Birth Rates Among Immigrants in America
Analysis of data collected by Census Bureau in 2002 shows that women from the top-10 immigrant- sending countries living in the United States collectively tend to have higher fertility than women in their home countries. As a group, immigrants from these countries have 23 percent more children than women in their home countries, adding to world population growth. Among the findings:
As the nation's immigrant (foreign-born) population has grown in recent years, a good deal of research has been devoted to their socio-demographic characteristics. However, less attention has been paid to their fertility. Children born to immigrants are probably the most important long-term effect of immigration. This study explores immigrant birth rates with particular focus on how they differ from women in their home countries. Studying immigrant fertility is necessary in order to understand immigration's impact on U.S. and world population growth as well as its effect on public services provided to children. In addition, fertility can be seen as a measure of immigrant integration. If people are choosing to have more children, this may indicate that they feel relatively optimistic about the future. Only recently has data become available to study immigrant fertility in any detail.
Data and Methodology
Calculating Fertility Rates. The ACS is designed to replace the Census long form and includes as many questions, including the respondent's age, race, sex, and whether a person is an immigrant and the country where each person was born. In addition, the survey also asks women if they have given birth in the past year. This information can be used to calculate what demographers call a Total Fertility Rate (TFR) for the entire population or for a segment of the population. A TFR is one of the most common measures of fertility used by demographers. It represents the average number of children a woman will have in her lifetime once she has passed through her reproductive years based on current trends.2 We calculate TFRs for women ages 15 to 49. While there are a tiny number of births to women younger than 15 and older than 49, the ACS does not capture births to women outside of this range. It should be noted that TFR is designed to control for differences in age structure between groups; thus if immigrants to the United States are much younger or older than is the overall population of their home country it should not unduly affect the findings of this study.
Alternative Calculations for TFR. While we use the ACS to calculate immigrant fertility, there is another way to calculate it. The National Center for Health Statistics (NCHS) collects administrative data from birth certificates, and when this data is combined with Census Bureau estimates of the total population, it can be used to calculate TFRs. This is done by using the NCHS's administrative data to report the number of births to women in each age group and then dividing it by the total number of women in that age group based on Census Bureau data, such as the ACS. When dealing with immigrants, however, there are some disadvantages to this approach. First the NCHS defines the foreign born somewhat differently than the Census Bureau. Moreover, NCHS public use data only specifically identify three countries (Mexico, Canada, and Cuba), while all other immigrant mothers are simply reported as being foreign-born. Third, the ACS, like all Census Bureau data, misses some fraction of the population; in contrast, the undercount in the NCHS birth data is supposed to be very small.3 Thus, combining administrative data with survey data like the ACS may tend to slightly overstate immigrant fertility because the undercount in the ACS makes for a denominator that may be too small.
Overall the ACS shows a TFR of 2.78 for all immigrant women ages 15 to 49. If we use the ACS in combination with the NCHS birth data the TFR would be 2.91, hardly a huge difference.4 Of course, it may be that the NCHS data combined with the ACS actually produces a more accurate TFR for immigrants. But it must be remembered that this approach uses two somewhat different data sources. Calculating a TFR using just the ACS on the other hand avoids this problem because if there is an undercount of the foreign born in the ACS, there is also an undercount of foreign-born women who have given birth recently. It's worth noting that most government publications, but not all, calculate the nation's TFR by using NCHS data combined with Census Bureau data and not from just one survey like the ACS. For the purposes of this study, the key point is that the differences between estimates of immigrant TFR using either approach are small.5
Fertility in the Home Countries. In order to compare immigrant fertility to that of their home countries, we use the latest figures for TFRs from the United Nations Population Division, Department of Economic and Social Affairs. The UN provides TFR for all member states for the time period 2000 through 2005. We compare home country fertility with the immigrants for the top-10 countries. We then weight the data to reflect the share that immigrants from each of the countries accounted for in the United States. Although the ACS is a very large survey, it is still not large enough to calculate fertility for immigrants from countries other than the top ten because the sample size for countries that send fewer immigrants is quite small. Moreover, only a small fraction of women have children in any year. By confining our analysis to the top countries, we are able to obtain more statistically robust estimates of immigrant fertility. It should be pointed out that the top-10 countries account for 58 percent of all immigrants in the United States. Moreover, the total fertility for immigrants from the top countries in 2002 was 2.86, which is very similar to 2.78 for immigrants overall.
Difference Between Immigrants and Their Countries. When comparing immigrant fertility to that of their home countries it is important to understand that immigrants can differ in important ways from the general population of the countries from which they come. It is probably not possible to control for all the ways in which immigrants to the United States may be different from the general population back home. While many factors impact fertility rates, it is a well-established principle in demography that throughout the world education levels are a key determinant of fertility, with more educated women having fewer children on average than women with less education. The ACS can be used to estimate the education level of immigrant women in the United States for the top sending countries. We can then compare this to fertility rates by education level in their home countries. Unfortunately, data for fertility by education levels for foreign countries are not always as complete or detailed as we would like.6 Despite these limitations, the available data can be used to adjust for the education level of immigrants in the United States in most cases. Fertility figures by education for the home countries come mainly from two sources: The Demographic and Health Surveys provided by Macro International, Inc., and the Information & Knowledge for Optimal Health Project at Johns Hopkins Bloomberg School of Public Health.7
Estimating Births to Illegal Aliens. Like all Census Bureau surveys, the ACS includes individuals in the country illegally. In fact, most researchers think about 90 percent of illegal aliens are included in Census surveys. Therefore it is possible to estimate fertility for illegals, or at least those illegals who respond to the ACS. Like almost all researchers in this field, we use the characteristics of individuals in the ACS to estimate the number of illegal immigrants. Based on the citizenship status, year of arrival in the United States, age, country of birth, educational attainment, sex, receipt of welfare programs, receipt of Social Security, veteran status, and marital status reported in the ACS, we assign probabilities to each survey respondent.8
This method is based on some very well-established facts about the characteristics of the illegal population. For example, it is well known that illegals are disproportionately young (under age 40), male, unmarried, and have few years of schooling, etc. We estimate that there were a total of 8.45 million illegal aliens in the 2002 ACS. We further estimate that of illegal alien women ages 15 to 49, 58 percent are Mexican, and 21 percent are from other Latin America countries.
Individually immigrants from seven of the top-10 countries have a higher TFR in the United States than in their home countries. For those countries in which the immigrants have higher fertility, the biggest differences are for women from Mexico and China. Mexico is especially important because immigrants from that country account for more than 40 percent of all births to immigrants and 54 percent of births among the top-10 immigrant-sending countries in Table 1. Thus immigrants from that country exert a very significant impact on the results. While immigrants from the top-sending countries tend to have more children than women in their home counties, this is not true in every case; the Philippines, India, and Viet Nam are the exceptions.
Immigrants Representative of Home Countries? Table 1 compares immigrants to those in their home countries. As discussed in the Data and Methodology section of this report, immigrants to the United States can differ in important ways from people in their home countries. Throughout the world, a key determinant of how many children a woman will bear in her lifetime is her education level. In almost every country in the world, more educated women have significantly fewer children than women with less education. If immigrants to the United States have a different educational profile than the general population of their home countries, then it can have a significant impact on fertility. Table 2 attempts to adjust for at least some of these differences.
To understand why adjusting for education level is necessary, consider the case of India. Although overall fertility in that country is 3.1, a 1998-99 survey found fertility in that country varies from 3.5 children for women with no education to 2.0 for women who have completed secondary school. The 2002 ACS shows that 90 percent of female Indian immigrants in the United States have completed secondary school. If Indian women in the United States had the same fertility rate as women in India with the same education level, then their fertility rate would have been 2.19. This is very similar to the 2.23 children the ACS shows is its actual fertility. Thus all of the difference between the fertility of women in India and Indian immigrants in the United States found in Table 1 disappears when we account for the education level of immigrants from that country.
Overall Table 2 shows that the difference between home countries and their immigrants in the United States is actually larger when we control for education. The table shows that immigrants from the top-10 sending countries should have 2.15 children on average, which is 33 percent less than their actual fertility. In the cases of India and Vietnam, where the straight comparison in Table 1 seemed to show that immigrants in the United States have lower fertility than women in their home countries, the fertility in the United States actually is about the same as back home once we control for education levels. For China and Mexico, the difference actually widens with the home country. Only in the case of the Philippines do immigrants seem to have fewer children in the United States and the difference does not seem to narrow even when we control for the mother's education level. The Philippines might be a special case, because such a large share of Filipino women are married to American servicemen and have high marriage rates to native-born Americans in general. Therefore, their child bearing may reflect the preferences for small families common among Americans. In any event, Table 2 shows that, at least when we account for education levels, immigrants from the top-sending countries tend to have significantly more children than women in the countries from which they come. Of course, comparisons of this kind are by no means perfect, but they at least allow us to account for some of the differences between immigrants in the United States and the countries they come from.
Estimated Fertility of Illegal Aliens. We estimated that the fertility of illegal aliens in 2002 was 3.06 children on average, or about half again as high as the fertility rate for natives. This compared to an estimated 2.61 for legal immigrants. In a study released in July of this year, we used birth certificate records and estimated that in total there were 380,000 births to illegal aliens in 2002, accounting for nearly one out of every 10 births in the United States. The high fertility rate of illegal aliens seems to be due primarily to factors other than their legal status, such as culture and educational attainment. It must be remembered that some 80 percent of the illegal population is from Latin America, and more than 60 percent lack a high school degree. Because the fertility of illegals is very similar to Hispanic immigrants in general, it suggests that illegals are not purposefully having children, who are automatically awarded US citizenship, in order to remain in the United States.
The large number of births to illegals is important for a number of reasons. Perhaps most importantly, it shows that the longer illegal immigration is allowed to persist, the harder the problem is to solve. As U.S. citizens, these children can remain permanently, and their citizenship can effectively prevent a parent's deportation. Moreover, once adults they can sponsor their parents for permanent residence. Births to illegals also have significant bearing on U.S. population growth. If illegal aliens are allowed to remain in the country, either as illegal aliens or as legal residents, their births alone will add some four million people to the U.S. population over the next decade.
Fertility by Education Level. As already discussed, education is one of the key determinants of fertility. One reason fertility rates have declined so much around the world in recent decades is that the education level for women has increased significantly in almost every country. Table 3 reports the fertility for immigrant and native women based on their education level. It confirms the basic fact that fertility tends to vary with education levels. But it also shows that, relative to natives, the differences between well-educated and less-educated immigrants are very large. For example, at 3.3 and 3.4 children, the fertility of immigrants without a high school degree or only a high school degree is three-fourths higher than the 2.0 and 1.9 rates for immigrants with some college or college graduates. In contrast, the 2.3 rate for natives without a high school or with only a high degree is not that different than the 1.8 rate for more educated natives.
The higher fertility of less-educated immigrants means that a much larger share of births to immigrants are to women with little formal education than is the case for natives. This is important because education is the best predictor of income, poverty, use of means-tested programs, and a host of other measures of social well being. For the children of immigrants it means that a very large share may grow up poor. These findings also have implications for likely educational attainment of the children of immigrants. Those born to less educated parents are themselves at higher risk for dropping out of high school. In addition to the cost to taxpayers, the results in Table 3 indicate that many children from immigrant families may have significant difficulty in reaching economic and social parity with the children of natives. Many researchers simply examine the education level of immigrants and natives. But examinations of this kind might be misleading when thinking about future generations because the number of children born to immigrants is not proportionate to the size of each educational group. When thinking about the second generation, it is important to realize that the children of less educated immigrants will comprise a large share of births, a share that is significantly larger than would be expected if one simply assumes that all immigrants have the same fertility.
Immigrants Do Not Account for High U.S. Fertility. In a study in April of this year, we found that immigrants had little impact on the nation's fertility.9 Using a different data source, we found that without immigrants the national fertility rate still would be two children on average. The 2002 ACS study shows the same basic results. If all the immigrants are removed from the data and the fertility rate recalculated, the overall fertility rates in the United States would still be about two children, or 2.05 children. It is true that America does have a higher fertility rate than other advanced industrialized democracies -- 1.4 for Europe or 1.3 for Japan. But that higher rate is due almost entirely to native-born American women. For whatever reason, they have significantly more children on average than women in other western countries. It must be remembered that nearly eight out of 10 births in the United States are to native mothers, thus it is their characteristics that will primarily determine the overall fertility rate in the United States.
Immigration Accounts for Most U.S. Population Growth. Although immigration has little effect on the nation's overall fertility rate, new immigrants (legal and illegal) plus births to immigrants add some 2.3 million people to the United States each year, accounting for most of the nation's population increase. In fact, because natives have only two children on average, absent the additions that come from immigrants, the U.S. population would be roughly stable in the long-run without continued high levels of immigration. Whatever one thinks about the costs and benefits of continued population growth, there is no question that immigration is the driving force behind it.
We also estimate that illegal-alien women had slightly less than 3.1 children on average in 2002, or about 50 percent higher than the two children natives have on average. The birth rate for legal immigrants is 2.6, or about one-third higher than that of natives. We have previously estimated from birth records that there were 380,000 births to illegal aliens in 2002, accounting for nearly one out of 10 of all births in the United States. The high fertility rate of illegal aliens seems to be due primarily to factors other than their legal status, such as culture and educational attainment.
The children born to immigrants are arguably the most important long-term legacy of immigration. The decision to have children is one of the most important any woman makes in her lifetime. The fact that immigrants tend to have more children once they come to the United States than do their counterparts in their home countries is an important finding. While the demographic data on which this study is based are clear, the reason for the finding is not. Perhaps it results from a feeling of optimism about the future immigrants experience after coming to America; perhaps it is the assistance the government offers to low-income women with children; or maybe it has something to do with the way immigrants differ from the populations of their home countries. These and other possible explanations are all areas in need of further research.
2. Demographers Arthur Haupt and Thomas Kane have defined TFR as, "the average number of children that would be born alive to a woman (or group of women) during her lifetime if she were to pass through all her childbearing years conforming to the age-specific fertility rates of the given year." All demographics textbooks provide a discussion of TFR; see for example David Yaukey and Douglas L. Anderton, "Demography: The Study of Human Population," 2001. Waveland Press, pp. 193-194.
3. The NCHS reports
that more than 99 percent of births in the United States are recorded. See
"National Vital Statistics, Births: Final Data 2002," Vol. 52 number 10, page 3.
4. It should be noted that the TFR of 2.91 is based on ACS and NCHS data and includes persons born aboard of American parents. This is true both for the NCHS birth data and the ACS population totals which are used as a denominator. Although the children of American citizens born in other countries are technically not immigrants and are not considered foreign born by the Census Bureau, they are counted as such in NCHS data.
5. It should be noted that using the 2002 June CPS, the TFR for foreign-born women was 2.4. This is quite a bit larger than the 2.8 TFR shown in the ACS and 2.9 using NCHS data. Sampling variability accounts for most of this difference; using a 95 confidence interval, the ACS and CPS are barely statistically different. Moreover, the question wording of the two surveys is not the same. Finally we use the ACS to measure fertility up to age 49, while the CPS does not question older women about their recent child bearing experience. One of the big differences between the two surveys is that the ACS reports some 860,00 births to immigrants while the CPS records only 630,000. It is not clear why there is the large difference, but the NCHS data, which is based on actual birth records from hospitals, shows 900,000 births to foreign born mothers (using the NCHS definition of foreign-born) in 2002. Thus the ACS seems to produce estimates that more closely match births figures from administrative data than does the CPS. This gives us more confidences in the ACS results.
6. In some cases the available data are relatively old and in other cases the education categorizes are highly aggregated.
7. Figures for fertility by education from the Demographic and Health Surveys can be found at http://www.measuredhs.com/contact.cfm . Fertility data by education level from the Johns Hopkins School of Public Health can be found at www.infoforhealth.org , see Table B. For China we use a 2000 paper by Cao Gui-Ying, which can be found at www.iiasa.ac.at/Publications/Documents/IR-00-026.pdf . For Mexico we use the data from the National Population Council of Mexico, which can be found at www.conapo.gob.mx . In the case of Mexico, China, India, and El Salvador the data for home country fertility by education is older than we would like, so we adjusted it to reflect the UN's 2000-05 fertility estimates assuming that fertility changed between the time of the survey and 2002 for each educational group in the same proportion as for the overall fertility rate. Because fertility is already so low for Cuba, Korea, Canada, and the United Kingdom, we did not adjust for the education level of those immigrants in the United States. If we had done so it would have probably increased the difference with home countries in each case because immigrants in the United States tend to be more educated than the general population of the home country. But any effect would be very small. One limitation of the data for the six countries that we did adjust is that in some cases the educational categories are quite aggregated. For example, for the Demographic and Health Surveys there are only three categorizes: 1) those with no education; 2) those with a primary education; 3) those with a secondary or higher education. Many immigrants to the U.S. are very educated, and this fact is lost because of the way the surveys group more educated women. If we had been able to find figures for fertility by education level in the home countries for more highly educated women it almost certainly would have further increased the gap with the home countries found in tables 1 and 2.
8. Those individuals who have a cumulative probability of 1 or higher are assumed to be illegal aliens. By design, the probabilities are assigned so that both the total number of illegal aliens and the characteristics of the illegal population closely match other research in the field, particularly the estimates developed by Jeffery Passel, formerly of the Urban Institute, now at the Pew Hispanic Center.