We polled 1,060 Americans about abortion. This is what they got wrong.

By: Sarah Kliff

Most Americans — Democrats and Republicans, men and women, pro-choice and pro-life — all share a belief about abortion: that it's rare.

Twenty-seven percent of Americans think fewer than 10 percent of women will have an abortion in their lifetime; 51 percent say it’s fewer than 20 percent, a new Vox poll of 1,060 registered voters conducted by communications firm Perry/Undem shows.

This turns out to be a significant underestimate. The best data available suggests that about 25 to 30 percent1 of American women will terminate a pregnancy at some point in their lives.

The most recent estimate available suggests that 30 percent of American women will have an abortion in their lifetime. This figure comes from the Guttmacher Institute, a reproductive health-focused nonprofit that supports abortion rights. In a 2011 paper in the journal Obstetrics and Gynecology, Rachel Jones and Megan Cavanaugh estimated that if the 2008 abortion rate held steady, 30 percent of women alive then would have an abortion during their lives. But since 2008, the abortion rate has dropped — and some have questioned whether the 30 percent figure is still a valid estimate. Those are fair questions and are the reason we identify the prevalence of abortion in the story as somewhere between 25 and 30 percent.

Abortion prevalence chart

Americans also overestimate the safety risks for women who have abortions, our poll shows. Most people think abortion is either "less safe" or "about as safe" for women as giving birth. But that's not true. In actuality, bearing a child causes more serious complications and deaths for mothers than abortion does.

Abortion is a common medical procedure. There are hundreds of thousands more abortions each year in the US than either appendectomies or hysterectomies.

Abortion is also safe for women who have the procedure. There are twice as many complications associated with wisdom teeth removal as with abortion. Complications for both procedures tend to be relatively minor (for abortion, a common complication is cramping, and for wisdom teeth it's inflammation).

These misperceptions aren't just unfortunate psychological quirks; they work together to contribute to a view of abortion as being infrequent and risky for the women who have one. That ultimately shapes the way we regulate abortion in the United States and how we judge which restrictions ought to stand.

Our misperception of abortion as rare means that we think few women are affected by new abortion laws and regulations, when many are.

And our misperception of abortion as dangerous for women who have them allows more laws that reduce abortion access — like Texas's new restrictions — under the guise of improving the safety of an already safe medical procedure.

Americans tend to think that abortion is rare — especially well-educated, high-income Americans

Social science shows that we tend to underestimate the frequency of experiences we hear less about. And that makes sense; if all your neighbors are talking about crime, or you heard about it a lot on the nightly news, it would be fair to start wondering whether a local crime wave has occurred. We overestimate things like teen pregnancy, perhaps because there are entire reality shows devoted to the subject.

"We overestimate the amount of people who get killed by gunshots and underestimate the amount killed by heart attacks, because we see a lot more news about gunshots," says Eric Beasley, a sociologist who studies risk perception at Michigan State University. "Abortion is hidden and not talked about, so nobody knows how often it happens."

Our poll finds groups of Americans that have the highest abortion rates — low-income and less educated women — tend to more accurately guess the prevalence of abortion. Demographics with lower abortion rates, meanwhile, tend to have less accurate guesses.

Demographics abortion chart

More educated and higher-income Americans are especially likely to believe that abortion is rare.

For example, 54 percent of Americans without a college degree underestimate abortion rates, compared with 70 percent of those with graduate degrees. And 51 percent of those earning less than $50,000 underestimate the frequency of abortion, compared with 69 percent of those earning more than $175,000.

The split happens when you look at gender, too. Women would near certainly have more experience with abortion than men. Our poll shows that 67 percent of men underestimate the frequency of abortion, compared with 57 percent of women.

Also, knowing someone who had an abortion increases the odds that that person will accurately approximate the overall abortion rate. Those who say they have discussed the experience of another's abortion are 9 percentage points less likely to underestimate the abortion rate (55 percent versus 64 percent).

Americans overestimate how dangerous abortion is for women undergoing the procedure

Studies repeatedly show that abortion rarely results in medical complications and almost never causes the mother to die. Of the 730,322 women who had abortions in the United States 2011, two died. That's a rate of 0.27 deaths for every 100,000 women having an abortion. The rate for childbirth in America is 17.8 deaths for every 100,000 women giving birth.

But our perception of abortion's safety is different — and influenced by our views on abortion itself.

Abortion safety chart

Overall, four out of five poll respondents (80 percent) said that childbirth was safer or "about as safe" as abortion.

This was especially true among respondents who identify as pro-life; 88 percent of that group estimated abortion to be equally safe or less safe than childbirth. That could reflect moral beliefs about abortion. If you believe abortion is wrong, that might lead you to overestimate how dangerous it is for the women who undergo it.

More interesting, though, is that abortion rights supporters — people in our poll who identify as pro-choice — also perceive abortion as dangerous. They presumably don't have the same moral objections to terminating a pregnancy. But 67 percent of pro-choice respondents incorrectly estimated abortion to be riskier or just as risky as giving birth.

Abortion tends to happen separately from the rest of the health care system, housed in its own clinics, because doctors and hospitals often don't want to deal with the political backlash and specific regulations that can come with providing that service.

A third of our poll respondents said they weren't sure whether doctors who performed abortions had to maintain medical licenses like other doctors. (The answer is yes, abortion providers are indeed required to maintain medical licenses.)

These misperceptions about dangers and training help set the stage for the most recent wave of abortion restrictions that impose additional health regulations on clinics. Most notable of these is Texas's law, which the Supreme Court will hear oral arguments on this week, and which requires all abortion clinics to become ambulatory surgical centers.

Where false beliefs on abortion come from

It's not totally clear why our perceptions of abortion skew toward high levels of danger and risk. It could relate to the fact that abortion rarely gets discussed, so it seems shadowy and unknown. Or an earlier era of unsafe and unregulated abortion may stick in our minds, from the 1950s and '60s and earlier.

"It might be a holdover from talking about illegal abortions, which are unsafe," says NYU sociologist Sarah Cowan, who studies how women talk about abortion.

Television shows and movies might also play a role, where plot lines have portrayed abortion as a dangerous medical procedure. One 2014 analysis found that 9.3 percent of abortion patients on TV and in movies ultimately die from complications of the procedure. (That's 34 times higher than the actual rate.)

"When abortion is shown on television, it's often a life-threatening procedure or a very extreme case," says Julia Reticker-Flynn, who runs the 1 in 3 Campaign, which encourages women to share their abortion stories. "Even when politicians talk about it, there's an overemphasis on the dangers."

Women are also reticent to share their abortion stories. Cowan at NYU has found that more people say they've had someone tell them about a miscarriage than an abortion, even though abortion happens more often. This suggests that abortion gets kept secret, allowing misperceptions to thrive.

Would the abortion debate be different if Americans had more accurate perceptions?

Our poll data does show that people who know someone who had an abortion are also more likely to support abortion rights.

But that doesn't necessarily mean that hearing about an abortion causes pro-choice views. It's possible that people feel more comfortable confiding about their abortions with those who already have liberal views on the topic.

A better understanding of the prevalence of abortion might normalize the experience and reduce the stigma around abortion. Some advocates think that if Americans were more aware of the frequency of abortion — if they knew a neighbor or friend who had terminated a pregnancy — they would become more likely to support abortion rights.

Or better knowledge might actually heighten the abortion debate. That's because supporters of abortion rights and opponents could react very differently to the new information.

"The numbers don't necessarily tell everyone the same story. To some, it might say, 'Look how normal and common this is,'" Cowan says. "But if you oppose abortion, you might see this and think the problem is way bigger than you realized."

By magnifying the size of the perceived problem, it could actually intensify the effort to restrict abortions.

PerryUndem Research/Communication conducted the survey among n = 1,060 registered voters 18 and older nationwide, January 20 through 27, 2016. The survey was administered among a nationally representative sample of voters, using GfK's KnowledgePanel. The margin of error is +/- 3.7 percentage points. Some results do not add to 100 percent as a result of rounding. Topline results are available here.