Download a PDF of this Backgrounder
Steven A. Camarota is the Director of Research at the Center for Immigration Studies.
There may be no more contentious issue in modern America than abortion. This report explores this important issue by using the Guttmacher Institute's Abortion Patient Survey to examine abortion rates for immigrants and natives separately. The findings show immigrants overall have nearly identical abortion rates to native-born Americans, but significant differences exist across racial and ethnic groups.
- Immigrants accounted for about one in six abortions (194,000) in the United States in 2008.
- For immigrant women in their primary reproductive years (ages 15 to 44), the abortion rate is 18.9 per thousand, very similar to the 19.5 per thousand for native women. This is not a statistically significant difference.
- Immigrants generally arrive after age 25, so only a modest share are in the youngest age groups, which have the highest abortion rates. Excluding those under age 25, the abortion rate for immigrants is 15.6 per thousand, somewhat higher than the 14.1 per thousand for natives in the same age range. This difference is statistically significant.
- Looking at abortion by demographic groups among immigrant women ages 15 to 44, the rate is 34.7 per thousand for blacks, 19 per thousand for Hispanics, 16.5 per thousand for Asians, and 9.1 per thousand for non-Hispanic whites.
- Among native-born women ages 15 to 44 the abortion rate is 41.6 per thousand for blacks, 36.7 per thousand for Hispanics, 26.2 per thousand for Asians, and 11.4 per thousand for non-Hispanic whites.
- Of immigrants who had an abortion, 78 percent identified as generally religious, 63 percent were Christians, 44.9 percent were Catholic, and 16.3 percent were specifically "born again" or fundamentalist Christians.
- Of natives who had an abortion, 71.5 percent identified as generally religious, 66 percent were Christians, 25 percent were Catholic, and 20.3 percent were specifically "born again" or fundamentalist Christians.
- Medicaid covers abortion costs in some states and under some circumstances. Of immigrant abortions, 20.6 percent were paid for by Medicaid, compared to 20.3 percent for natives. This is not a statistically significant difference.
To calculate the abortion rate per thousand for women in their reproductive years, we follow the example of prior research and use data from the Guttmacher Institute's Abortion Patient Survey (APS) as the numerator and Census Bureau population figures as the denominator.1 Immigrants in the APS are defined as individuals not born in the United States. Figure 1 reports abortion rates by age. The figure shows that, overall, the abortion rates for all immigrants and natives ages 15 to 44 are very similar, at 18.9 and 19.5 per thousand respectively. This is not a statistically significant difference.2
Abortion Rates by Age. The abortion rate for all women 15 to 44 does not take into account the distribution of immigrants and natives across age groups. For example, immigrants generally arrive after age 25, so a smaller share than natives are in the youngest age group, which has the highest abortion rate. Only 21 percent of immigrants were ages 15 to 24 at the time of the APS, compared to 36 percent of natives. Figure 1 shows that when those under age 25 are excluded, the abortion rate for immigrants is somewhat higher than the native rate, at 15.6 vs. 14.1 per thousand. This difference, while modest, is statistically significant. Looking at other age groups in Figure 1 shows that immigrant abortion rates are also statistically significant and higher than the native rate in the 35 to 44 cohort.
Abortion Rates By Race. Table 1 reports abortion rates for demographic groups by both race and age. The table shows significant variation across racial groups, even of the same age. Among immigrants, the abortion rate for non-Hispanic whites is a good deal lower than other groups. Non-Hispanic whites tend to be slightly older than other racial/ethnic groups, but among immigrants the differences with other groups are large even for those who are the same age. This is also the case for natives. White natives have much lower abortion rates than native-born blacks, Hispanics, or Asians and the differences are statistically significant.3 The relatively high abortion rates for Hispanics and Asians may be surprising, but it must be remembered that abortion rates in Latin America and Asia tend to be higher than in the United States.4
Abortion rates have been shown to be correlated with income and lower-income women are more likely to have abortions.5 However, the situation for immigrants and natives is complex. For example, Hispanic natives have higher incomes than Hispanic immigrants, yet the abortion rates for Hispanic natives are higher than for Hispanic immigrants. Asian natives have slightly higher incomes than their foreign-born counterparts. Yet, as Table 1 shows, Asian natives have much higher abortion rates than Asian immigrants. The situation for blacks and non-Hispanic whites is somewhat different. Black and non-Hispanic white immigrants have slightly higher incomes than their U.S.-born counterparts and have somewhat lower abortion rates.6
The complex ways culture and socio-economic status impact abortion rates is well beyond the scope of this analysis. What we can say is that abortion is common among both immigrants and natives overall, with significant differences between groups. Some commentators, such as David Brooks and Michael Potemra, have argued that immigrant communities are socially conservative and continued high levels of immigration will move the country to the right on social issues.7 At least as reflected in their abortion rates, it would be difficult to argue that immigrants or their children are particularly conservative. Moreover, because there is so little difference in the abortion rates of natives and immigrants, the top of Table 1 shows the presence of immigrants in the country has no meaningful impact on the nation's overall abortion rate. However, as continued high levels of immigration increase the share of the population comprised of U.S.-born Hispanics and Asians in the long-run, the overall abortion rate in the country may be higher than it otherwise would have been.
Religious Affiliation. Table 2 reports the religious affiliation of women getting abortions from the APS. In general, a somewhat larger share of immigrant women getting abortions report that they are religious compared to natives. Of immigrants who are religious, Catholics make up a plurality of women getting abortions, while among natives, Protestants make a plurality of religious women getting abortions. These differences almost certainly reflect the larger share of immigrants who are Catholic and the larger share of natives who are Protestant.8
Abortions Subsidized by Medicaid. There is other interesting information in the APS that could be used for future research. One area worth mentioning is the role of Medicaid. About one-fifth of abortions in the United States are paid for by Medicaid. States pay for abortion if the mother's life is in danger and in the case of rape or incest. A number of states also provide Medicaid funding much more broadly for abortion. In 2008, the APS shows that 20.6 percent of abortions to immigrants were paid for by Medicaid as were 20.3 of abortion for natives. Overall, 15.5 percent of all abortions paid for by Medicaid were for immigrant women.
One might assume that immigrants' use of Medicaid for abortion would be very low, as newly arrived legal immigrants and immigrants in the country illegally are barred from using Medicaid. However, two of the top three states of immigrant settlement are California and New York, which provide Medicaid assistance for abortion. This likely explains why a significant share of immigrants getting abortions use Medicaid. Moreover, a much larger share of immigrants are poor and this, too, likely helps explain why a significant share of abortions for immigrants are paid for by Medicaid, despite legal barriers to Medicaid enrollment for some immigrants.
This analysis has shown that immigrants are about as likely as natives to have an abortion. While some commentators have suggested that immigrants are more socially conservative than native-born Americans, in terms of their abortion rates immigrant and native women are quite similar. In fact, when we control for age, immigrants are slightly more likely to have abortions than natives in some cases. The findings also show that abortion rates are high for U.S.-born Hispanics and Asians. The relatively high rates of abortion among U.S.-born Hispanics and Asians may impact their partisan political preferences.
Our primary dataset is the 2008 Abortion Patient Survey (APS). The APS was conducted by the Guttmacher Institute and the data are freely available on the organization's website.9 Guttmacher sampled 95 participating hospitals, clinics, and other places where abortions are performed, eventually collecting 9,493 responses from a representative sample of women who had abortions in 2008. Survey weights were provided to account for sampling and non-response bias, and all of our tabulations use those weights.
We calculated abortion rates by following the same basic procedure used by Rachel K. Jones and Megan L. Kavanaugh in a 2011 academic paper.10 The APS can tell us the proportion of abortions that can be ascribed to various subgroups, but the APS cannot by itself give us the number of abortions per 1,000 women in each subgroup. To calculate the abortion rates, we also need the total number of abortions in the United States in 2008 — 1,212,350 according to Guttmacher — and the number of people in the United States from each subgroup, which we obtain from the Census Bureau's American Community Survey. 11
For example, to determine the abortion rate for Hispanic immigrants ages 15 to 44, we first observe that 7.64 percent of abortion patients in the APS were Hispanic immigrants ages 15 to 44. If we then multiply 7.64 percent by 1,212,359, we get a total of 92,624 abortions for Hispanic immigrants in 2008. The ACS shows that there were 4,884,437 Hispanic immigrant women ages 15-44 living in the United States in 2008. Therefore, the abortion rate for Hispanic immigrants in 2008 was 92,624 divided by 4,884,437, which equals 0.0190, or 19.0 per 1,000 women.12
Due to data limitations, this study uses a definition of "immigrant" that is slightly different from the past practices of CIS and the Census Bureau. Normally, an immigrant is any U.S. resident who was not a U.S. citizen by birth. Any person born in the United States is a citizen, as is anyone born in outlying U.S. territories (Puerto Rico, Guam, etc.) and anyone born abroad of U.S. citizens. However, the APS question on foreign-born status does not contain that level of detail. Respondents are simply asked where they were born, without regard to parental citizenship. Therefore, any APS respondents who were born abroad of U.S. parents are included in the "immigrant" category for this study. We define immigrant (foreign-born) in the ACS in the same way to create denominators that are comparable to the APS.
1 The methodology section at the end of this report provides greater detail about how abortion rates are calculated. Also see Rachel K. Jones and Megan L. Kavanaugh, "Changes in Abortion Rates Between 2000 and 2008 and Lifetime Incidence of Abortion", Obstetrics and Gynecology, Vol. 117, No. 6 (June 2011), pp. 1,358-1,366.
2 When discussing statistical significance in this report, we use a 90 percent confidence interval.
3 The category "Asian" includes people of both East Asian and South Asian descent; Pacific Islanders are not included in this category, however they are included in the overall numbers for immigrants and natives.
4 Facts on Induced Abortions World Wide, Guttmacher Institute, January 2012; Safe and unsafe induced abortion: Global and regional levels in 2008, and trends during 1995–2000, World Health Organization.
5 The APS survey itself shows that more than two-thirds of women having abortions have incomes below 200 percent of the poverty threshold. See Rachel Jones, Lawrence Finer, and Susheela Singh, Characteristics of U.S. Abortion Patients, 2008, Guttmacher Institute, May 2010. This is more than double the low-income population's share of the women in their reproductive years based on the Current Population Survey.
6 Income figures for immigrants and natives by demographic groups can be found in this Center for Immigration Studies report, Immigrants in the United States 2010: A Profile of America's Foreign-Born Population. Rates of poverty can be found in Table 10 (p. 29), and household income can be found in Table 14 (p. 36).
7 David Brooks, "Beyond the Fence", The New York Times, May 6, 2013; Michael Potemra, "Could Mass Immigration Save Social Conservatism?" National Review, April 5, 2013.
8 Religious Landscape Survey", Pew Religion and Public Life Project, April 2015.
9 Public-use datasets available from the Guttmacher Center for Population Research Innovation and Dissemination.
10 Jones and Kavanaugh, "Changes in Abortion Rates Between 2000 and 2008 and Lifetime Incidence of Abortion".
11 Jones and Kavanaugh used the Current Population Survey instead of the ACS. We prefer the ACS because the larger sample size is useful when generating population counts for relatively small subgroups. Our calculated abortion rates are very slightly different from Jones and Kavanaugh's as a result. One difference between the APS and ACS is that the ACS alone allows respondents to choose more than one race. About 2 percent of non-Hispanic women ages 15 to 44 picked more than one race in ACS. To create comparable denominators using the ACS we allocated multi-race non-Hispanics to a single race using the approach developed by Anne Polivka at the Bureau of Labor Statistics; see "Methodologies for Maintaining Data Comparability for the Current Population Survey: One Year Into the Implementation of the 1997 Race and Ethnicity Standard", U.S. Bureau of Labor Statistics, 2003, paper presented at the American Statistical Association's Joint Statistical Meetings in San Francisco, Calif., August 5, 2003.
12 The figures in the text are rounded for readability. The actual calculations preserve all of the available precision.