Each month the Census Bureau conducts the Current Population Survey (CPS), the primary purpose of which is to collect employment data. The March CPS uses an extra-large sample of Hispanics and is considered the best source for information on persons born outside of the United States — referred to as foreign-born by the Census Bureau, though for the purposes of this report, foreign-born and immigrant are used synonymously.1 The March 1998 CPS found that 26.3 million immigrants now live in the United States, the largest number ever recorded in the nation's history, and a 33 percent increase over 1990.2 As a percentage of the population, immigrants now account for nearly one in 10 residents (9.8 percent), the highest percentage in seven decades.
Other findings in the CPS include:
- The immigrant population is growing six and one-half times faster than the native-born population — slightly over 4 percent per year compared to .6 percent per year for natives.
- The number of immigrants living in the United States has almost tripled since 1970, from 9.6 million to 26.3 million.
- As a percentage of the U.S. population, immigrants have more than doubled, from 4.8 percent in 1970 to 9.8 percent in 1998.
- By historical standards, the number of immigrants living in the United States is unprecedented. Even at the peak of the great wave of immigration in the early 20th century, the number of immigrants living in the United States was only about half what it is today (13.6 million in 1910).
- Immigration has become the determinate factor in population growth. The 8.6 million immigrants who indicated that they had arrived between 1990 and 1998 represent 42 percent of the 20.4 million increase in the total U.S. population since 1990. Additionally, 5.4 million children were born to immigrant women over the same period. Thus, immigration and births to immigrant women are equal to 70 percent of the increase in the U.S. population in the 1990s.
- Immigration has dramatically increased the supply of unskilled workers in the United States. In 1998, 31 percent of the high school dropouts in the labor market were immigrants.
- The poverty rate for immigrants is 50 percent higher than that of natives, with immigrants accounting for one in seven persons living in poverty.
- The proportion of immigrant households receiving welfare is 30 to 50 percent higher than that of natives.
The rapid growth in the immigrant population is a direct consequence of the large number of legal and illegal immigrants entering each year. Since all children born to immigrants are by definition natives, the sole reason for the increase in the immigrant population is new immigration. While some immigrants die and others return home, the issuance of 800,000 to 900,000 permanent residency visas and the settlement of more than 400,000 illegal aliens each year greatly exceeds deaths and outmigration.3 Because of the sampling error that exists in any survey, small differences between countries of origins and states reported in this backgrounder should be interpreted with caution.
There is no doubt that immigration has played an important role in American history. However, the level of immigration has varied considerably over time and, as a result, so has the size of the immigrant population. Figure 1 shows the number of immigrants living in the United States over the course of this century. The 26.3 million immigrants living in the United States in 1998 are the most ever recorded. Even during the great wave of immigration at the turn of the century, the immigrant population was about half what it is today.
Figure 1 shows that, after growing in the early part of this century, the immigrant population stabilized at around 10 or 11 million for about four decades. In the mid-1960s, changes in immigration law and other factors caused the annual level of immigration to rise steadily, from about 300,000 in the 1960s to 900,000 in the 1990s. As a result, the total immigrant population has grown dramatically. Between 1970 and 1980, the number of immigrants living in the United States grew by a record 4.5 million. Reflecting the continuing increase in the level of legal and illegal immigration in the 1980s, the immigrant population grew by 5.7 million between 1980 and 1990, another record. And, in just the first eight years of the 1990s, the number of immigrants living in the United States grew by 6.5 million, already surpassing the previous record.
The foreign-born population's growth rate since 1970 is higher than at any other time in history, far surpassing growth at the beginning of the century. Between 1900 and 1910 the immigrant population grew by 31 percent, less than the 47 percent increase in the 1970s and the 40 percent increase in the 1980s. Between 1990 and 1998, a mere eight years, the immigrant population grew by 33 percent. And if, as expected, it continues to grow at its current rate, the 2000 Census will find an increase of 40 percent since 1990.
Also, in contrast to the past, immigrants now account for a much larger share of the increase in the total U.S. population. For most of this century, the growth in the immigrant population accounted for little or none of the increase in the size of the U.S. population as a whole. Even during the first decade of this century, when immigration was an important part of population growth, the immigrant contribution to U.S. population growth was much less than it is today. The 3.2 million increase in the size of the immigrant population between 1900 and 1910 accounted for only 20 percent of the total increase in the size of U.S. population at that time. In contrast, the 6.5 million increase in the immigrant population from 1990 to 1998 accounted for 32 percent of the growth in the total U.S. population over the same period. This is because fertility rates are much lower today. As a result, around the turn of the century, the population grew regardless of immigration. Moreover, the effect of immigration on population growth is even higher — 42 percent — if the number of immigrants who responded in the CPS that they had arrived in the 1990s is used to measure the impact of immigration on the growth of the total U.S. population.4
Table 1 shows the percentage of the U.S. population comprised of immigrants for each decade in this century. While the number of immigrants and the growth rate of the immigrant population are both higher now than at any other time in the 20th century, the immigrant percentage of the population was larger in the first few decades of this century. In 1910, the immigrant population reached an all-time high of 14.7 percent of the total U.S. population, compared to 9.8 percent today. As a result of World War I and immigration law changes in the early 1920s, the level of immigration began to fall, as did the foreign-bon percentage of the population. The 1930 Census was the last time the percentage of immigrants was as high as it is today.
In terms of the impact of immigrants on the United States, both the percentage of the population made up of immigrants and the number of immigrants are clearly important. The ability to assimilate and incorporate immigrants is partly dependent on the relative sizes of the native and immigrant populations. Still, 26 million immigrants are likely to have an enormous effect on the socio-economic life of the United States regardless of whether this represents 10 or 15 percent of the nation's population, especially since the immigrant population is concentrated in only a few states.
Table 2 ranks the states by the size of their immigrant populations. It also shows the number of immigrants who reported arriving in the 1990s. California clearly has the largest immigrant population; New York, the state with next largest number of immigrants, has fewer than half as many. Table 2 also shows how concentrated the immigrant population is: Only a few states represent the vast majority of the foreign-born population. The nearly eight million immigrants in California account for 30.3 percent of the nation's total immigrant population, followed by New York (13.8 percent), Florida (8.8 percent), Texas (8.8 percent), Illinois (4.5 percent), and New Jersey (4.5 percent). Despite having only 39 percent of the nation's total population, these six states account for 71 percent of the immigrant population.
Table 3 ranks states by the percentage of their populations composed of immigrants. Reflecting the high concentration of the foreign-born population found in Table 2, the proportion of immigrants exceeds the national average in only 12 states. While the rankings by percent immigrant are similar to those in Table 2, there are some significant differences. Because of their relatively small populations, several states with high percentages of immigrants rank lower in terms of number of immigrants. Hawaii, for example, is 21st in number of immigrants, but third in terms of percentage and Rhode Island is 28th in number of immigrants, but is eighth in terms of percentage.
Table 4 compares the 1990 census with the March 1998 CPS and ranks the top 12 states by the numerical increase in their immigrant populations. While the states that had large immigrant populations in 1990 continue to account for most of the growth in the immigrant population, Table 4 shows substantial growth in the immigrant populations of Arizona, Oregon, Maryland, and North Carolina.5
Region and Country of Origin
Table 5 shows the distribution of immigrants by region, with Mexico and Canada treated separately. The CPS found that 7.1 million of the immigrants in the United States are from Mexico. Even though it is only one country, the Mexican share of the immigrant population (27.1 percent) is larger than the number of immigrants from any other part of the world. Immigrants from Latin American, the Caribbean, and Asia make up the majority of the immigrant population, with 77.5 percent of the foreign-born coming from these areas. Africa and Europe make up a relatively small portion of the immigrant population, accounting for only one in five immigrants. Table 6 ranks the top 15 immigrant sending countries by the number of post-1970 immigrants living in the United States as of March 1998. Mexico is, of course, the largest sending country, accounting for more than five times as many immigrants as the combined total for China, Taiwan, and Hong Kong. As is clear from Tables 5 and 6, Latin America, the Caribbean, and Asia dominate the list of immigrant-sending countries, accounting for 12 of the top 15.
Labor Market Characteristics
In March 1998, immigrants comprised 12.2 percent of the nation's total work force. This is somewhat higher than the 9.8 percent of the total U.S. population because, in comparison to natives, a slightly higher percentage of immigrants are of working age. Table 7 reports the educational attainment and other characteristics of immigrants and natives in the work force. In 1998, 31 percent of immigrants did not have a high school degree, and of those who arrived in the 1990s, 35 percent lacked a high school degree. In comparison, only 9 percent of natives in the work force lacked a high school education, making immigrants more than three times as likely as natives to be dropouts. At the highest level of education, immigrants tend to be slightly more educated than natives, with 10 percent of immigrants holding a graduate or professional degree compared to 8 percent of natives.
The large number of immigrants with low levels of education means that immigration policy has dramatically increased the supply of workers with less than a high school degree, while increasing other educational categories more moderately. The last column in Table 7 shows the portion of each educational category composed of immigrants. Although about 12 percent of the total work force is immigrant, they comprise 31 percent of the high school dropouts in the work force. This means that any effects on the wages or job opportunities of natives will disproportionately affect less-skilled workers.
Table 7 also provides information on the earnings, unemployment rate, and age of immigrants in the work force. Given the large proportion of immigrants with few years of schooling, it is not surprising that immigrants are significantly poorer as a group than are natives. Table 7 shows that the median earnings of immigrants are only about 75 percent that of natives, and for the most recent immigrants, median earnings are only about half that of natives. While as a group immigrants earn significantly less than natives, the earnings data by entering cohort suggest significant progress over time.
Since the cohort data is only based on one point in time, March 1998, it is possible that the seeming economic progress of immigrants is at least partly caused by the departure of those immigrants who did not fare well in the U.S. labor market. Moreover, the age data in Table 7 indicate that immigrants who entered prior to 1980 are, on average, older than natives in the work force. Because greater work force experience comes with age, one would expect this to translate into higher earnings. Despite this, the median earnings of immigrants who arrived in the 1970s are actually below that of natives, even though they are on average older than natives. Only the cohort that arrived before 1970 had higher earnings than natives. It should be noted, however, that the average age of this group is much older than that of natives. In addition to their age, the higher earnings of pre-1970s immigrants may also be explained by the fact that most were admitted under the pre-1965 immigration system, which tended to produce a more educated flow of immigrants relative to natives than today's policies.
In addition to education level, earnings, and age, the distribution of immigrants across occupations can also provide important information about the characteristics of immigrants in the labor market as well as their likely impact on native workers. Table 8 shows the occupational concentration of immigrants and information on each occupation. The upper half lists those occupations in which the immigrant component is less than or equal to their proportion in the overall work force. The lower half lists those occupations in which immigrants comprise a proportion larger than their representation in the work force (henceforth referred to as high-immigrant occupations). While the occupations are highly aggregated, the table does suggest a clear pattern to immigrant employment. Given the low level of educational attainment of a large proportion of immigrants, it is not surprising that high-immigrant occupations are those that tend to require fewer years of education. For example, while immigrants make up only 9 percent of individuals in managerial and professional jobs, they comprise 19 percent of those holding non-private household service jobs, such as janitor, security guard, and child care worker.
Table 8 reveals that only 26 percent of natives are employed in occupations that have high concentrations of immigrants. This suggests that for most natives the effect of immigrant competition will be relatively modest.6 However, as discussed below, this is not necessarily as desirable a situation as it may first appear.
Commensurate with their low skill requirements, high-immigrant occupations pay an average of only 60 percent of what low-immigrant occupations pay. Additionally, high-immigrant occupations have a much higher unemployment rate than low-immigrant occupations, 8.7 percent compared to 3.1 percent. This does not necessarily mean that immigrants have lowered the wages or increased unemployment in these occupations. What it is does mean, however, is that immigration's effect on the jobs and wages of natives will likely fall on the 25 million native-born workers who already have the lowest wages and the highest unemployment.
The much higher unemployment rate in high-immigrant occupations is especially important because it strongly suggests that demand for labor in these occupations is weaker than in low-immigrant occupations. In contrast, the tight labor market in low-immigrant occupations is likely to make it much easier to absorb immigrant workers into these occupations without affecting the opportunities available to natives. This also means that studies that attempt to estimate the wage effects of immigration by assuming that each additional immigrant will have the same impact, regardless of the occupation he is employed in, may fail to capture the varying conditions prevailing in different segments of the labor market.
Another distinguishing characteristic of high-immigrant occupations is the large percentage of blacks they employ: 42 percent of blacks work in high-immigrant occupations compared to only 22 percent of whites. The much larger share of the black population employed in these occupations means that African-Americans are more likely to be negatively affected by immigration than are whites.
In March of 1998, 20 percent of immigrants lived in poverty, half-again more than the 12.7 percent of natives, and the poverty rate is 29 percent for immigrants who entered in the 1990s, almost two and a half times that of natives. While the most recent immigrants have very high poverty rates, year-of-entry data suggest economic progress over time. For immigrants who entered prior to 1970, the poverty rate is 11 percent; for 1970s immigrants it is 13.1 percent and for 1980s immigrants it is 19.7 percent. As is the case with the wage data in Table 7, it is unclear to what extent this progress reflects the out-migration of unsuccessful immigrants, the different selection criteria used in the past, and the changing origins of immigrants over the last few decades. This last point is especially important because immigrants from different countries have very different poverty rates, even after they have been in the United States for several decades. For example, 21 percent of Mexican immigrants who arrived prior to 1970 still live in poverty, but fewer than 8 percent of pre-1970 European immigrants live in poverty. Since the sending countries have shifted in the last 30 years, it is possible that poverty among today's immigrants will not shrink as it did among their predecessors.
Table 9 reports poverty rates for persons from the top 15 post-1970 immigrant sending countries. The data indicate that there are enormous variations in poverty rates among immigrants from different countries. For example, the 38 percent poverty rate for Dominicans is more than 6 times that of persons from India. Table 9 also shows the proportion of persons 21 years of age and older from each country who have not completed high school. This educational data indicates that a good deal of the variation in poverty rates is explained by educational differences.
The high incidence of poverty among immigrants as a group has significantly increased the overall size of the population living in poverty. In 1998, immigrants accounted for 15 percent of those living in poverty. While this is a large percentage, it would be even larger if immigrants' native-born children (17 and under), who are included in the poverty figures for natives, were counted with their parents. Since a child's standard of living is a function of his parents' income, it is not unreasonable to view poverty among the native-born children of immigrants as attributable to their immigrant parents. The lower portion of Table 9 shows poverty rates for immigrants and natives by age and mother's nativity for children 17 and under. (To avoid double counting, only the mother's nativity is used.) Table 9 also breaks down poverty rates for immigrants and natives if the U.S.-born children of immigrants are counted with immigrants.
Of the 31 million natives living in poverty, 2.5 million (8 percent) are the U.S.-born children of immigrant mothers. If the native-born children of immigrants are excluded, poverty among natives drops from 12.7 percent to 12.1 percent. And if the 5.3 million immigrants in poverty are excluded, along with their U.S.-born children, from the nation's overall poverty counts, the overall number of people living in poverty drops by 7.8 million. This means that immigrants and their U.S.-born children account for 22 percent of the 36.2 million people living in poverty in the United States and 24 percent of the children in poverty.
Table 10 shows the percentage of immigrant- and native-headed households in which at least one member of the household receives public assistance (including TANF and general assistance programs), Supplemental Security Income (SSI), and Food Stamps. It indicates that, in general, welfare use by immigrant-headed households is higher than that of native-headed households. This is true for all three programs and for all entering cohorts after 1970 with the exception of SSI use by 1990s immigrants. The higher welfare use rates by immigrant households is at least partly explained by the large proportion of immigrants with few years of schooling: Because less-educated persons tend to have lower incomes and higher unemployment, it is not surprising that immigrant use of welfare programs is between 30 and 50 percent higher than that of natives.
It should also be clear from Table 10 that most households, immigrant or native, do not use means-tested programs. On the other hand, even though a relatively small portion of the population uses welfare, the total cost of just the programs listed in Table 10 is over $80 billion a year. Moreover, there are a number of additional means-tested programs not listed that are linked to those in Table 10. In the March 1998 CPS, for example, 14.9 percent of immigrant households reported that at least one child in the household used the subsidized school lunch program, compared to only 6.5 percent of native households. Finally, there is the question of whether native use of welfare is the proper yardstick by which to measure immigrants. Some may reasonably argue that since immigration is supposed to benefit the United States, our admission criteria should, with the exception of refugees, select only those immigrants who are self-sufficient.
While on the whole immigrant households have higher welfare use rates, this is not true for immigrants from all countries. Table 11 shows welfare use rates for the top 15 post-1970 immigrant sending countries. As is the case with poverty rates, immigrants from those countries with higher education levels tend to have lower welfare use rates. From the list of countries in Table 11, it is also clear that refugees tend to use more welfare than other immigrants. Russian and Vietnamese immigrants, most of whom were admitted as refugees, have very high welfare use rates. On the other hand, Mexican and Dominican households have welfare use rates that are as high or higher than Russian or Vietnamese immigrants, and virtually none of these immigrants are refugees. Thus, higher use of welfare by immigrants is not caused simply by immigrants admitted for humanitarian reasons.
Recently, many school districts have been faced with dramatic increases in enrollment without corresponding increases in revenue. Absent from much of the public debate over this situation has been the role of immigration policy. Table 12 shows the proportion of the school-age population (ages 5 to 17) with immigrant mothers. While only one in five of these children are immigrants themselves, the use of public education by the native-born children of immigrants is a direct consequence of their parents having been allowed into the country.
Table 12 indicates that immigration has had a significant effect on the size of the school-age population. In March 1998, almost 15.9 percent of children ages 5 to 17 had an immigrant mother. This situation results from the higher proportion of immigrant women in their childbearing years as well as their high fertility rates. The effect of immigration on public schools will be even larger in the coming years because 17.9 percent of children approaching school age have immigrant mothers.
Table 12 also shows the effect of immigration on public education in the top seven immigrant-receiving states. Immigration policy has had an enormous effect on the size of the school-age population in these states. Of course, a dramatic increase in enrollment may not create a problem for public education if tax revenue increases proportionately. But, as we have seen, immigrants generally have lower earnings than natives so the tax contributions of immigrants are unlikely to entirely offset the costs imposed on schools. This is especially true due to the higher costs associated with teaching children whose first language is not English.
Clearly, the absorption capacity of American public education is an important issue that needs to be taken into account when formulating a sensible immigration policy. Table 12 suggests that the failure to consider this question may have significant consequences for public schools.
While immigration's impact on the United States continues to be a subject of intense national debate, there can be no doubt that the large number of immigrants now living here represents an enormous challenge. No nation has ever attempted to incorporate over 26 million newcomers into its society. Moreover, without a change in immigration policy, immigration's impact will only grow as the size of the immigrant population continues to increase rapidly.
1 While generally considered the best source of information on immigrants, the CPS is thought to undercount the foreign-born by perhaps 5 or 10 percent. In the past, the Census Bureau has also had to revise the numbers in the CPS after problems were discovered with the sampling weights. Thus, it is possible that the numbers reported in this Backgrounder will have to be slightly adjusted if problems are found in the March 1998 survey.
2 All persons not born in the United States, one of its outlying territories or of U.S. parents living abroad are considered immigrants. All persons born in the United States, including the children of illegal aliens, are considered natives. The immigrant population in the CPS includes some illegal aliens and a small number of persons on temporary visas, such as students, which allow them to stay for multiple years in the United States.
4 The CPS asks immigrants when they came to the United States. In the March 1998 CPS, 8.6 million immigrants responded they had arrived between 1990 and 1998. However, the immigrant population also shrinks as a result of deaths and outmigration. Thus, even though 8.6 million immigrants arrived in the 1990s according the year-of-entry question, the immigrant population grew by 6.5 million because slightly over 2 million of the immigrants here in 1990 had either died or gone home by 1998. It is not unreasonable to view the 8.6 million immigrants who arrived between 1990 and 1998 as accounting for 42 percent of the 20.4 million person increase in the total U.S. population.
5 Given the small sample size of immigrants in these states, the large numerical and percentage increases in these states should be interpreted with caution. The estimates in Table 4 should be used to make determinations of the relative differences between states, and not as quantified absolute differences.
6 There are a number of specific occupations that are included in the upper portion of Table 8 that do have a large concentration of immigrants, such as doctor. Because of the sample size, however, it is not possible to break out these occupations separately even though the effect of immigrants on these specific occupations is likely to be significant.