In 1987, President Reagan gave a speech to a pro-life group, saying that while a personhood amendment was not in the cards for the foreseeable future, concrete steps were being taken by his administration to show that abortion harms women. Reagan had directed his Surgeon General, the conservative pro-lifer C. Everett Koop to find this evidence. To Reagan’s surprise, after Koop interviewed dozens of women and reviewed 250 studies, the Surgeon General found nothing. However, he did consider that the studies that had been done so far had methodological flaws and called for better research. This call would go unheeded for twenty years. Until 2007, when The Turnaway Study by Dr. Diana Foster began.
Abortion is somewhat common in America; one in four to one in three women will have one in their lifetimes. However, being common does not make it easy to study. In particular, one problem is avoiding comparisons between wanted pregnancies and unwanted pregnancies. For example, poverty might make a pregnant woman desire an abortion, but that same poverty will lead to negative life outcomes for the woman and the baby. The unique aspect of The Turnaway Study is that it compares women with unwanted pregnancies who received an abortion to women who wanted an abortion but were denied it.
The study started recruiting women between 2008 and 2010. This is before Dobbs vs. Jackson’s Women’s Health and abortion was mostly banned in twenty-one states. However, clinics have limits that they will perform abortions up to, and women may not be eligible for an abortion. Since a few days can make the difference between getting or not getting an abortion, there are few differences in the cohorts beyond whether or not they get the abortion. They recruited 1,000 pregnant women from 30 clinics in 21 states. For each woman denied an abortion, they recruited two women who received an abortion in the second trimester and one woman who received an abortion in the first. Women who were seeking an abortion due to fetal anomalies or health risks were excluded from the study.
For five years, they followed the women, following up every six months with a battery of questions about their lives. The study also did in-depth interviews with 31 women, and 10 women’s stories are featured in the book. Each chapter begins with one of these women’s stories to provide examples of women who sought abortions, their reasoning, and their experience. The cohort of women closely resembles the profile of women who receive abortions nationally. Just over a third (37%) were white and not Latina, just under a third (29%) were African American, one in five (21%) were Latina, 4% were American Indian, and 3% were Asian American. Similar to abortion patients nationwide, more than half of the women (60%) were in their twenties at the time of the abortion. Almost one in five (18%) were teenagers, and just over one in five (22%) were 30 or older. Half were living in poverty, although women seeking abortion later in pregnancy were more likely to be below the federal poverty line — 40% of those in the first trimester and 57% of those who sought abortions just above or below the clinic gestational limit. 60% had children, and 45% had a prior abortion. 61% were in a current romantic relationship with the man with whom they became pregnant; 39% reported that the man was a friend, ex-partner, or acquaintance, or that they had no relationship with him. One in five women reported a history of sexual assault or rape; 11 women (1%) were pregnant as a result of rape.
The study asked women open-ended questions as to why they sought an abortion and classified it into various categories. This table shows the primary reason women gave for wanting to terminate a pregnancy:
|Not financially prepared||40%|
|Not the right time for a baby||36%|
|Needs to focus on other children||29%|
|Interferes with future opportunities||20%|
|Not emotionally or mentally prepared||19%|
|Fear that alcohol, tobacco or drug use had hurt their fetus||5%|
That being said, in two-thirds of cases, there was more than one reason. For example, while 40% gave “financial reasons” as a reason, only 6% cited this as the only reason. A small number of women (1%) reported coercion from others as a reason for getting an abortion. Either this is because clinics screen for coerced abortions or coercion itself is rare.
Second-trimester abortions are more stigmatized, both for the woman getting one and for the doctors who perform them. They are much rarer than first-trimester abortions (10% of all abortions). The main reason women wait until the second trimester is that they didn’t know they were pregnant. The average second-trimester abortion patient found out they were pregnant at 12.5 weeks.
80% of women reported something slowed them down in getting an abortion (Not realizing they were pregnant, financing the abortion, making the decision, etc;) 37% said deciding slowed them down. However, this was a matter of a few days, not weeks. Time to decide did not vary between first-trimester and second-trimester abortions.
What are the cause of unplanned pregnancies? While most couples use contraception, the most common reason for an unplanned pregnancy is gaps in contraceptive use. But even with perfect contraceptive use, there will still be unplanned pregnancies. This table shows how many unplanned pregnancies can be expected using a single method:
|Method||Expected Number of Unplanned Pregnancies per Woman Over Lifetime|
|Condoms||.5 - 3.8|
|Oral Contraceptives||.1 - 2.0|
|Contraceptive Patch||.2 - 2.5|
|Contraceptive Ring||.1 - 2.0|
|Injectable Contraceptives||.1 - .8|
|Withdrawal||1.0 - 6.8|
It’s also true that contraception is an endeavor requiring money, forethought, and tolerance of side effects. Many women report having trouble accessing contraceptives. A separate study the author did showed 40% of women seeking abortions had trouble accessing reliable contraceptives. For example, 20% of women experienced running out of pills before they could refill them. Another study the author did showed that in California, the abortion rate could be lowered by supplying birth control pills for one year at a time rather than requiring monthly refills. Another one of the women featured in the book asked for a tubal ligation after her second through fifth unplanned child, but the hospital kept “losing the paperwork”. They were a Catholic hospital and had religious objections to it, but didn’t tell her that.
The author thinks that more research needs to be done into getting better methods of contraception, ideally one that’s easy to use, has no side effects, is effective, and can be easily turned on or off. African-American women had slightly different preferences, specifically wanting menstruation to continue normally while using the method.
Justice Anthony Kennedy wrote in Gonzales vs. Carhart, that “It seems unexceptionable to conclude some women come to regret their choice to abort the infant life they once created and sustained. Severe depression and loss of esteem can follow.“.
Women seeking abortion had mental health statistics that broadly mirrored national estimates. 25% of women had been diagnosed with anxiety or depression, 14% had experienced violence or the threat of violence from an intimate partner in the past year, 20% had a history of sexual assault or rape, 14% reported illicit drug use, and 30% reported problem alcohol use.
Receiving an abortion did not have any effect on mental health or suicidal ideation. Being denied an abortion was associated with short-term stress, depression, and lower life satisfaction, but long-term, both groups’ mental health and life satisfaction improved over time, especially after the pregnancy was over. The timing of the abortion didn’t matter either; women who had first-trimester and second-trimester abortions had the same mental health outcomes.
Abortion had no effect on suicidal ideation. The study found that the traditional predictors of suicidal ideation - depression/anxiety and alcohol use - predicted suicidal ideation in the women in the study.
39% of women reported Post Traumatic Stress (PTS) symptoms, and 16% were at risk of PTSD. Of the 39% with PTS symptoms, 19% reported the pregnancy was the cause of the symptoms. Women who received or were denied an abortion had the same incidence of PTS symptoms. This indicates the abortion did not cause the PTS symptoms.
Before finding out they were pregnant, all cohorts had the same reported drug and alcohol use. After being denied an abortion, women reported drinking and smoking less, while those who received an abortion drank and smoked the same amount. However, women who had an alcohol use disorder did not report drinking less during their pregnancy.
At every interview during the five years of the study, 95% of women said that abortion was the right decision for them.
Initially, women who were denied an abortion still wanted one after one week (65%). After the birth, only 12% of women still wished to have the abortion, declining to 7% at the child’s first birthday and 4% after five years. Women with the least social support (the author regrets not studying financial support) were the most likely to still want an abortion. Women who placed the child for adoption were more likely to wish they could have had the abortion.
Pregnancy involves major bodily changes and health risks. What are the risks of having an abortion versus giving birth, as well as the long-term health effects?
Abortion has a 2% risk of minor complications, compared to childbirth (29%.) Major complications for abortion are extremely rare at .25%, while birth has a 27.5% chance. One woman in 160,000 dies from abortion, while one woman in 11,300 dies from childbirth. Two women in the Turnaway study died from giving birth. This was considered an aberration by the authors and is not statistically significant.
6.3% of women in the study who gave birth study reported life-threatening complications. 1.1% of women who received a second-trimester abortion reported life-threatening complications and .5% of women receiving a first-trimester abortion did.
Most health outcomes were similar over the long term for women who gave birth versus had an abortion. Women who delivered were more likely to suffer migraines (23% vs. 18%). However, on self-reported measures of health rather than specific diseases, women who gave birth were more likely to rate their health poorly than women who received an abortion (27% vs. 21%).
Abortion did not cause infertility. In fact, having an abortion was a risk factor for having another pregnancy. Over the next five years, 39% of women in the study had at least one additional pregnancy, with the average woman in the study having 1.5. Women who received an abortion were more likely to have future pregnancies than women denied the abortion. 29% of unintended subsequent pregnancies are aborted.
Having a baby is a life changing experience with new rewards and responsibilities. How did women’s lives turn out after seeking an abortion?
Women were asked about their plans in the next year and next five years. These plans were classified as aspirational (“get a better job”), neutral (“kids will be older”), or negative (“have less money”). Most of the plans were aspirational (80%). Women who were denied an abortion were much less likely to have an aspirational plan than women who received an abortion (86% vs 56%).
Among the plans that were both aspirational and measurable, women achieved almost half (47%) of their one-year goals. Women were most likely to achieve child-related plans (89%) and financial plans (73%), and least likely to achieve educational plans (31%) and relationship plans (18%). We did not find any difference in achievement in aspirational plans among those who set one by whether they received or were denied an abortion.
With five-year plans, the difference between women who received or were denied an abortion were much smaller. 83% of women denied an abortion had aspirational plans and 91% of women who received one had aspirational plans. 56% of women achieved their five year goals.
Women who were denied the abortion were more likely to be seeking a high school diploma than women who received one (40% vs. 24%). This is likely because the women denied were 1.5 years younger on average. Over five years, 34% graduated and 38% dropped out. This was the same regardless of whether or not the woman had an abortion. Among graduates, participants who were denied a wanted abortion were less likely to complete a postsecondary degree (27%) than those who received an abortion (72%), but they were also less likely to be pursuing such a degree to begin with.
About half of women looking for an abortion are below the federal poverty level, and three-quarters reported they didn’t have enough money to pay for basic expenses. Women who sought earlier abortions were less likely to be poor. In the initial analysis, women denied an abortion were more likely to be unemployed than women who received one. (60% vs. 45%). In the short term, women denied the abortion are more likely to live with family members, but over the long term, they are more likely to raise their children alone than women who received the abortion.
One in twelve women who received an abortion received welfare six months later, while one in six who were denied received welfare. Half of women denied participated in WIC compared to 8% of women who received one, and 44% of women denied an abortion used food stamps compared to 33% who received an abortion. 61% of those denied an abortion were living below the poverty level, compared to 45% who received an abortion.
Another study using Turnaway data looked at the credit reports of women who received or were denied an abortion. Before seeking the abortion, the women were very similar credit-reporting-wise. They found that being denied abortion services increased past-due bills by $1,750 on average, a 78% increase relative to the amount that was past due on their credit reports prior to the pregnancy. Over the five years after seeking abortion, past-due bills remained essentially flat among women who received the abortion. The incidence of very bad financial events recorded in public records, like evictions, bankruptcies, and court judgments for bill nonpayment, also increased significantly, by about 81%, for women who were turned away.
One year after seeking an abortion, 86% of women who got the abortion and 81% of women denied an abortion used contraception. Four years later, 17% of women denied an abortion got a tubal ligation, versus 6% of women who received an abortion. Women who sought an abortion were twice as likely to use Long Acting Reversible Contraception such as an IUD or implant as the general population. Women who got an abortion were more likely to use a barrier or hormonal method than those denied (43% vs. 28%).
As you’ll recall, 60% of women seeking an abortion already had kids. So far we’ve examined the effects of abortion on women, but what effect does abortion have on the child who is born, future children and children the woman already has?
Children of the mother denied the abortion were more likely to live in poverty (72% vs. 55%), more likely to live in a household that receives public assistance (19% vs. 10%), and to live with adults that don’t have enough money to pay for basic necessities (87% vs. 70%). Existing children of a mother denied an abortion were also slightly less likely to hit major developmental milestones on time, as defined by Parents’ Evaluation of Developmental Status: Developmental Milestones (PEDS: DM). 77% of children hit PEDS:DM milestones on time when the mother received an abortion, vs 73% where the abortion was denied.
In the study, they compared the outcomes of children of the women who were denied an abortion to children of women who received an abortion and then had children later on. The author argues this is the ideal comparison group because it allows you to compare if having children at a different time improved the outcome of the children while keeping all other possible confounding factors the same. The unwanted pregnancy will be referred to as the “index child” and future children as “subsequent children”. Birth outcomes were the same for both groups: 10% were born premature, 8% had low birth weight and 13% had to spend time in the NICU. 17% of index children were born within 21 months of a previous child vs 3% for subsequent children. (The WHO recommends 24 months between births)
Subsequent pregnancies are much more likely to be planned (24%) than the index pregnancy. The mother is likely to be 3 years older (27 vs 24). Subsequent children were on average living 32% above federal poverty level, but both households received public assistance at the same rate. Mothers of subsequent children were less likely to report having not enough money for basic necessities (55% vs 72%).
Mothers of subsequent children were more likely to be living with a male romantic partner (49% vs 35%). It was not asked if that male partner was the biological father of the children, but the author thinks they most likely are.
After being assessed for psychological bonding with their baby, 9% of mothers denied an abortion had poor bonding with their baby, compared to 3% for subsequent children.
9% of women put up their child for adoption. Those who did were the most likely to have feelings of regret or report negative emotions.
Some men are involved in the decision making to get an abortion, while some men are the reason women opt for an abortion. Unreliable or abusive partners do not make great fathers. Women also realize that having a kid with a man may tie them to the father for the rest of their lives, and not want to make that commitment.
The study did not ask any of the men involved about their feelings or thoughts about the pregnancy, but they did ask their partners about the men involved and their relationships/impressions of them. The author thinks studying the father’s perspective is a thing worth doing, but outside the scope of the Turnaway study.
31% of women reported that the man involved was the reason for abortion. Of those, 33% said that their relationship wasn’t good enough to be sustained with a child, 25% said that the partner would not or could not support a child, and 20% said that they were the wrong man to have a baby with, for example, being a career criminal or incarcerated.
Of the 31 women who were sat down for qualitative interviews, two said their then-boyfriends pressured them to have an abortion. But only one woman in the study (not interviewed) reported pressure as the sole reason for getting an abortion
2.5% of women cited partner abuse as the top reason for seeking an abortion, similar to other surveys of women seeking abortions. 1 in 20 women reported violence from the father, and 1 in 30 reported fear of violence. Women who had the abortion had drastically lower rates of violence after a year than women denied the abortion.
80% of women were in a relationship with the father at the time of conception. By the time the women sought an abortion, it had declined to 61%, by two years 37%, and four years 26%. Receiving or being denied the abortion did not affect whether or not the relationship dissolved.
Initially, they didn’t ask any questions about future relationships but added the questions at the two year mark. There was no difference between women who received or denied an abortion and whether or not they had a future relationship; however, women who received an abortion were twice as likely to rate their relationship as good (47% vs. 28%).
Note: This was written in 2020. Laws have changed since then.
Let’s contrast Louisiana and California. The average distance to an abortion clinic in Louisiana is 60 miles; in California it is 7. Louisiana requires two visits, one to hear the state education program to discourage abortion and a waiting period, then another visit to get the abortion. California does not have a waiting period. Medi-Cal covers abortions; Louisiana Medicaid doesn’t. The woman featured in this chapter takes an extra month to raise $500 to get an abortion. This is common: women who can’t raise the money take longer to get the abortion, making it even more expensive and potentially pushing it into the second trimester.
As of 2019, 11 states prohibit private insurance from covering abortions, and twenty-six states restrict it on public exchange plans.
About 50% of women seeking abortions are living below the federal poverty level, so these laws are quite effective at restricting abortion.
Women who got second-trimester abortions were more likely to report delays due to trouble funding the abortion (41% vs. 20%).
Because of the Hyde Amendment, the federal government cannot use funds for abortion except in cases of rape, incest, or severe health problems in the mother. Fifteen states pay for abortion out of their own medicaid funds, but some women reported having a hard time signing up for Medicaid. Of the 11 women in the study who reported getting pregnant due to rape, only 2 of them got money from Medicaid. Other problems include that Medicaid reimburses doctors less than the cost of the procedure, so many are reluctant to accept it.
25% of women reported getting help from an abortion fund, and funds such as the National Network of Abortion Funds provide support for 1 in 7 abortions nationally. The author had an intern call these hotlines, and it took about 6 hours to speak to someone.
23% of women seeking a first-trimester abortion traveled 100 miles, versus 30% for second-trimester abortions. First-trimester patients spent an average of $23 on travel costs, while women seeking later abortions spent an average of $100 on transportation.
Not all abortion clinics are the same; clinics have a maximum gestational limit they will perform abortions up to. All offer it up to 8 weeks, 72% offer it up to 12, 25% 20 weeks and 10% 24 weeks. Women who need a later abortion have to call around to find the right facility or get referred from clinic to clinic, with each step taking a little more time. 58% of women seeking a later abortion visited more than 1 facility, and 12% visited 3 or more.
Many states have laws requiring the woman to get an ultrasound before receiving the abortion, and sometimes verbally describe it. They don’t have to view it, however. 79% of women went to a clinic that offered it voluntarily or had no policy; 21% went to a clinic where state law required it. 48% of women were offered the opportunity to view an ultrasound, and 65% of those women chose to view it. 68% of women who had never had a baby before chose to view it. Women in states where the ultrasound was mandatory were much less likely to want to view it (48% vs. 82%).
Of those women who did get an ultrasound and viewed it, they asked the women open-ended questions about how they felt about it, and classified it into various emotions. The most common emotion was neutral (33%), the next most common were negative emotions such as guilt or sadness, and some were even positive about seeing the ultrasound. Nobody changed their mind, and there was no relationship between fetal development and emotional response.
The author thinks that it’s a way of coming to terms with the pregnancy, but doesn’t provide any citations or stats to support it.
Some states require that doctors give certain information about abortion before having the procedure, including false statements such as abortion causing breast cancer. Sometimes, getting this information can be a burden. For example, in Pennsylvania, you have to call a state hotline and keep dialing until they pick up, which was reported as very random and erratic.
Some clinics and states require some sort of counseling. Two-thirds of women reported receiving some kind of counseling. 99% of counselors said they would support either decision; 3% encouraged abortion, and less than 1% discouraged it. 40% reported that the counseling they received was extremely helpful, 28% quite, 17% moderately, 10% a little, and only 4% reported it was not at all helpful. Women who received counseling at a clinic where a standardized counseling script was mandated by the state were significantly less likely to report finding counseling extremely or quite helpful (60%) compared to 75% of women who went to a clinic without state-mandated counseling.
The Turnaway Study only recruited women who made it into the abortion clinic, so it can’t measure if protesting turns them away. Slightly less than half of women reported seeing protestors. A third saw them, a third were spoken to by the protestors, and a third experienced an attempt to stop them from entering. Half didn’t find the protestors upsetting, a quarter found them a little bit upsetting, and 16% found them very upsetting.
The more contact they had, the more likely they were to be upset by the protestors. Two-thirds of the women who were stopped by a protestor found them upsetting, compared to 36% who only saw them.
3% of the women in the study thought that abortion should be banned in any situation, compared to the 15-20% of adults in America who think that. 20% believed abortion is morally wrong, compared to 47% of Americans. Women denied an abortion were more likely to say they had become less supportive of abortion rights, while women who received one were more likely to say they were more supportive of abortion rights. Latina and African-American women were more likely to think abortion was wrong or oppose legal abortion than White women.
Like any good scientist, the author includes criticism of her study. I also found this article but it’s been retracted.
Pro-life critics argue that UCSF and the author are biased because UCSF is a major medical school that trains doctors in performing abortions. The author argues that yes, UCSF has a bias towards abortion, but that many different scientists across different institutions analyzed this data and created the results, not just her and UCSF. She says she didn’t intend to make a political statement with the Turnaway study, though she is pro-choice. I don’t find it terribly convincing that she approached this from a politically neutral perspective, based on the way the book is written.
One concern is that the Turnaway study might not be representative of women seeking an abortion. 3,045 women were approached, and only 1,132 of those women agreed to hear more about the study. Of those, 956 participated. But 95% of the women tracked completed the course of interviews. It is, of course, a big ask to participate in 30- to 60-minute interviews every six months for five years and disclose a lot of contact information.
Participation varied a lot by recruitment site. The top three sites enrolled 2/3ds of the women involved, and the bottom 5 enrolled less than quarter of the participants. This was attributed to the skills and charisma of the recruiters. It was not the case that more conservative areas were underrepresented. For example, one of the most successful recruiters was in Fargo, ND. The recruiter there said that most survey recruited women from the coastal states and this was a chance for their voices to be heard, encouraging more women to enroll.
The author says that even if you restrict the data set to sites with a 50% participation rate, the results are the same, though they lose their statistical significance. She also points out that findings from the Turnaway study are pretty similar to prior studies on abortion in America.
Regardless of your position on abortion, we cannot make good policy without information about the consequences of the policies we choose. I don’t expect the facts from The Turnaway Study to change any minds - it is a debate over values, not science. But without knowing how our choices play out in real life, a debate about values is empty.