[An updated version of this article can be found at Immigration in the 2nd edition.]
Immigration is again a major component of demographic change in the United States. Since 1940 the number of legal immigrants has increased at a rate of 1 million per decade. By the eighties about 600,000 legal immigrants were being admitted each year, making for a rate of about 6 million per decade (see table 1). Large numbers of illegal aliens also enter the country. In 1986, for instance, the Border Patrol apprehended 1.8 million persons attempting to enter illegally, or more than three people per minute.
In the early 1900s, when immigration reached historically high levels, half of the growth in total U.S. population and in the labor force was due to immigration. In the seventies only about one-quarter of the growth in population and one-eighth of the growth in the labor force were due to immigration. In the eighties immigration was back up and accounted for just under 40 percent of the growth in population and for a quarter of all new labor market entrants. These proportions are approaching the earlier ones, not only because of the large number of immigrants, but also because of the declining fertility rate of American women.
Just as numbers of immigrants change, so has the means of selection. Between 1924 and 1965 immigrants were selected mainly on the basis of national origin. The United Kingdom and Germany received over 60 percent of the visas allocated outside the Western Hemisphere. (Visa applicants originating in North or South America were not subject to these quotas.) That all changed with the 1965 amendments to the Immigration and Nationality Act. Under the new system most visas are reserved for relatives of U.S. residents. By the late eighties 75 percent of immigrants were admitted because of family ties, and an additional 20 percent were refugees.
The 1965 amendments dramatically altered the mix of immigrants. In the fifties 53 percent of immigrants originated in Europe, 25 percent in Latin America, and 6 percent in Asia. By the eighties only 11 percent of immigrants originated in Europe, 42 percent in Latin America, and 42 percent in Asia.
Two major changes in immigration policy were enacted recently. First, the 1986 Immigration Reform and Control Act granted amnesty to 3 million illegal aliens and introduced penalties for employers who hire undocumented workers. Although the act's purpose was to stem the illegal flow, its effectiveness is already in doubt. The Border Patrol apprehended 1.2 million illegal aliens in 1990, the same number it apprehended in 1983. The second piece of legislation was the 1990 Immigration Act, which permits the entry of an additional 175,000 immigrants per year, with half of the extra visas reserved for skilled applicants.
Because of the increasing importance of immigration, the nineties will likely witness the renewal of the debate over the "immigration problem." This debate will be guided by three issues:
1. How well do immigrants adapt to the United States?
2. What is their impact on the labor market opportunities of natives?
3. What immigration policy is most beneficial for the country?
Immigrant Performance in the Labor Market
When immigrants enter the United States, they typically lack skills, such as proficiency in the English language, that American employers value. Hence, it is not surprising that new immigrants earn less than native workers. As immigrants acquire these skills, however, their economic status catches up to that of natives. But because the recent waves of immigrants are relatively less skilled than earlier ones, the wage disadvantage of newly arrived immigrants has worsened over time. Immigrants who arrived in the late fifties earned 12 percent less than natives at the time of arrival. This wage disadvantage upon arrival increased to 15 percent in the late sixties, to 25 percent in the late seventies, and to 27 percent in the late eighties. Because recent immigrants start so far behind, they cannot attain wage parity with natives even after two or three decades in the United States (see chart 1).
Chart 1. Earnings over the Working Life for Immigrant and Native Men
SOURCE: Adapted from George J. Borjas, Friends or Strangers: The Impact of Immigrants on the U.S. Economy, 1990: 114.
Enlarge in new window
In 1980 newly arrived immigrants from India or Iran earned 20 percent less than natives; newly arrived Mexican or Haitian immigrants earned 50 percent less. Compare this to immigrants from Sweden or the United Kingdom, who earned about 20 percent more than natives.
The large disparity in earnings for various nationalities arises partly because skills acquired in advanced, industrialized economies (like the United Kingdom or Sweden) are more easily transferable to the American labor market. But another reason is that the typical worker who emigrates from, say, Sweden, differs substantially from the typical worker who leaves Mexico. The Swedish government taxes skilled workers heavily and subsidizes the unskilled. Hence Sweden suffers from what is called a brain drain: skilled Swedes migrate to the United States. In contrast, there is a great deal of income inequality in Mexico, where unskilled workers have few economic opportunities and skilled workers are well rewarded. Therefore, it is the unskilled who wish to emigrate. Because incomes are highly unequal in many of the less developed countries that are the source of U.S. immigrants today, current waves of immigrants are less skilled than earlier waves.
An important consequence of the shift toward a less skilled immigrant flow is a sizable increase in the costs associated with welfare use among immigrants. Between 1970 and 1980 the fraction of immigrant households on welfare rose by 3 percentage points (from 5.9 to 8.8 percent). Welfare recipiency among some national origin groups is disturbingly large. Some 11 percent of Filipino, 18 percent of Cuban, and 26 percent of households originating in the Dominican Republic receive public assistance, as compared to 7.9 percent of native households.
The Impact of Immigrants on Native Earnings
There are two opposing views about how immigrants affect the labor market opportunities of American natives. One view is that they have a harmful effect because immigrants and natives tend to have similar skills and compete for the same jobs, thus driving down the native wage. The other view is that the services of immigrants and natives are not interchangeable, but rather complement each other. For instance, some immigrant groups may be unskilled but particularly adept at harvesting crops. Immigration then increases native productivity and wages because natives can specialize in tasks for which they are better suited.
The first view is more likely correct. Economists who have rejected this view on the basis of evidence have looked at somewhat superficial data. These economists speculated that if the services of natives and immigrants are interchangeable, natives should earn less in cities where immigrants are in abundant supply, such as Los Angeles or New York, than in cities with few immigrants, such as Nashville or Pittsburgh. Although natives do earn somewhat less in cities that have large immigrant populations, the correlation between the native wage and the presence of immigrants is weak. If one city has 10 percent more immigrants than another, the native wage in the city with the most immigrants is only 0.2 percent lower.
The results of the Mariel boatlift provide further evidence of how weak the correlation is. In April 1980, when Fidel Castro declared that Cubans wishing to emigrate could leave from the port of Mariel, 125,000 people accepted the offer. Miami's labor force suddenly grew by 7 percent. Yet the trends in wages and unemployment rates in Miami between 1980 and 1985, including those of black workers, resembled those observed in comparable cities.
But all this evidence is superficial. Why? Because it ignores the fact that labor and capital are mobile between cities. If an influx of immigrant workers reduced wages substantially in a particular city, native workers and some immigrants would leave that city and find work elsewhere. And natives who had contemplated migrating to that city would choose another destination. Also, capital may "migrate" to cities with large numbers of unskilled immigrants, where capitalists can earn a greater return on their investment. Large-scale immigration, therefore, may not drive down wages in particular cities. Rather, its depressing effect on wages is nationwide.
Striking evidence for this aggregate effect of immigration is given by the deteriorating economic conditions for native workers with less than a high school education. During the eighties the wage of native high school dropouts fell by 10 percent relative to the wage of workers with more schooling. About a third of this decline is attributable to the increase of unskilled immigrants in the work force, who went from 13 percent of the high school-dropout work force in 1980 to 26 percent by 1988. Thus, a good case can be made that immigration reduced unskilled wages in the United States by about 3 percent (one-third of 10 percent).
Economic Impact of Immigration
Although the entry of immigrants reduces the wage of comparable native workers, it produces a slight increase in the income of U.S. natives overall. Using a well-known formula in economics, we can estimate that immigration increases the real income of natives, but only by about 0.1 percent. (This calculation is based on what is called the Harberger triangle.) That 0.1 percent increase translates to about a $5 billion a year gain from immigration for U.S. natives. [Editor's note: the data in this section are from 1991.] Of course, not everyone benefits equally from immigration; workers with competing skills lose, while owners of land and capital gain.
Many people believe that because a comparatively large percentage of immigrants goes on welfare, the costs to native American taxpayers wipe out the gains from immigration. But this has not been the case in recent years. The numbers show why. The present value of cash welfare benefits (such as Aid to Families with Dependent Children) received by the typical immigrant family over its lifetime is $8,700. With 6.4 million immigrant households, the total cost of assistance programs for immigrants is about $56 billion.
But immigrants also pay taxes. The present value of lifetime earnings for the typical immigrant man is $380,000. With 7.6 million working immigrants (both men and women), total earnings of immigrants are at most $2.9 trillion, of which about 40 percent, or $1.2 trillion, are paid in taxes of all forms. Because 3 percent of total revenues are allocated to cash welfare benefits, immigrants pay about $36 billion ($1.2 trillion times .03) in taxes to fund welfare programs. Comparing the $36 billion that immigrants contribute to welfare to the $56 billion they consume, immigrants consume $20 billion more over their lifetimes than they contribute. Thus, the welfare system causes U.S. natives to lose about $1.1 billion per year (in present value terms). Subtracting this $1.1 billion from the $5 billion annual increase in national income, the United States benefits from immigration, but the economic gains are small.
The net benefit is even smaller when immigrants are relatively unskilled. For instance, suppose that all immigrants have the skill level of those who came in the late seventies. Lifetime welfare costs per household would then be $13,600, and the immigrant population would add $87 billion to welfare costs. These less skilled immigrants only earn $313,000 over their working lives, so that total earnings are about $2.4 trillion. They would then pay about $960 billion in taxes, of which $29 billion is allocated to funding cash benefit programs. The immigrants would drain the U.S. Treasury by about $58 billion over their lifetime, for a net loss of about $3.2 billion per year. Because national income increases by somewhat more, immigration is still beneficial. Note, however, that these calculations do not include the costs of other components of the welfare state, particularly health care. The introduction of these additional programs would further reduce the meager economic benefits associated with the immigration of less skilled workers.
Because the gains from immigration depend on the skill level of immigrants, other host countries (Australia and Canada, for instance) now use a "point system" to allocate visas. Applicants are graded on the basis of such factors as education and occupation, and only those applicants who "pass the test" are awarded entry visas. It is not surprising that people migrating to those countries are more skilled than those admitted by the United States. The United States, in effect, is losing the competition for skilled workers in the immigration market.
The evidence also suggests that a policy allowing unrestricted immigration will not necessarily increase the skill level of immigrants. A good example of unrestricted migration is the population flow between Puerto Rico and the United States. Despite the lack of any legal impediments, Puerto Ricans who migrated to the United States in the sixties and seventies are less skilled, on average, than those who remained in their birthplace. The typical Puerto Rican migrant residing in the United States has about 1.3 years less schooling than his compatriots back on the island. Because Puerto Rico's wage structure has substantial inequality and offers little opportunity for less skilled workers, it is not surprising to find that these are the workers who choose to leave, just as in the earlier example of Mexico.
Changes in immigration policy since 1965 greatly altered the number, national origin mix, and skill composition of immigrants. Importing unskilled workers helps fill menial jobs at low wages, but these immigrants also impose substantial costs, mainly by being disproportionately on welfare. Although U.S. natives benefit from immigration on net, the benefits are small.
George J. Borjas is the Pforzheimer Professor of Public Policy at the John F. Kennedy School of Government, Harvard University. He is also a research associate at the National Bureau of Economic Research. He was a member of the National Academy of Sciences Panel on Immigration Statistics.
Abowd, John M., and Richard B. Freeman, eds. Immigration, Trade, and the Labor Market. 1991.
Borjas, George J. "Self-Selection and the Earnings of Immigrants." American Economic Review 77 (September 1987): 531-53.
Borjas, George J., Friends or Strangers: The Impact of Immigrants on the U.S. Economy. 1990.
Borjas, George J., and Richard B. Freeman, eds. Immigration and the Work Force: Economic Consequences for the United States and Source Areas. 1992.
Borjas, George J., and Stephen J. Trejo. "Immigrant Participation in the Welfare System." Industrial and Labor Relations Review 44 (January 1991): 195-211.
Card, David. "The Impact of the Mariel Boatlift on the Miami Labor Market." Industrial and Labor Relations Review 43 (January 1990): 245-57.
Simon, Julian L. The Economic Consequences of Immigration. 1989.