January 1973, every state legalized abortion, at least to save the life of the mother). The real problem is that the characterization is vague, to the point of being misleading. If the polls described what the rulings actually did, their results would yield far less public support. As Saad at Gallup told me, “The public generally has a favorable impression of Roe, but when you look at the specifics, you can make a case that Roe is at odds with public opinion and that the public would favor overturning Roe.”

Baby Steps

The disconnect between the Supreme Court’s actual abortion standards and most polls’ interpretations of them isn’t without consequences. Forget for a moment whether Justice Blackmun, in consulting polls about abortion, used them to shape his rulings in Roe and Doe. Consider how this disconnect corrodes trust in the polls themselves. By misunderstanding the issues, pollsters raise unnecessary questions about their motives. After all, Roe and Doe were decided more than three decades ago.

Was it a coincidence, for example, that the polls claimed that the public supported Roe before President Bush nominated someone to the high court? Absent any evidence, proving that pollsters deliberately “cooked” the results is impossible. Gene Ulm, a pro-life pollster with Public Opinion Strategies, largely rejects the charge that pollsters are misleading the public. “Are people releasing selective polls? Yes. When you see a huge variance between the Pew and LA Times polls, you know something’s wrong…. [But] they’re not trying to get a result. They say, ‘It looks right to me.’ It’s a desire to put a poll in a headline. It’s not an in-depth study.”

But others—especially Republican conservatives—disagree. As conservative commentator Hugh Hewitt claimed in a profile in the New Yorker, “Polling is an activism tool. Every time the Washington Post or New York Times runs a poll, they are attempting to influence legislation. They are engaged in activism.” Or as author Matthew Robinson claimed in Mobocracy, a 2002 book on polls and the polling industry, “It seems to be a law of media polling: You can always expect the most underhanded and irresponsible tactics from the media when polls will augment their agenda and frame the issues that favor their allies in times of crisis or controversy.”

Is this questioning of motives necessary? It needn’t be. Pollsters ought to describe the Supreme Court’s abortion standards as they are today: legalized abortion at nearly any time for virtually any reason. But as things stand, the best poll is that of the Los Angeles Times from June 2000. As Keeter noted—correctly, though for reasons different from what he meant—”It’s an outlier.”

Mark Stricherz


Mark Stricherz is the author of Why the Democrats Are Blue: Secular Liberalism and the Decline of the People's Party.