• Open Access

Who Seeks Abortions at or After 20 Weeks?

Authors

  • By Diana Greene Foster,

    Corresponding author
    1. Diana Greene Foster is associate professor, and Katrina Kimport is assistant professor, both at the Advancing New Standards in Reproductive Health program in the Bixby Center for Global Reproductive Health, Department of Obstetrics, Gynecology and Reproductive Sciences, School of Medicine, University of California, San Francisco
    Search for more papers by this author
  • Katrina Kimport

    1. Diana Greene Foster is associate professor, and Katrina Kimport is assistant professor, both at the Advancing New Standards in Reproductive Health program in the Bixby Center for Global Reproductive Health, Department of Obstetrics, Gynecology and Reproductive Sciences, School of Medicine, University of California, San Francisco
    Search for more papers by this author

Abstract

Context

Recent years have seen the introduction of state bills seeking to ban abortions after 20 weeks, but little empirical data exist on who is affected when such bans become law.

Methods

As part of a larger study, 272 women who received an abortion at or after 20 weeks’ gestation and 169 who received first-trimester abortions at 16 facilities across the country in 2008–2010 were interviewed one week after the procedure. Mixed effect logistic regression analyses were used to determine the characteristics associated with later abortion (i.e., at 20 weeks or later). Causes of delay in obtaining abortion were assessed in open- and closed-ended questions; profiles of women who received later abortions were identified through factor analysis.

Results

Women aged 20–24 were more likely than those aged 25–34 to have a later abortion (odds ratio, 2.7), and women who discovered their pregnancy before eight weeks’ gestation were less likely than others to do so (0.1). Later abortion recipients experienced logistical delays (e.g., difficulty finding a provider and raising funds for the procedure and travel costs), which compounded other delays in receiving care. Most women seeking later abortion fit at least one of five profiles: They were raising children alone, were depressed or using illicit substances, were in conflict with a male partner or experiencing domestic violence, had trouble deciding and then had access problems, or were young and nulliparous.

Conclusion

Bans on abortion after 20 weeks will disproportionately affect young women and women with limited financial resources.

Recent years have seen the introduction of state bills seeking to ban abortions after 20 weeks.[1] Citing contested science suggesting that the fetus can feel pain at 20 weeks,[2, 3] this legislation has drawn attention to a small minority of abortions. The considerable majority of abortions (91%) take place in the first trimester, before the end of the 13th week of pregnancy; 7% are performed at 14–20 weeks, and 1% at 21 weeks or later.[4] Nonetheless, this final category constitutes a substantial number of abortions: Given an estimated 1.21 million abortions in the United States annually,[5] more than 15,000 likely take place after 20 weeks. It is these procedures that have captured legislative attention.

Although scholars have rigorously contested claims of fetal pain,[6, 7] bans on abortions after 20 weeks’ gestation have provoked little outcry. Perhaps this is because acceptance of abortion wanes as gestational age increases, even among supporters of abortion rights;[8] some abortion providers have acknowledged their discomfort with performing the procedure at later gestations.[9-13] Some supporters of abortion rights have called for the prochoice movement to cease advocating for rights to later abortions, emphasizing the emerging personhood of the fetus.[14-16] As a pregnancy progresses, the fetus takes on more of the appearance of a baby, and for some people, this development may tilt the balance between the rights of the woman and those of the fetus.

Concern over the increasing personhood of the fetus, however, is not the only cause of diminished support for later abortion. Lack of support also stems from judgments about the women seeking these terminations. Contemporary discourses identify reproduction as a process that can and should be controlled, and place the responsibility for this control on individual women,[16, 17] presuming that women know early in a pregnancy whether they want to carry it to term and positing early abortion as morally and ethically preferable to later abortion. According to these discourses, women who seek later abortions have not only failed to use contraceptives adequately to avoid an unwanted pregnancy, they have also failed to terminate that pregnancy early enough.

This characterization, however, is not informed by empirical data on women seeking later abortions. The body of research on women who have dealt with fetal anomalies or life endangerment during pregnancy describes their stories as narratives of pregnancy wantedness and tragic circumstances.[18-20] We do not know how accurately these narratives characterize the circumstances of women who seek later abortions for reasons other than fetal anomaly or life endangerment. But data suggest that most women seeking later terminations are not doing so for reasons of fetal anomaly or life endangerment.[5, 21]

Indeed, we know very little about women who seek later abortions. Random samples of abortion clients capture few women at gestations past the middle of the second trimester. For this reason, the most commonly cited research on post–first-trimester abortion focuses primarily on women in the early second trimester.[22-24] The most salient findings are that women seeking second-trimester abortions did not realize they were pregnant until much later than women seeking first-trimester abortions, and that myriad logistical barriers slow down access to abortion once a woman is beyond 13 weeks. Certain physical health conditions, such as obesity[25, 26] and a lack of pregnancy symptoms,[26] increase the risk of late discovery. Research from the United Kingdom has identified uncertainty about what to do if pregnant and changing personal circumstances, such as dissolution of romantic partnerships or job loss, as associated with delay in seeking abortion.[27] Jones and Finer[24] offer some empirical data on later abortion, but because of the constraints of random sampling, they report only on women who have abortions at 16 weeks or later.

Without adequate understanding of who has later abortions and under what circumstances, the effect of legislative bans on procedures at or after 20 weeks cannot be known. This renders the question of who will be affected by these bans of increasing and timely importance. The current study addresses this question by analyzing data on women who sought and received an abortion at or after 20 weeks’ gestation for reasons other than fetal anomaly or life endangerment.

Methods

Study Sample and Procedures

To offer a fuller portrait of women who seek abortions at 20 or more weeks’ gestation, we draw on data collected as part of the Turnaway Study, a project that investigates the consequences of receiving or being denied a wanted abortion. Details about study design, recruitment sites and recruitment can be found elsewhere.[28, 29] The Turnaway Study was approved by the University of California, San Francisco, Committee for Human Research.

Recruitment sites were identified using the National Abortion Federation membership directory and professional contacts; recruitment took place at 30 abortion facilities, all of which had the latest gestational limit of any facility within 150 miles. The facilities set their limits on the basis of state law, the availability of trained physicians, clinician and staff comfort, and facility regulations. Gestational age limits at these clinics ranged from 10 weeks to the end of the second trimester; 16 sites had a limit beyond 20 weeks. As is true generally of abortion care in the United States, women could self-refer; facilities’ waiting periods and counseling differed because of variation in state law and facility policy.

Between January 2008 and December 2010, women who presented for an abortion near or just past the gestational limit were recruited.1 Women were told that the purpose of the study was to learn more about how women's decisions and experiences with unintended pregnancy and abortion affect their lives—for example, in the areas of health, employment, relationships and sense of well-being. After an initial interview, the women were followed for five years. We limit the analyses here to participants who received abortions at the 16 sites with a gestational limit beyond 20 weeks. These facilities were distributed across the United States: three in the West, three in the Northeast, four in the Midwest and six in the South. The data come from participants’ first interview, one week after their abortion.

Measures

We examined a broad range of social and demographic characteristics: age; race or ethnicity; marital and employment status; income; previous births and abortions; whether the woman has a child younger than a year; and whether she is raising a child on her own. Participants were also asked if they were uninsured, privately insured, publicly insured prior to this pregnancy or publicly insured only for this pregnancy (a category possible in states where Medicaid covers abortion services).

To assess general health characteristics and behaviors, we asked participants whether they had experienced any of four chronic health conditions (asthma, diabetes, high cholesterol and hypertension); several types of chronic pain; alcohol or substance use; and recent physical abuse. Any episode of consuming more than five drinks in the past month was considered “heavy drinking.” Participants were also asked about their weight (those with a body mass index of 30 or more were considered obese) and mental health history; history of major depression was established by the 12-month Composite International Diagnostic Interview[30] or a reported past diagnosis of depression.

A number of measures explored the pregnancy and abortion-seeking experience: gestation at discovery of the pregnancy (dichotomized to before and after eight weeks), travel time to the clinic, number of facilities contacted and visited, and costs.

Women were asked whether eight specific factors slowed them down in obtaining abortion care: not knowing they were pregnant, not knowing where to go for an abortion, difficulty getting to the abortion facility, raising money for travel costs, the cost of the procedure, difficulty securing insurance coverage, trouble deciding whether they wanted an abortion, and disagreeing with the man involved in the pregnancy on the decision to have an abortion.

We also asked women several open-ended questions: “What are some of the reasons you decided to have an abortion?” “What would you say was the main reason you decided to have an abortion?” “Did anything slow you down and prevent you from coming to the [recruitment site] earlier in your pregnancy?” and “What would you say slowed you down the most?” The general phrase “slow you down” allowed participants broad leeway to report on their experience.

Analysis

Our analysis had several parts. We began by describing the women who received later abortions and compared their characteristics to those of women who had first-trimester procedures. (The numbers of women at gestational ages in the early second trimester were too small to permit us to treat gestational age as a continuous measure.) Bivariate and multivariate models were assessed using mixed effects logistic regression with random intercepts for site. Models included five social and demographic characteristics, three variables about the pregnancy and abortion-seeking experience, and four health characteristics. Because the variable describing the gestational age at which the woman discovered the pregnancy was closely associated with gestational age at abortion, we estimated our model both with and without it to check the sensitivity of the results to its inclusion.

Second, the participants’ responses to the questions about obstacles to obtaining care were used to examine reasons for delay in seeking abortion. Participants were considered to have cited a reason for delay if they identified that reason as slowing them down in response to either the closed-ended prompts or the open-ended questions. Broadly, these reasons applied both to women who obtained later abortions and to those who obtained first-trimester abortions. To discern differences between these two groups, we used t tests to compare mean costs, gestational age, number of facilities contacted and visited, and travel time.

Third, we conducted a factor analysis to identify profiles of women who received abortions at or after 20 weeks on the basis of responses to specific items that tend to be reported together. This analysis included seven social and demographic items,2 six causes of delay in seeking an abortion,3 three physical health items,4 and three mental health and substance use items.5

Results

Of the eligible women approached at the 16 facilities with a gestational limit beyond 20 weeks, 44% consented to participate in the five-year longitudinal study. The analyses below include the 272 of these women who received a later abortion and the 169 who received a first-trimester abortion. The average gestation at time of abortion was 22 weeks among the later abortion group and eight weeks among the first-trimester group. Some 38% of the women in our sample were black; 31% were white, 19% were Latina and 12% were of mixed or some other race. Seventeen percent were aged 15–19, and 63% were already mothers. Our sample is similar in race and ethnicity, age and parity to the population of women who receive abortions nationally.[31]

Women Who Had Later Abortions

In many ways, women who had later abortions were similar to those who obtained first-trimester procedures. The two groups did not differ by race or ethnicity, number of live births or abortions, mental or physical health history, or substance use (Table 1). Nonetheless, several significant differences were evident in their social and demographic characteristics and, particularly, in their pregnancy and abortion-seeking experience. In bivariate analyses, women seeking later abortion were younger than women seeking first-trimester abortions: Fifty-nine percent of the former were younger than 25, compared with 41% of the latter. Women seeking later abortions were also less likely than those who had first-trimester procedures to be married (4% vs. 9%) and to be employed (50% vs. 66%). Further, a lower proportion of women in the later abortion group than in the first-trimester group had private insurance (23% vs. 33%). Women who sought later abortions were also much more likely than first-trimester patients to have been eight or more weeks pregnant at the time they discovered their pregnancy (68% vs. 12%) and to have to traveled more than three hours to get to the abortion facility (21% vs. 5%).

Table 1. Percentage distribution of women who obtained abortions at 16 facilities with gestational limits beyond 20 weeks, by selected characteristics, according to timing of abortion, United States, 2008–2010
CharacteristicFirst-trimester (N=169)Later (N=272)
  1. a

    p<.05.

  2. b

    Asthma, diabetes, high cholesterol and hypertension. In the last month. Notes: “Later abortion” refers to an abortion taking place at or after 20 weeks. Percentages may not total 100 because of rounding.

Agea  
15–191519
20–242640
25–345035
≥3597
Race/ethnicity  
White3429
Black4136
Latina1721
Other914
Marital statusa  
Never-married8187
Married94
Separated/divorced98
Employmenta  
Not employed3450
Employed6650
Insurance statusa  
No insurance2328
Medicaid for this pregnancy only1218
Medicaid before pregnancy3231
Private3323
No. of live births  
03738
≥16363
No. of previous abortions  
05052
≥15048
Gestation at discovery of pregnancya  
<8 weeks8832
≥8 weeks1268
History of depression  
No7381
Yes2719
Any of four chronic health conditionsb  
No5356
Yes4744
Travel time to the clinica  
≤3 hours9579
>3 hours521
Heavy drinking  
No8388
Yes1712
Use of recreational drugs  
No8490
Yes1610
Total100100

In a multivariable model, women aged 20–24 were significantly more likely than women aged 25–34 to present at or after 20 weeks (odds ratio, 2.7—Table 2). Employed women were less likely to seek a later abortion than women who were unemployed, and compared with women who discovered their pregnancy at eight weeks or later, those who discovered it sooner were less likely to seek a later abortion (0.5 and 0.1, respectively). Traveling more than three hours to get to the abortion facility was associated with increased odds of later abortion (4.6). Results were similar in a model that excluded late discovery of pregnancy (not shown).

Table 2. Odds ratios (and 95% confidence intervals) from multivarate logistic regression analyses examining associations between selected characteristics and the likelihood of seeking later abortion
CharacteristicOddsratio
  1. a

    p<.05.

  2. b

    Asthma, diabetes, high cholesterol and hypertension.

  3. c

    In the last month. Notes: “Later abortion” refers to an abortion taking place at or after 20 weeks. ref=reference group.

Age
15–191.72 (0.78–3.79)
20–242.71 (1.46–5.04)a
25–34 (ref)1.00
≥351.13 (0.39–3.27)
Race/ethnicity 
White (ref)1.00
Black1.13 (0.57–2.22)
Latina1.40 (0.65–3.01)
Other2.20 (0.88–5.47)
Marital status 
Never-married (ref)1.00
Married0.73 (0.24–2.24)
Separated/divorced1.14 (0.44–2.94)
Employment 
Not employed (ref)1.00
Employed0.48 (0.28–0.83)a
Insurance status 
No insurance (ref)1.00
Medicaid for this pregnancy only1.18 (0.51–2.69)
Medicaid before pregnancy0.55 (0.28–1.11)
Private0.77 (0.38–1.54)
No. of live births 
00.99 (0.52–1.90)
≥1 (ref)1.00
No. of previous abortions 
01.00 (0.58–1.71)
≥1 (ref)1.00
Gestation at discovery of pregnancy
<8 weeks0.07 (0.04–0.11)a
≥8 weeks (ref)1.00
History of depression 
No (ref)1.00
Yes0.79 (0.42–1.48)
Any of four chronic health conditionsb 
No (ref)1.00
Yes0.85 (0.51–1.43)
Travel time to the clinic 
≤3 hours (ref)1.00
>3 hours4.61 (1.89–11.26)a
Heavy drinkingc 
No (ref)1.00
Yes1.06 (0.50–2.25)
Use of recreational drugsc 
No (ref)1.00
Yes0.62 (0.28–1.36)
Constant8.97 (3.28–24.52)a

Reasons for Delay

Women in both abortion groups experienced delays. Indeed, 94% of later abortion patients and 80% of first-trimester patients reported that something slowed them down (Table 3, page 214).

Table 3. Percentage of women reporting specific reasons for delay in seeking abortion services, by timing of abortion
ReasonAllb (N=384)First-trimester (N=166)Later (N=218)
  1. a

    p<.05 in both chi-square tests and mixed-effects models that account for clustering by site.

  2. b

    Includes only women who replied to open-ended questions or closed-ended prompts. Note: “Later abortion” refers to an abortion taking place at or after 20 weeks.

Any barriera888094
Not knowing about the pregnancy434045
Trouble deciding about the abortion373340
Disagreeing about the abortion with the man involved181620
Not knowing where to go for an abortiona301838
Difficulty getting to the abortion facilitya201227
Raising money for procedure and related costsa513165
Difficulty securing insurance coveragea322041

We describe below the seven most common experiences that accounted for delay in seeking abortion in the order in which they typically occurred. Notably, these barriers related to multiple aspects of pregnancy and abortion care; they were financial, medical, social and geographic, and also reflected a lack of access to information. Although women across the sample reported similar delays, some reasons were more common among later abortion patients than among those who obtained first-trimester procedures.

Not recognizing pregnancy

Overall, 43% of women reported that not realizing they were pregnant delayed them in seeking abortion care; the proportion did not differ between the two groups. However, women seeking later abortions were generally much farther along when they discovered their pregnancy than were women seeking first-trimester abortions: 12 weeks, on average, compared with just five weeks (p<.05—not shown). For example, a 22-year-old white woman from Illinois discovered she was 23 weeks pregnant and had her abortion the same week. She said, “I didn't know I was pregnant in the beginning.”

Trouble deciding about the abortion

Having recognized their pregnancy, 37% of women reported that the process of deciding whether to have an abortion slowed them down. For some, difficulty deciding resulted strictly from indecision about what to do and a desire not to rush into anything. A 30-year-old Latina woman from California, who had her abortion at 23 weeks’ gestation, talked about being slowed down because she was “deciding whether or not to do it. I had to make sure I was okay with myself.” Other women explained that their personal stance against abortion made the decision difficult. A 28-year-old multiethnic woman from California who sought an abortion at 22 weeks’ gestation explained that she was delayed in seeking care because she “didn't believe in abortion prior to this.”

As with delay in discovering pregnancy, the proportion of women who cited this factor was similar for both groups, but the gestation at which women first considered abortion differed. Women seeking later abortions were, on average, already in the second trimester when they first considered abortion, whereas women in the first-trimester group initially considered abortion at six weeks’ gestation (p<.05).

Disagreeing with the man involved

One in five participants said that disagreement with the man involved in the pregnancy over their decision to have an abortion slowed them down. A 19-year-old Latina woman from Illinois, who had an abortion at 22 weeks, explained, “My partner, we broke up. And when we got back together, he didn't want me to have the abortion.” Similarly, conflict with the man involved in the pregnancy could make women fearful about obtaining an abortion. A 26-year-old Latina woman in New Mexico, who had an abortion at 28 weeks’ gestation, said, “I was afraid of my boyfriend finding out, and I went [to the abortion clinic] once he was in jail.” The proportion of women who reported this reason for delay did not differ significantly between the two groups.

Finding and getting to a facility

Once participants decided to have an abortion, logistics often complicated their ability to obtain the procedure. Some women had trouble finding a place to go. This was especially the case for women seeking later abortions, because later procedures are less commonly available;[5, 32] 38% of these women reported delay for this reason, compared with 18% of those seeking first-trimester procedures. Women seeking later abortions called an average of 2.2 abortion facilities before they found the one that was able to perform their procedure, while women seeking a first-trimester abortion contacted an average of 1.7 facilities (not shown). Fifty-eight percent of later abortion patients visited more than one facility, and 12% visited three or more facilities. In contrast, only 33% of women seeking a first-trimester abortion visited more than one clinic (p<.05).

Highlighting the difficulty later abortion patients faced finding a facility, a 23-year-old black woman from New York, who had her abortion at 22 weeks’ gestation, said that she had never before heard of the clinic where she obtained her abortion: “I found it on a Web site after two other clinics I called said they couldn't do it.” Some women who sought later abortions also had physical conditions associated with abortion delay, such as obesity.[25] A 27-year-old black woman from California, who sought an abortion at 21 weeks’ gestation said, “The first clinic that I went to [turned me away because of my weight and] didn't tell me that there was another option.”

On top of challenges in locating a facility, women faced difficulty getting there. A 35-year-old multiethnic woman in California, who had her abortion at 22 weeks’ gestation, described how travel considerations slowed her down: “I didn't trust [my car] on the freeway and had to get it fixed before I could [go to the clinic].” Women seeking later abortions were more than twice as likely as first-trimester patients to report that difficulty getting to the abortion facility slowed them down (27% vs. 12%).

Costs

Many women faced difficulty covering the costs associated with their abortion. The average prices participants in our study reported were $519 for a first-trimester procedure and $2,014 for a later abortion, somewhat higher than the prices reported in other research.[5] A 21-year-old black woman in California, who sought an abortion at 22 weeks’ gestation, said that the “financial part [slowed me down], because I could get there, but paying for it was the biggest problem.” For some women, raising money for the procedure took so long that by the time they had gathered enough money, their pregnancy had progressed to a stage that necessitated a more expensive procedure. A 28-year-old Asian woman from Kentucky, who was at 21 weeks when she received her abortion, explained, “I couldn't afford it. They told me it was going to be $650, [but] by the time I was able to raise the $650, they had to do a different procedure, and so the price went up. The price jumped to $1,850,…and they don't take insurance.”

Financial challenges were not restricted to the cost of the procedure. Longer distances to an abortion facility could be associated with additional transportation costs. Women seeking a first-trimester abortion spent an average of $18 (range, $0–400) on transportation to the facility, while women seeking later abortions spent an average of $100 (range, $0–2,200).

Almost two-thirds of the women seeking later abortion and fewer than one-third of those seeking a first-trimester abortion said they were delayed because they were raising money for travel, the procedure and other costs.

Insurance coverage

Finally, difficulties securing insurance coverage for the abortion delayed some patients. A 16-year-old white woman in California, who had her abortion at 23 weeks’ gestation, said that what slowed her down was Medi-Cal, the California Medicaid program, which covers abortion for low-income women: “I got [the] runaround from the offices about getting on it.” Delays were not limited to women with public insurance coverage. A 28-year-old multiethnic woman from Washington, who was at 24 weeks’ gestation at the time of her abortion and had private insurance, said that what slowed her down was “difficulties with [her] insurance company.”

Likely owing to the higher cost of later procedures and associated importance of insurance coverage, women seeking later abortions were twice as likely as women seeking first-trimester abortions to report delays because of difficulties securing public or private insurance coverage for the abortion (41% vs. 20%).

Profiles of Later Abortion Patients

Our factor analysis identified five general profiles of women who sought later abortions, describing 80% of the sample. Forty-two percent of later abortion patients fit only one group; 38% belonged in two or more groups.

Raising children alone

Angel, a 24-year-old white woman from Maryland, represents the profile that describes the largest proportion of our sample (47%). At the time of her abortion, she had a 10-month-old daughter, whom she cared for full-time while she looked for paid employment. Her husband had recently been incarcerated, leaving her with no household income. As Angel explained, her daughter was her top priority. When she realized she was pregnant, at 22 weeks, her principal concern was for her daughter. Angel's experience of being a new mother interfered with her ability to detect her pregnancy and, moreover, convinced her that having another child was a bad idea. She believed that having another child would compromise the care she could give her infant daughter: “I knew I couldn't continue with [the pregnancy]. My daughter isn't even a year.” Deciding to have the abortion was very easy for Angel.

Angel had difficulty finding a clinic where she could obtain an abortion. After visiting one facility that could not help her, she found another, three hours away. As late as she was in her pregnancy, the cost of the procedure was daunting: $2,700. But as Angel said, “I was determined.” She paid $300 herself, borrowed $400 from her mother and received aid from three funds that help low-income women pay for their abortions. She had her abortion at 24 weeks.

Depressed or using illicit substances

The second most common profile of women seeking later abortion (describing 30% of the sample) is characterized by substance use or mental health problems. Representing this group, Rose, a 25-year-old American Indian woman from Oklahoma, was living with her boyfriend and was financially stable when she had her abortion. She was employed full-time as a manager and, aside from some chronic back and joint pain, enjoyed good physical health. Her life was not without challenges, however. She had a history of major depression and bipolar disorder, for which she was receiving medication and cognitive behavioral therapy. Twice, when she was 12 and 22, she was physically assaulted, experiences she says have had a negative impact on her life. Rose used Xanax recreationally about once a week and, in the month prior to her interview, used amphetamines and had episodes of binge drinking. Both depression and illicit substance use are associated with delays in recognition of pregnancy.[22, 23]

Rose explained in her interview, “I don't want kids.” When she found out she was pregnant, at 19 weeks, she was “shocked and not happy about it.” Rose had been using oral contraceptives and had trusted their efficacy. Indeed, her main feeling about her pregnancy was, she said, “frustration.”

Deciding to have an abortion was easy for Rose. Finding a clinic where she could obtain an abortion, however, was much more difficult. Rose called four clinics and visited another one before finding a clinic four and a half hours from her home that could perform her abortion at 20 weeks, for which she paid $1,750.

Conflict with male partner or domestic violence

Even women who know of their pregnancies early on may seek later abortions because of the circumstances of their lives, including domestic violence. Lesley, a 34-year-old white woman from Kentucky who worked as a medical professional, was representative of this profile, which describes 24% of our sample. She lived with her eight-year-old daughter and cared for her disabled mother. For the previous 11 years, since she married, she had been physically abused by her alcoholic husband. This has had an extremely negative impact on her life, including causing episodes of depression and symptoms of posttraumatic stress disorder. Research shows that experiences of intimate partner violence can be a risk factor for unintended pregnancy.[33, 34]

Even women who know of their pregnancies early on may seek later abortions because of the circumstances of their lives.

Lesley's husband was the man responsible for her pregnancy. She discovered the pregnancy at five weeks’ gestation and was deeply conflicted about whether to carry it to term. Victims of violence, like Lesley, may have different reasons for seeking abortion than nonabused women,[35] and they may delay the decision-making process because of the abusive relationship.

Lesley left her husband. After counseling from a private therapist and with the support of two friends, Lesley got an abortion at 20 weeks’ gestation. The procedure cost her $1,700. Although she felt guilty about the abortion, she said it was the right decision for her, and she had high hopes for the future. She hoped that in a year, she would be divorced and raising her daughter as a single mother.

Trouble deciding and access problems

Difficulty in accessing abortion alone can contribute to the need for later abortion, even if a pregnancy is recognized early; this was true for 22% of the sample. Amber, a 24-year-old black woman from Alabama, had two children, the younger of whom was four years old. Amber worked part-time and, because the father of her children failed to provide the child support he was responsible for, rarely had enough money. She received food stamps for her family and was insured through the state Medicaid program.

Amber became pregnant with her ex-boyfriend. She discovered the pregnancy at five weeks and was considering abortion by six weeks. Her ex-boyfriend was not supportive of her continuing the pregnancy, and Amber knew she could not financially support another child. But she initially equivocated, unsure if she wanted an abortion. She eventually decided to have an abortion because, as she said, “I couldn't afford another child. The dad didn't want to be with me. Me and him weren't going to be together, and he told me that I was going to have to raise the baby myself.”

New obstacles emerged after Amber decided to terminate her pregnancy. Amber called two clinics and visited another before finding a clinic three hours from her home that could perform her abortion. She had trouble securing transportation, and during this delay, her gestation advanced such that she required a different and more costly procedure, which she struggled to afford. In the end, she received nearly $600 from an abortion fund, her ex-boyfriend paid $600 and she paid $300 for the abortion. She also paid $150 in transportation costs. By the time she had secured this money to cover the procedure and her travel, Amber was 20 weeks pregnant. Although Amber “tries not to think about” the abortion, she said it was the right decision for her.

Young and nulliparous

Twelve percent of the women were young and nulliparous. Like many women seeking later abortion, Lana, a Hispanic woman from California, did not learn she was pregnant until 21 weeks into her pregnancy. Just 15 years old, Lana was a full-time high school student. She lived with her aunt and brother in a household that received food stamps. Lana became pregnant with her boyfriend, despite their condom use. Although her boyfriend wanted her to continue the pregnancy, Lana decided on abortion immediately. Her age and inability to support a baby financially were key factors in her decision. She explained, “I was too young, and I barely started going back to school and getting my life back on track. I wouldn't have enough things to support a baby.” Nonetheless, she described the decision as “somewhat difficult.”

Lana had her abortion at 23 weeks’ gestation. She had to travel four hours to reach a facility that could perform the procedure, the cost of which was covered by public insurance. After the abortion, Lana said she felt relief.

Discussion

In contemporary discussions of later abortion, few empirical data exist about the women who seek these procedures, but speculation abounds, including the presumption that these women are intrinsically different from those seeking early abortions. Despite this common narrative, women in our study who obtained first-trimester abortions and women who obtained abortions at or after 20 weeks’ gestation were remarkably similar. Demographically, the only significant differences between the two groups were in age (women aged 20–24 were more likely than women aged 25–34 to seek later abortion) and in employment status (employed women were less likely than unemployed women to have later abortions).

Participants’ reports of their experiences of deciding to obtain an abortion and of accessing care further show similarities that run counter to the popular narrative of difference between these two groups. Indeed, three reasons for delay we identified—the woman's not knowing she was pregnant, having trouble deciding whether she wanted an abortion and disagreeing with the man involved in the pregnancy on the abortion decision—were equally common among early and later abortion seekers. These commonalities point to the important finding that obstacles to receiving abortion care exist regardless of gestational age. Moreover, they are consistent with research findings on reasons for second-trimester abortion.[22-24, 27] Our findings deviate from the literature on delays in other ways. We did not find that changing personal circumstances contributed to delays in seeking abortion care, as they do in the United Kingdom,[27] perhaps because access and finance-related obstacles in the United States overwhelm any such effect.

There are important limitations to analogizing the experience of delay across gestational ages. Obstacles to receiving care were common among both study groups, but these barriers had greater consequences for later abortion seekers. Most women in the later abortion group were already in the second trimester by the time they had decided to have an abortion. The subsequent logistical challenges were more common among the later abortion group and of greater magnitude, as second-trimester abortion providers are harder to find and procedures more expensive; these findings suggest a snowballing effect, by which each delay increases the likelihood and magnitude of the next. The greater magnitude of these delays increased the divergence between the two groups from a seven-week gap at pregnancy recognition (five vs. 12 weeks) to an average gap of 14 weeks by the time they received their abortions (eight vs. 22 weeks). Rather than intrinsic characteristics, the mounting complications of each obstacle faced resulted in some women's being in the first-trimester group and others’ in the later abortion group. This finding is an important corrective to presumptions that women in these two groups differ because of individual characteristics such as personality, disposition or sense of responsibility.

Obstacles to receiving care were common among both study groups, but…had greater consequences for later abortion seekers.

Addressing the social bias against women seeking later abortions becomes of increasing importance as states consider legislative bans on abortion at or after 20 weeks. In some cases, this deadline for abortion may speed up women's decision making. However, our findings suggest that several of the reasons women delay seeking abortion (e.g., late discovery of pregnancy, conflict with the man involved) cannot be reduced by shortening the window for abortion care. As legislative bans on post–20-week abortions are already in effect in some states, future research should explore the consequences of these laws. Our data show that such bans will disproportionately affect young women and women with limited financial resources. These bans may make exceptions for women seeking abortions for reasons of fetal anomalies or life endangerment, but women seeking later abortions for any other reasons—including the reasons we identify above—are unlikely to get their abortions.

Our study has several important limitations. Our data are limited by the exclusion of women who sought later abortions on grounds of fetal anomaly or life endangerment. Because of waiting time for testing and diagnosis, delay in seeking care among that population likely differs significantly from the delays faced by women in our study. In addition, while some aspects of our findings may be generalizable beyond the United States, the results are largely limited in their applicability to other cultural and legal environments.

The account we have presented of women who seek later abortions is not meant to supersede consideration of the fetus in the abortion debate. Rather, we are offering an important corrective to discussions that have conceptualized later abortions exclusively in terms of the fetus, without a portrait of the women seeking them.

Acknowledgments

The authors thank Rana Barar, Heather Gould and Sandy Stonesifer for study coordination and management; Janine Carpenter, Undine Darney, Ivette Gomez, Selena Phipps, Claire Schreiber and Danielle Sinkford for conducting interviews; Michaela Ferrari and Elisette Weiss for project support; Jay Fraser and John Neuhaus for statistical and database assistance; and all the participating providers for their assistance with recruitment. This study was supported by research and institutional grants from the Wallace Alexander Gerbode Foundation, The David and Lucile Packard Foundation, the William and Flora Hewlett Foundation and an anonymous foundation.

Endnotes

  1. 1

    To meet the objectives of the larger study, for every woman we recruited who was turned away because of advanced gestation, we recruited two women who received the procedure at a gestation within two weeks of the facility limit and one woman who received a first-trimester abortion.

  2. 2

    The social and demographic items were being younger than 20, being 35 or older, having no children, having had no previous abortions, having a baby younger than one year, raising a child alone and having a monthly income of less than $1,000.

  3. 3

    The six causes of delay were having discovered the pregnancy late (defined as after 13 weeks, or the end of the first trimester), been discouraged from having an abortion by the people who knew about the pregnancy, had conflict with the man involved in the pregnancy, had trouble deciding, had difficulties accessing abortion services and had difficulties paying for the abortion.

  4. 4

    The physical health items were being obese, having experienced physical abuse in the past year and having any of the four specified chronic health conditions.

  5. 5

    The mental health and substance use items were use of recreational drug in the past month, history of major depression and reports of episodes of heavy drinking.

Ancillary