Characteristics of Immigrants Obtaining Abortions and Comparison with U.S.-Born Individuals

Abstract

Background: Little information exists about individuals born outside of the United States who seek abortion services from U.S.-based providers. Baseline data are necessary to identify future changes in the profile of this population.

Materials and Methods: Using the Guttmacher Institute's Abortion Patient Survey, we pooled two national samples of individuals obtaining abortions from 2008–2009 to 2013–2014 to provide data on 17,873 respondents, 16% of whom were immigrants. We estimated the distribution of immigrant and U.S.-born respondents across demographic and circumstantial characteristics such as age, poverty level, and gestational age at abortion. We compared the distribution of characteristics by nativity status using chi-square tests.

Results: The majority of immigrants obtaining abortions were in their 20s (51%), had poverty-level (50%) or near poverty-level incomes (23%), and had graduated from high school (78%). Almost half (45%) were uninsured and a similar proportion had been in the United States for less than 10 years (44%); nearly one-quarter completed their survey in Spanish. Compared with U.S.-born respondents, a larger proportion of immigrants were older, uninsured, and had not completed high school. A smaller proportion of immigrants compared with nonimmigrants had their abortions after 12 weeks (8% vs. 11%) or traveled over 50 miles to obtain their abortion (9% vs. 16%).

Conclusions: Particularly with the continued rise in both restrictive abortion and immigration policies in the United States, it is critical to monitor how immigrants' use of and access to abortion services are impacted in the changing environment. Ensuring that policies and clinical practices facilitate abortion access for immigrants will serve to better support the reproductive health needs of all women.

Introduction

Although immigrants account for 17% of women of reproductive age in the United States and 23% of births,¹ little is known about their use of abortion services. In the context of the U.S. health care system, immigrants are more likely to be uninsured, to receive culturally and linguistically inappropriate care, and to make fewer health care visits relative to the U.S.-born population.^2–4

Immigration policies may contribute to these inequities.³ For example, immigrants are barred from using public insurance programs, such as Medicaid, based on their legal status or duration of residence in the United States.⁵ Stricter enforcement of federal immigration laws, coupled with increased state-level anti-immigrant legislation, has been shown to deter immigrants and their families from seeking needed health care for fear of discrimination or legal action.^6,7 Combined with mounting legal restrictions on abortion,⁸ the present political climate could widen disparities in reproductive health access and outcomes between immigrants and nonimmigrants.

Research on the reproductive health of immigrant women suggests that they are less likely than nonimmigrants to use sexual and reproductive health services such as contraceptive counseling, reproductive cancer and sexually transmitted infection screenings, and pre- and postnatal care.^9–12 A recent study also found that undocumented women are less likely than other immigrant groups to obtain adequate levels of prenatal care.¹³ Indeed, immigrants in the United States are not a monolith and heterogeneity in immigration status, length of stay in the United States, country of origin, and other dimensions of the immigration experience could constrain some groups' use of services more than others.

When reproductive health services are obtained, a higher proportion of immigrant than nonimmigrant women seek care from publicly funded clinics, which are also considered the usual source of medical care for almost three-quarters of foreign-born women.¹⁴ Immigrant women may face even greater difficulty than before to obtain routine reproductive health care, given recent policy efforts to curb Title X funding for comprehensive family planning services and restrict public funding to providers such as Planned Parenthood health centers.¹⁵

On the surface, it would seem that differences in reproductive health care use and access may not extend to abortion care insofar as immigrants obtain abortions at a rate comparable with nonimmigrants.¹⁶ Yet, despite similar use of abortion services, identifying differences between immigrants and nonimmigrants obtaining abortions is necessary to assess the reproductive health service and policy needs of immigrants in the United States. The majority of women obtaining abortions in the United States are in their 20s, unmarried, people of color, graduates of high school, and already parents.^17,18 However, it is unknown whether immigrant women obtaining abortions have the same demographic profile. To that end, this study aims to describe the characteristics of immigrants obtaining abortions in the United States in 2008 and 2014 and compare these to nonimmigrants obtaining abortions during the same time period.

This article builds on prior studies of individuals obtaining abortions in the United States by focusing on a population that has not previously been examined in the published literature on abortion, addressing a critical gap in both immigrant and reproductive health research. Findings from this study may help inform health policies and clinical practices to better serve and support the reproductive health needs of immigrants. Furthermore, the continued rise in abortion restrictions, coupled with stricter immigration-related policies, could adversely impact immigrants' ability to obtain abortion services or make it more difficult for certain subgroups than others to do so. Findings from this study should provide important baseline data to identify future changes in the profile of immigrants obtaining abortions in the United States, which could also inform emergent policy issues affecting this population.

Materials and Methods

Data for this analysis came from the Guttmacher Institute's 2008 and 2014 Abortion Patient Survey (APS), a national sample of individuals obtaining abortions in the United States. Both rounds of data collection used a similar sampling design, questionnaire, and fieldwork protocol to that of previous APS surveys in 1987, 1994–1995, and 2000–2001.^18–21 One notable exception was that the 2014 survey did not include individuals obtaining abortions in hospitals. As hospital abortions made up only 4% of total abortions in that year, excluding these facilities from the sampling frame was not expected to have a substantive impact.²² Participating facilities were randomly selected and recruited, and all women obtaining abortions at these facilities during a specified fielding period were asked to complete a four-page, paper-and-pencil, self-administered questionnaire, available in English and Spanish. Respondents were provided with a sealable envelope in which to return the survey so that their responses would not be seen by staff. The 2008 APS had a 74% respondent response rate; the survey collected information from 9,493 respondents from 95 abortion care facilities across the United States. The 2014 APS collected information from 8,380 respondents from 87 nonhospital abortion care facilities for a 76% respondent response rate. Data from each survey round were weighted to create a nationally representative sample of abortion patients in 2008 and 2014. A detailed description of the data collection and weighting procedures can be found in previously published studies.¹⁷ The APS and data collection procedures were approved by the Guttmacher Institute's Institutional Review Board.

To robustly study the characteristics of immigrants obtaining abortions, we pooled the 2008 and 2014 APS data to increase the sample size of immigrants. The demographic profiles of all abortion patients in 2008 and 2014 have been previously compared and found to be similar. While a smaller proportion APS respondents in 2014 were adolescents and uninsured, and a larger proportion were poor,^16,17 these changes were attributed to changes in the abortion patient population, not to differences in the survey design. We constructed new probability weights to appropriately account for the change in the number of abortions from 2008 (N = 1,212,350) to 2014 (N = 926,187) using two steps: (1) dividing the number of total abortions in year_x by the number of abortion respondents in the APS from year_x (where x is either 2008 or 2014); and (2) multiplying this adjustment factor by the existing weights in each APS round. Although hospitals were not sampled in the 2014 APS, the hospital data from 2008 were retained in the pooled sample as 4.5% of immigrant respondents from that survey round obtained care at a hospital. The pooled sample included a total of 17,873 individuals who had obtained an abortion in 2008–2009 and 2013–2014.

Nativity status was constructed based on responses to the yes/no survey question: “Were you born in the United States?” Respondents who answered “yes” were categorized as U.S.-born and those who answered “no” as immigrants. Some 426 respondents, 2% of the sample, did not answer this question, responses were imputed using the information from respondents with similar characteristics (see description of imputation strategy in the next paragraph). We conducted a sensitivity analysis that excluded cases with missing information on this variable, and the findings were unaltered (not shown). Immigrants comprised 16% (n = 2,790) of the analytic sample, proportional to the share of immigrants in the U.S. population of women of reproductive age during both time periods.^16,23 We examined key demographics, including age, race and ethnicity, health insurance status, relationship status, poverty level (measured relative to an annual income of ∼$24,000 for a four-person household, as per the Health and Human Services poverty guidelines²⁴), level of education (among women 20 years and older), residence in a Metropolitan Statistical Area (MSA), region of residence, survey language, and length of stay in the United States (for immigrants only).

We also assessed differences by nativity status in select situational variables such as gestation when the abortion was obtained (categorized as ≤12, 13–15, and ≥16 weeks since the last menstrual period), prior births (0, 1–2, and ≥3 births), prior abortions (yes or no), distance traveled to obtain services (<25, 25–49, 50–100, and >100 miles), experience of intimate partner violence (yes or no), and exposure to disruptive life events experienced over the last year (0, 1, ≥2 events). Disruptive life events included exposure to any of eight events such as death of a friend, financial difficulties, and medical issues. We also examined if and how respondents had ever attempted to self-manage an abortion; although abortion is legal in the United States, some individuals still obtain, or attempt to obtain, abortions outside of formal clinic settings.²⁵ In addition, immigrant women who previously resided in a country where abortion was so restricted as to be practically illegal may have had no option but to rely on self-managed abortion before emigrating.

Most variables had missing information for ∼1%–4% of cases. Three exceptions were distance traveled to obtain an abortion, family income, and length of stay in the United States, with missing values of 10%, 14%, and 19%, respectively. Missing values for all key demographic variables, including family income, were imputed using a “hot-deck” procedure. This strategy identifies variables most strongly associated with each item requiring imputation, and sorts the data file accordingly to replace the missing value with that from a similar, adjacent case. Analysis of variables that were not imputed—length of stay, distance traveled, exposure to intimate partner violence, and prior attempts of self-managed abortion—excluded individuals who did not answer the relevant question(s). We conducted a sensitivity analysis that excluded respondents who had imputed values on the variables of age, race and ethnicity, health insurance, union status, income/poverty, education, gestation, and prior births and abortions. Patterns between immigrants and nonimmigrants were virtually the same on all characteristics.

In this article, we compare the demographic and reproductive health profiles of immigrant and U.S.-born individuals receiving abortion services in the United States in 2008 and 2014, with a focus on describing the characteristics and circumstances of immigrants. We use chi-square to test for significant associations between respondent characteristics and nativity status. Differences were considered significant at a level of p < 0.05. All analyses were conducted using Stata version 15.1 and weighted to account for individual-level nonresponse, variation from the original facility sampling plan, and the change in number of abortions between 2008 and 2014.

To assess if demographic differences by nativity status were proportional to differences in the underlying population of immigrant and nonimmigrant women in the United States, we conducted a sensitivity analysis comparing the percent distribution of these groups in the pooled APS sample to the percent distribution of these groups in the American Community Survey (ACS) across select demographic characteristics. Full methodological details for the ACS are available through the U.S. Census Bureau.²⁶ We pooled the 2008 and 2014 1-year supplemental files of the ACS to estimate distributions of age group, race and ethnicity, poverty status, education level (among women 20 years and older), and region of residence. Comparisons were calculated as the proportion of immigrant abortion patients in a given subgroup (e.g., a particular age group) relative to the proportion of all reproductive-aged (15–44 years) immigrant women in the United States in that same subgroup. If the ratio from this comparison was the same as that for nonimmigrants, the difference in the distribution of this characteristic by nativity status was considered proportional to the difference observed between immigrants and nonimmigrants in the underlying population. Ratios that were not similar would suggest that the observed difference between immigrant and nonimmigrant abortion patients was specific to women seeking abortion.

Results

The majority of immigrants obtaining abortions were in their 20s (51%), had poverty-level (<100% of the federal poverty level [FPL]) (50%) or near poverty-level (100%–199% of the FPL) incomes (23%), and had graduated from high school (78%). Most commonly, immigrants in the sample were Hispanic (49%) and one in five was Asian (20%). They were about equally likely to live in the south (33%) or west (31%) of the country. Almost half (45%) were uninsured and a similar proportion had been in the United States for less than 10 years (44%); nearly one-quarter completed their survey in Spanish. This demographic profile was substantially different from that of U.S.-born individuals obtaining abortions (Table 1). While the majority of both groups were in their 20s, a larger proportion of immigrants were older; 40% were aged 30 and older compared with 23% of nonimmigrants. As noted, immigrant women were predominantly Asian and Hispanic, while the most common racial identities among nonimmigrants were non-Hispanic white (43%) and non-Hispanic black (31%). Furthermore, compared with immigrant women, a smaller share of nonimmigrants obtaining abortions lacked insurance coverage (32%) and were not married (31% vs. 12%). Although the majority of both groups had attended some college or obtained a college degree, a larger proportion of immigrants (ages 20 and older) had not completed high school (22% vs. 8%). A smaller proportion of immigrants (7%) compared with nonimmigrants (17%) lived in the Midwest region of the country. Each of these associations was significant at p < 0.001.

Table 1.

Weighted Distribution of Immigrants and Nonimmigrants Obtaining Abortions in U.S. Facilities by Selected Demographic Characteristics, 2008–2014

Characteristic	All abortion patients (N = 17,873)
	Immigrantsa(n = 2,790)		Nonimmigrantsa(n = 15,083)		p
	N	%	N	%
Age, years
<18	76	3	857	6	<0.001
18–19	152	5	1,570	11
20–24	692	24	5,406	35
25–29	759	27	3,773	25
30–34	554	20	2,001	13
≥35	557	20	1,476	10
Race/ethnicity
Hispanic	1,361	49	2,928	20	<0.001
Asian	554	20	340	2
Non-Hispanic black	428	15	4,726	31
Non-Hispanic white	296	10	6,510	42
Otherb	151	6	579	4
Health insurance
No coverage	1,236	45	4,805	32	<0.001
Medicaid	756	27	5,067	34
Private	766	27	4,968	32
HealthCare.gov/State exchangec	32	1	243	1
Relationship status
Married	845	30	1,739	12	<0.001
Cohabiting, not married	655	24	4,706	31
Never-married, not cohabiting	892	32	7,243	48
Previously married, not cohabiting	398	14	1,395	9
Poverty status, %
<100	1,402	50	6,769	45	<0.01
100–199	657	23	4,019	27
≥200	731	27	4,295	29
Highest level of educationd
Less than high school	561	22	1,017	8	<0.001
High school graduate/GED	682	26	3,603	29
Some college	693	27	5,526	43
College graduate	626	25	2,510	20
Resides in MSA
No	163	6	1,720	12	<0.001
Yes	2,327	94	12,090	88
Region of residence
Northeast	698	27	3,325	23	<0.001
Midwest	266	9	2,534	17
South	903	33	5,469	36
West	917	31	3,755	24
Survey language
English	2,105	76	15,026	99.6	<0.001
Spanishe	685	24	57	0.4
Length of stay in the United Statesf
<5 years	490	22	—	—	—
5–9 years	498	22	—	—
10+ years	1,283	56	—	—

^aCounts may not sum to the total number of abortion patients due to missing data.

^bIncludes respondents who identified as American Indian or Alaskan Native, Native Hawaiian or Pacific Islander, or other race.

^cThese data were only collected in 2014.

^dAmong women ages 20 years or older.

^eIncludes seven respondents who completed the survey in Portuguese.

^fOnly asked of foreign-born (i.e., immigrant) respondents.

MSA, metropolitan statistical area.

With the exception of variations by race and ethnicity and poverty level, the demographic differences reflected those in the composition of the underlying population of immigrant and nonimmigrant women in the United States. For example, a higher proportion of immigrant abortion patients did not have a high school degree compared with nonimmigrants, and this educational difference was also seen among the larger population of U.S. women. Similar to the demographic profile of all individuals seeking abortion in the United States,¹⁷ a larger share of immigrants in our sample were younger and had poverty-level incomes compared with the population of immigrant women in the United States (Appendix Table A1).

Nearly all immigrants (92%) in our sample had their abortion in the first trimester (≤12 weeks) and 80% traveled under 25 miles for their procedure (Table 2). Over half (68%) were already parents and 52% had not had a prior abortion. Less than 2% reported ever having attempted to self-manage an abortion. These patterns differed from nonimmigrants on most characteristics. A smaller proportion of immigrants compared with nonimmigrants had their abortions after 12 weeks (8% vs. 11%) or traveled over 50 miles for the procedure (9% vs. 16%) (chi-square p < 0.001). Some 41% of nonimmigrants had not had a prior birth compared with 33% of immigrants (chi-square p < 0.001); still, the majority of women in both groups were already parents. Although nearly one-third of each group reported exposure to one disruptive event in the last year, 25% of nonimmigrants reported exposure to two or more events compared with 15% of immigrants (chi-square p < 0.001). Notably, there was no difference between the proportion of immigrant and nonimmigrant women who reported a prior attempt to self-manage their abortion or reported intimate partner violence.

Table 2.

Weighted Distribution of Immigrants and Nonimmigrants Obtaining Abortions in U.S. Facilities by Select Reproductive Health and Situational Characteristics, 2008–2014

	All abortion patients (N = 17,873)
	Immigrantsa(n = 2,790)		Nonimmigrantsa(n = 15,083)		p
	N	%	N	%
Gestation (weeks since LMP)
≤12 weeks	2,568	92	13,510	89	<0.001
13–15 weeks	123	4	954	6
≥16 weeks	99	4	619	4
Prior births
0	930	33	6,241	41	<0.001
1–2	1,597	58	7,920	53
≥3	263	9	922	6
Prior abortions
No	1,369	52	7,715	53	0.827
Yes	1,216	48	6,756	47
Prior attempt to self-manage an abortion
Yes, using misoprostol	29	1	142	1	0.733
Yes, using other substances	21	1	98	1
No attempt	2,740	98	14,843	98
Distance traveled to facility
<25 miles	1,974	80	9,484	70	<0.001
25–49 miles	268	11	1,931	14
50–100 miles	120	5	1,316	9
100+ miles	110	4	945	7
Experience of intimate partner violenceb
No	2,496	95	13,722	94	0.181
Yes	146	5	908	6
Exposure to disruptive life eventsc
0 events	1,462	53	6,450	43	<0.001
1 event	904	32	4,816	32
≥2 events	424	15	3,817	25

^aCounts may not sum to the total number of abortion patients due to missing data.

^bBy the man with whom respondent became pregnant.

^cTypes of disruptive events include: a close friend died; behind on rent/mortgage; had a baby; partner was incarcerated or arrested; moved 2+ times; separated from partner; unemployed for a month or more; and a family member had a serious medical problem.

LMP, last menstrual period.

Discussion

Individuals obtaining abortion in the United States have long been an economically and socially marginalized population.^17,21,23 Combined data from the 2008 and 2014 national surveys of individuals obtaining abortions suggest that some of these vulnerabilities were even more pronounced for respondents born outside of the United States. In particular, a larger proportion of immigrants in the APS had not graduated from high school and did not have health insurance. The latter may be influenced by policies such as the 5-year ban on Medicaid that bars many immigrants from eligibility for public health insurance programs, as well as immigration enforcement that deters participation in these programs.²⁷

The data also find that a higher proportion of immigrants obtaining abortions had family incomes below the FPL. Immigrant status can constrain options for employment or limit opportunities to low-wage jobs, which could impact family incomes. The overwhelming majority of immigrants in our sample were women of color; given the pervasive history of racism and xenophobia in the United States, these individuals may face discrimination and hostility based on their race, culture, and nativity, in general and within the medical system, that could impede their access to other reproductive health care.^28–30 Similarly, that nearly one in five immigrants filled out the survey in Spanish could indicate that some have limited proficiency in English. Such language barriers could serve as an additional source of discrimination and also limit individuals' ability to find and obtain necessary care in communities and health systems where only English is used.

Immigrants obtaining abortions differed from their U.S.-born counterparts in several other ways. That immigrants were older reflects the age distribution of the underlying population, and may indicate, for example, how some immigrants do not enter the United States until adulthood. Abortion patients born outside of the United States were more likely to live in a metropolitan area compared with U.S.-born individuals and, relatedly, they traveled shorter distances to obtain care. These patterns may reflect the concentration of immigrants in urban areas, which also have a higher density of abortion providers.³¹ This proximity to abortion services may help offset other potential barriers to care. Alternately, it is possible that immigrants were unable to find abortion providers located outside of their metropolitan area or, even if located, lacked the resources to travel for care.³² Indeed, previous research found that white, educated, and higher income women—groups that traditionally have access to resources and privilege—travel farther for abortion services than their counterparts.^33,34

Our findings suggest that a smaller proportion of immigrants in the APS obtain abortions after the first trimester compared with those U.S.-born. Similar to the above, it is possible that the higher cost of second-trimester abortions, coupled with the lack of facilities that provide this care, requires individuals to travel farther for their abortion, potentially making these services less accessible to immigrant women. Alternately, as immigrant abortion patients were older and a higher proportion had had children, it is possible they were able to recognize their pregnancies earlier than nonimmigrants. Subsequent research should explore the factors that contribute to patterns in gestation among immigrants and nonimmigrants obtaining abortions.

There were several characteristics on which immigrant and U.S.-born respondents did not differ or differed in unexpected ways. Among both groups, similar proportions reported having a prior abortion and, despite prior evidence, ever attempting to self-manage an abortion.³⁵ Given that some immigrants in our sample undoubtedly came from countries where abortion is highly restricted, we expected a lower proportion to have had a prior abortion but a higher proportion to report prior attempts to self-induce. The proportion of immigrant and U.S.-born abortion patients who reported experiencing intimate partner violence was comparable, and the proportion of immigrants who reported experiencing one or more disruptive events in the past year was lower than that for U.S.-born individuals. However, none of the experiences we asked about pertained directly to immigration-related stressors (e.g., had to produce proof of legal residency or family separation), and, in turn, we may have not captured the full range of disruptive events for this population. Alternately, since immigrants were older and a higher proportion was married compared with nonimmigrants, they may have had more stable lives and experienced fewer disruptive events.

This study has several limitations. Perhaps most importantly, it only contains data from individuals who were able to access clinical abortion services. Women who wanted an abortion but were unable to have one—whether due to lack of information, economic resources, or providers in their geographic area—were not captured in these data. Similarly we did not obtain information from individuals who were able to successfully self-manage their abortions outside of a clinical setting. Immigrants could be differentially excluded from this study if they are more likely than their counterparts to obtain abortions in nonclinical settings or face barriers that altogether prevent access to abortion. It is also possible that respondents born outside of the United States provided inaccurate information on their nativity status or were more likely than nonimmigrants to decline to fill out the survey. Similarly, the survey may have been too time-consuming or difficult to complete for individuals for whom English (or Spanish) was not their primary language. Such language barriers could differentially impact survey and study participation among immigrants compared with nonimmigrants. Furthermore, we were unable to examine specific immigrant groups in this study, despite the heterogeneity of the immigrant population. Because of substantial missing data on length of stay in the United States; small cell sizes by their race and ethnicity; and no data on country of origin, or immigration status (our survey did not collect this information), we were unable to disaggregate our analyses by these factors, which have been documented to influence immigrants' health service use.⁹ Still, our study provides more information about immigrants obtaining abortions than was previously known. Finally, combining noncontinuous abortion data from 2008 and 2014 may have masked temporal changes in the population of individuals obtaining abortions; however, we derived and used pooled weights as one approach to account for such population shifts.

Conclusions

For the last decade, nearly one in five individuals obtaining abortions was an immigrant, and findings from this study suggest several clinical implications. Nearly one-quarter of immigrants filled out the survey in Spanish, which suggests the need for multilingual clinic resources (e.g., consent forms, follow-up instructions) and staff to facilitate accessible and comprehensive care for all people seeking abortion. Similarly, fewer years of education and lower levels of insurance among immigrants may impose additional challenges to navigating the health care system, including abortion care. This study also provides an important baseline for comparing subsequent studies of immigrants obtaining abortions. Increased immigration enforcement has already had a documented chilling effect on health care use and access among immigrants.^27,36 Further study of immigrants' abortion access, particularly related to the influence of immigration policy on abortion care as well as the experiences of immigrants who are unable to access this care, will be critical. This information will help bring to bear if and how the changing policy environments impact immigrant women's abortion use, and ultimately, better serve and support the reproductive health needs of all women seeking abortion.

Appendix Table A1.

 Percent Distribution of Individuals Obtaining Abortions in the United States and of All U.S. Women Ages 15–44 by Nativity Status, and Comparison of Both Populations by Selected Characteristics, 2008–2014

Characteristic	Individuals obtaining abortionsa					All women ages 15–44b		Ratio of all U.S. women to those obtaining an abortion
	Immigrantsc(n = 2,790)		Nonimmigrantsc(n = 15,083)		p	Immigrants	Nonimmigrants	Immigrants	Nonimmigrants
	N	%	N	%		%	%	%	%
Age, years
<18	76	3	857	6	<0.001	3.9	10.9	0.7	0.5
18–19	152	5	1,570	11		3.4	7.6	1.6	1.4
20–24	692	24	5,406	35		11.5	18.0	2.1	2.0
25–29	759	27	3,773	25		16.5	16.8	1.6	1.5
30–34	554	20	2,001	13		20.3	15.4	1.0	0.9
≥35	557	20	1,476	10		44.3	31.3	0.5	0.3
Race/ethnicity
Non-Hispanic white	296	10	6,510	42	<0.001	14.7	67.6	0.7	0.6
Non-Hispanic black	428	15	4,726	31		8.5	15.8	1.8	2.0
Hispanic	1,361	49	2,928	20		49.1	12.8	1.0	1.6
Asian	554	20	340	2		26.6	2.6	0.7	0.9
Otherd	151	6	579	4		1.1	1.2	5.3	3.2
Poverty status, %
<100	1,402	50	6,769	45	<0.01	22.6	17.3	2.2	2.6
100–199	657	23	4,019	27		24.6	18.0	1.0	1.5
≥200	731	27	4,295	29		52.8	64.7	0.5	0.4
Highest level of educatione
Less than high school	561	22	1,017	8	<0.001	25.1	7.1	0.9	1.2
High school graduate/GED	682	26	3,603	29		22.0	22.5	1.2	1.3
Some college	693	27	5,526	43		22.9	39.1	1.2	1.1
College graduate	626	25	2,510	20		30.0	31.3	0.8	0.6
Region of residence
Northeast	698	27	3,325	23	<0.001	21.0	17.1	1.3	1.3
Midwest	266	9	2,534	17		11.7	22.7	0.8	0.7
South	903	33	5,469	36		32.4	38.5	1.0	0.9
West	917	31	3,755	24		34.9	21.8	0.9	1.1

^aData source: Abortion Patient Survey, combined 2008–2009 and 2013–2014.

^bData source: American Community Survey, combined 2008 and 2014 1-year supplemental files.

^cCounts may not sum to the total number of abortion patients due to missing data.

^dIncludes respondents who identified as American Indian or Alaskan Native, Native Hawaiian or Pacific Islander, or other race.

^eAmong women ages 20 years or older.

Author Disclosure Statement

No competing financial interests exist.