On Jan. 25, hundreds of thousands of abortion opponents assembled in Washington, D.C., for the March for Life. The weather was freezing, but they’re used to that. Roe v. Wade, the Supreme Court case that declared abortion a constitutional right, was decided in January 1973. Every year, pro-lifers hold the march to mark Roe’s anniversary and renew their commitment to overturning it. “It might be 20 degrees out here,” activist Ryan Bomberger told the crowd, “but it has not put out this fire.”
Why do these people come out in the bitter cold, year after year? Because to them, abortion is mass slaughter. “Fifty-five million Americans have died as a result of legalized abortion in the last four decades,” march organizer Jeanne Monahan lamented in her opening remarks. That belief—that horror—keeps pro-lifers marching, voting, donating, and phone-banking.
By most measures, the United States is a pro-choice country. Half of Americans prefer the term pro-life, and most of us have moral qualms about abortion. But 40 years after Roe, most of us still don’t think that ruling should be overturned. Most think abortion should be legal in the majority of cases. Most think the abortion decision should be left to a woman and her doctor.
But the pro-choice majority doesn’t run this country, because a lot of its constituents don’t really care. They don’t march or phone-bank for women’s health or equal rights. They just want to be left alone. They don’t vote on abortion unless they’re scared. When the debate is quiet, it’s pro-lifers who hold the upper hand.
In polls, there’s a glaring gap between the views of the general population and the views of people who care about this issue. With the help of several pollsters, I’ve been looking at surveys with a wide range of intensity levels. A low intensity level, by my definition, is when few respondents—20 percent or fewer, let’s say—report that the issue is very important to them or that they would vote only for a candidate who shares their abortion views. A high intensity level is when more respondents elevate the issue to “very important” or “must share” status.
Across six Gallup polls taken from 1996 to 2012, self-identified pro-choicers consistently outnumbered self-identified pro-lifers. But in each of these surveys, among respondents who said they would vote only for a candidate who shared their abortion views, pro-lifers outnumbered pro-choicers. The pattern held in 1996, 1999, 2000, 2004, 2008, and 2012.
In Gallup’s data, the percentage of respondents who say a candidate must share their abortion views has fluctuated between 13 and 20 percent. What happens when the intensity level goes higher? To answer that question, I looked at surveys by the Pew Research Center. Last month, when Pew asked whether Roe should be overturned, only 29 percent of respondents said yes. But among respondents who said abortion was a “critical issue facing the country,” 62 percent said yes. That matches Pew’s 2009 survey, in which 79 percent of the “critical issue” respondents, but only 47 percent of respondents as a whole, said abortion should be illegal in most or all circumstances. In both polls, high-intensity pro-lifers clearly outnumbered high-intensity pro-choicers. The intensity level was 15 percent in 2009 and 18 percent in 2013.
But in 2006, the only other year in which Pew asked the “critical issue” question, the intensity level was much higher. Twenty-eight percent of respondents said abortion was a critical issue. Among this group, 49 percent said they opposed “making it more difficult for a woman to get an abortion.” Only 46 percent said they favored making abortions more difficult. This appears to be the only poll in which Pew or Gallup has found a pro-choice advantage among high-intensity respondents.
Pew’s data are intriguing but problematic. The “critical issue” question stayed the same in all three Pew surveys, but the pro/con question changed from making abortion more difficult (2006) to making it illegal (2009) to overturning Roe (2013). What we really need is the same pro/con question under different intensity levels. The Quinnipiac University Polling Institute has such a question. In 2007, and again in 2012, it asked, “If you agreed with a presidential candidate on other issues, but not on the issue of abortion, do you think you could still vote for that candidate or not?” And second, “Do you think abortion should be legal in all cases, legal in most cases, illegal in most cases or illegal in all cases?”
In 2007, 22 percent of Quinnipiac respondents said abortion was a deal-breaker for them in choosing a candidate. Among this group, 46 percent said abortion should be legal in all or most cases. Forty-nine percent said it should be illegal in all or most cases—a 3-point pro-life advantage. But in 2012, the intensity level went up. Twenty-nine percent of respondents said abortion was a deal-breaker. And this time, pro-choicers held the advantage: In the high-intensity group, respondents who said abortion should be legal outnumbered those who said it should be illegal, 52 to 45 percent.
Polls taken by CBS News and the New York Times vary in question format and, for a host of reasons detailed below*, are too complicated for straightforward comparison. But they’re consistent with the overall trend. In 2012, at an intensity level of 34 percent, intense pro-lifers outnumbered intense pro-choicers by six percentage points. In 1999, at an intensity level of 37 percent, the pro-life advantage was four points. In 2004, at an intensity level of 38 percent, the advantage was zero.
These data are complex and open to many interpretations. Pro/con questions and intensity questions varied from pollster to pollster and sometimes from year to year. But one basic takeaway stands out: When the intensity level stays low, around 20 percent or less, the population of Americans most likely to base their voting decisions on this issue is more pro-life than pro-choice. To cancel out that advantage, pro-choicers have to raise the intensity level. When the percentage of respondents who call abortion critical, very important, or a deal-breaker rises toward 30 percent or more, the balance of power begins to shift.
Think about that when you see pro-lifers winning elections, passing laws, and marching in the cold. They don’t win because they’re a majority. They win because they care enough to fight.
*Details of CBS/NYT surveys: In 1999, CBS/NYT asked, “How important is it to you that the presidential candidate you vote for shares your views on the issue of abortion?” Thirty-seven percent of respondents said it was very important. In the sample as a whole, 34 percent said “abortion should be generally available to those who want it,” while 22 percent said “abortion should not be permitted.” (The poll’s middle option—“abortion should be available but under stricter limits than it is now”—is too ambiguous to assign to either side.) In the high-intensity subsample—those who said it was “very important” that a candidate share their abortion views—the percentage who said abortion shouldn’t be permitted rose 12 points. But the percentage who said abortion should be generally available fell only 4 points. The resulting pro-life advantage—34 to 30 percent—was tighter than in any of the Gallup polls.
In 2004, CBS/NYT asked, “Is it possible that you would ever vote for a candidate who does not share your views on the issue of abortion?” Thirty-eight percent of respondents said no. In the overall sample, 34 percent said abortion should be generally available, while 21 percent said abortion shouldn’t be permitted. In the high-intensity subsample, the percentage who said abortion shouldn’t be permitted rose 13 points. But the percentage who said abortion should be generally available didn’t decline at all. The result was a dead heat among high-intensity respondents.
In 2012, a CBS News poll combined the 2004 intensity question with a different pro/con question: whether abortion should be “permitted in all cases,” “permitted, but subject to greater restrictions than it is now,” “permitted only in cases such as rape, incest or to save the woman's life,” permitted only “to save the woman's life,” or not “permitted at all.” Thirty-four percent of respondents said they couldn’t vote for a candidate who didn’t share their abortion views. In the overall sample, 35 percent said abortion should be permitted in all cases, while 48 percent said it should be prohibited with few or no exceptions. (Again, the “permitted but with greater restrictions” option is too ambiguous to assign to either side.) In the high-intensity subsample, the percentage who said abortion should be prohibited increased by only 4 points. Furthermore, this very slight bump was matched by a 3-point increase in the percentage who said abortion should be permitted in all cases. This is the only time I’ve seen no measurable net decline in pro-choice sentiment, relative to pro-life sentiment, when the pool narrows to high-intensity respondents.
Caveats: Pro/con questions, and sometimes intensity questions, differed among the Pew and CBS/NYT polls, and among polling organizations generally. In the CBS/NYT polls, shifts in support for the “permitted with greater restrictions” option may signify factors I’ve missed. The five restriction scenarios offered in the 2012 CBS/NYT poll, with the rape exception as the middle option, may have skewed this poll to the pro-life side.
This analysis of previously unpublished cross-tabulations would not have been possible without the generous assistance of several polling organizations. Thank you to Jennifer De Pinto of the CBS News Polling Unit, Jocelyn Kiley of the Pew Research Center, Alena Naff of Gallup, April Radocchio and Ralph Hansen of the Quinnipiac University Polling Institute, Jon Cohen of the Washington Post, and Slate intern Zahra Ahmed.