The Public Interest, Fall 1998, issue no. 133
Partly as result of reforms undertaken in the 1960s, the United States is currently experiencing the largest sustained wave of immigration in its history. Each year the United States admits between 700,000 and 900,000 legal immigrants; additionally, the Immigration and Naturalization Service estimates that 5 million illegal aliens now live in the country, with 400,000 new illegal aliens settling here annually. This influx of legal and illegal immigration has caused the foreign-born share of the population to double from roughly 5 percent of the population in 1970 to 10 percent today. While less than the 15 percent recorded in 1910, the 27 million immigrants now living in the country is more than twice the 1910 number.
A fierce national debate has erupted over the wisdom of allowing in such a large number of people. The most notable recent study of the issue comes from the National Academy of Sciences' National Research Council (NRC). Prepared by many of the top economists, demographers and sociologists in the field, the study, The New Americans: Economic, Demographic, and Fiscal Effects of Immigration, contains an extensive analysis of the latest research on the subject as well as a good deal of original work. The media has devoted considerable attention to the study. However, misled by the study's rosy press release, they have claimed that The New American unequivocally confirms the benefits of immigration. In fact, the NRC study found that the total effect of immigration on workers, employers, consumers, and taxpayers is probably small. But it also finds - and this has generally gone unreported in the media - that the negative effect of immigration on unskilled workers and taxpayers in high-immigration states is substantial.
The disparity between the study's actual findings and its portrayal in the media caused George Borjas and Richard Freeman, two Harvard economists who helped write the study, to chastise the report's editor, James Smith of Rand, in a New York Times. They accused him and Senator Spencer Abraham, chairman of the Senate Immigration Subcommittee, of distorting the study by portraying immigration as a "free lunch." In this essay, I will explore what The New Americans actually says about who wins and who loses from immigration and what policy implications might be drawn from its findings.
Immigration's Effect on Wages
One of the most hotly debated aspects of immigration is its effect on the wages and job opportunities available to natives. After reviewing the literature in the field, the NRC study concluded that the negative effect of immigrants primarily takes the form of wage losses for workers who lack a high school degree. This is the only educational group adversely effected by immigration because such a large percentage of recent immigrants lack a high school education - about 40 percent. And it is only in this skill category that the proportion of immigrants is large enough, 25 percent by the mid-1990s, to exert a significant downward pressure on wages.
To arrive at its estimates, the NRC study used Census and other government survey data to compare increases in the supply of one type of worker relative to other workers brought about by immigration. The wage consequences of immigration are then calculated using an existing body of research that has measured the wage effects resulting from shifts in the relative supply of labor. An increase in the supply of unskilled workers is assumed to lower wages for those natives in competition with immigrants. In contrast, those who tend not to be in competition with immigrants should see their wages rise as demand for skilled labor increases. Additionally, consumers should benefit from price reductions for goods produced by unskilled labor and owners of capital should also realize a better return on their investments from falling labor costs.
Using this method, the NRC study estimated that immigration was responsible for 44 percent of the decline in wages that high school dropouts experienced between 1980 to 1994. Some research, including my own, indicates that the wage effects of immigration on workers with few years of schooling are larger than those reported in the NRC study. But, if we accept the estimates in the study, it means that 13 million workers, a group that roughly corresponds to the poorest 11 percent of the labor force, are experiencing an immigration-induced reduction in wages of roughly 5 percent or $13 billion a year. These losses should generate economic benefits for owners of capital, and more skilled workers who can now be paid more. The estimated size of the gain is thought to exceed the loss to the unskilled by roughly $5 billion or around one-tenth of one percent of our $7 trillion economy.
It is important to note that these calculations do not take into account the marginal utility of income. That is, each dollar lost or gained is more valuable to those with low incomes because they have less money to begin with. If this had been included in the calculation, then the overall effect on the economic well-being of natives is likely to be negative. Putting aside this question, as well as immigration's effect on public coffers, which will be discussed later, the NRC report's estimates tell us two things: First, the overall impact of immigration is minuscule given the enormous size of the American economy. And second, the primary economic effect of immigration is to redistribute income from the poor and unskilled to skilled workers and owners of capital.
The study relies on changes in the relative supply of labor to determine the effects of immigration because it is now widely accepted that the labor market effects of immigration are likely to be national in scope. In the past, economists had compared cities with differing percentages of immigrants and generally found little effect from immigration. However, this method is now widely viewed as inadequate. Research by University of Michigan demographer William Frey and others indicates that native-born workers migrate out of cities and states as immigrants move in, especially natives with few years of schooling. The movement of natives in response to immigrant competition spreads the labor-market effects of immigration from high-immigrant areas to the rest of the country.
In addition to internal migration, the huge volume of goods and services exchanged between cities across the country creates pressure toward an equalization in the price of labor. For example, newly arrived immigrants who take jobs in light manufacturing in a high-immigrant city such as Los Angeles come into direct and immediate competition with natives doing the same work in a low-immigrant city like Pittsburgh. The movement of capital seeking to take advantage of any immigrant-induced change in the local price of labor should also play a role in preserving wage equilibrium between cities. Besides the response of native workers and firms, immigrants themselves tend to migrate to those cities with higher wages. In short, the mobility of labor, goods, and capital, as well as choices made by immigrants, make it very difficult to determine the impact of immigration by comparing cities. By calculating the increases in the relative supply of labor nationally, the NRC study avoids this problem.
The findings of the NRC study also suggest that some groups are more adversely affected by immigrant competition than others. For example, because blacks in the work force are 33 percent more likely to lack a high school degree than whites, a much higher proportion of them are hurt by immigration. There is also a good deal of anecdotal evidence, and some systematic research, to indicate that many employers see immigrants as more reliable and hard working than native-born blacks. A recent study of the Harlem labor market, by Katherine Newman and Chauncy Lennon of Harvard, found that 41 percent of the immigrants in their sample of low-wage workers were able to find jobs within one year, while only 14 percent of native-born blacks were able to do so. The authors conclude that immigrants fare better in the low-wage labor market because employers see them as better workers. If this is the case, then it means that there is an added effect on unskilled blacks in competition with immigrants beyond just the wage reduction experienced by all unskilled workers.
Another group that is disproportionately harmed by immigrant competition is the long-term welfare population. Because two-thirds of the this population lacks a high school diploma, immigration is likely to make it more difficult for them to find jobs that pay enough for self sufficiency. And probably those most adversely affected by immigration are earlier waves of immigrants, who often find themselves in direct competition with newer immigrants. Overall, those groups that are disproportionately represented among the poor and unskilled will bear the brunt of the negative wage effects of immigration.
Immigration's Fiscal Impact
Immigrants not only affect natives by their presence in the workforce, but also affect natives through their consumption of public services. Based on Census and other government survey data, the NRC study concluded that, at the present time, immigrant-headed households consume more in public services than they pay in taxes. The current fiscal burden - tax revenue minus expenditures - imposed on all levels of government by immigrant households is estimated to range from $11 billion to $20 billion. That's more than the report's estimate of the net gain from having immigrants in the work force. While large in absolute terms, when one considers that total expenditures by all level of government exceeds $2 trillion each year, the added fiscal burden is relatively small. However, as is the case with the wage effects of immigration, the negative fiscal consequences are not evenly distributed across the population.
Six states - California, New York, Texas, Florida, New Jersey and Illinois - which have only 38 percent of the nation's total population, account for three-fourths of the immigrant population. It is in these high-immigrant states that the negative fiscal effects of immigration are felt. In New Jersey, for example, the average household headed by an immigrant used $1,484 more in state and locally provided public services than they paid in taxes each year; in California the figure was $3,463. If one were to multiply these averages by the total number of immigrant households in each state, and then divide it by the number of native households, one would discover that the added tax burden imposed on the each native household in New Jersey is $232; in California, $1,178.
The study also found that immigrants have a net positive effect at the federal level. In New Jersey and California the average immigrant household currently uses $520 and $127 less, respectively, in federally provided services than they pay in federal taxes. This translates into a net benefit of $3 and $4 to each native household in the United States. Combining the positive fiscal effect at the federal level with the negative fiscal effect at the state and local level reveals that the negative fiscal impact of the average immigrant household is $964 in New Jersey and $3,336 in California.
Immigrants have a negative fiscal impact for three reasons: First, they have more children than natives and so consume more in educational services. Second, immigrant households are poorer than native households and, therefore, receive more in state and locally funded income transfers. Third, because immigrant households have significantly lower incomes than native households, they pay less in taxes. In 1996, for example, immigrants who entered after 1980 reported average annual earnings that were only 65 percent that of natives. The reason that immigrant households have a positive effect on the federal budget is that they are assumed to impose no additional costs on pure public goods provided by the federal government, primarily defense. Thus, expenditures on defense are applied only to native households.
While at present immigrants do not pay enough in taxes to cover their consumption of services, the NRC study also attempted to measure their long-term fiscal impact. The study found that if today's immigrants and their descendants use services and pay taxes over the course of their lifetimes in a manner similar to immigrants and their descendants in the past, then an immigrant who enters today will begin to have a positive effect on public coffers in about 22 years. It takes an additional 18 years for immigrants and their descendants to pay back the negative burden they imposed for the first 22 years. This long-term fiscal benefit depends entirely on the taxes that the descendants of immigrants are expected to pay and not on the tax contributions of the immigrants themselves. When only the tax payments and public services used by the immigrant himself are considered, the NRC study found that the average immigrant does not pay enough in taxes to cover his consumption of public services during his lifetime.
It is important to note that these estimates, in accordance with standard fiscal practice, are discounted at an annual rate of 3 percent a year to reflect the fact that dollars lost or gained in the future are less valuable than are dollars today. This means the negative fiscal effect from immigrants that occurs in the short and medium term is weighted more heavily than the positive effects that are projected to come after 22 years. But there is great uncertainty about the future structure of the economy, tax rates, public expenditures and the economic progress of immigrants and their American-born children. The NRC study's long-term multi-generational projections make certain key assumptions about the future. For example, the study assumes that, because the current rate of growth in spending on Medicare, Medicaid and Social Security is unsustainable, the entitlement crisis will be solved in the year 2016 by an even split of taxes and spending cuts. If this assumption turns out to be wrong, the study concludes that immigrants and their descendants will be a fiscal burden in the next century.
Not surprisingly, the study also found that the fiscal effect of an immigrant depends heavily on his skill level and earnings. Immigrants who have few years of schooling are always a net burden. In contrast, skilled immigrants are almost always a net benefit. The reason for this is simple. Individuals, immigrant or native, with few years of schooling tend to have low earnings and therefore pay less in taxes on average and use more in public services. Thus to reduce the fiscal burden from immigration, we must select immigrants with more education and higher earnings.
Costs versus Benefits
The combined economic and fiscal effects of immigration in the NRC study point to a fundamental dilemma. The economic benefit from immigration comes from the fact that immigrants are significantly less skilled than natives. The resulting shift in the relative supply of labor is what causes the wage reductions that, in turn, generate the net gain for employers, and others. But unskilled immigrants also have a negative fiscal impact. In other words, it is precisely those workers who create the economic benefit who are responsible for the fiscal burden.
This paradox makes it very difficult for a society such as ours, with a well-developed state and extensive social programs, to benefit from immigration. Of course, if we were to fundamentally reduce the size of government at the state and local level, then perhaps the country could benefit economically from immigration without creating a fiscal burden. The cuts would have to come in such areas as education, law enforcement, and infrastructure maintenance, because this is where state and local government spend most of their money. But, even assuming that such cuts are desirable, they remain politically unlikely. A second possible solution would be to cut immigrants off from as many public services as possible. However, this too seems neither desirable or likely.
More important, neither of these solutions gets at the real issue. Immigrant use of federally funded welfare programs, though higher than that of natives, has little to do with the fiscal burden they create at the state and local level. That burden is caused primarily by their lower incomes and the resulting lower tax payments.
A Question of Policy
The findings of the NRC study basically tell us this: We can now say with some confidence that immigration is not the benefit to the economy that it was once thought to be. While the overall economic impact of immigration is modest, the effect on the unskilled and on taxpayers in high-immigrant states is substantial. We should, therefore, change our immigration policy not because the negative effects are so large for the country as a whole but because they are so concentrated.
The most important policy reform would be to give greater weight to immigrant skills as criteria for admission. At present, roughly 70 percent of legal immigrants entering each year do so because they are related to a U.S. citizen or a legal permanent resident (legal residents who have not become US citizens). The rest consists of employment-based immigration, which accounts for about 13 percent, humanitarian-based immigration (refugees and asylees), which is slightly smaller in size, and a visa lottery that gives away 55,000 visas a year to persons who mail in a post card and "win" the opportunity to come to the United States. With the exception of employment-based immigrants, who are allowed entrance because they are thought to have needed talents or skills, over 85 percent of legal immigrants are granted admission without regard to their likely economic or fiscal impact.
The bipartisan U.S. Commission on Immigration Reform, headed by the late Barbara Jordan, recognized the need to increase the skill level of entering immigrants. It suggested limiting family immigration to the spouses, minor children and parents of citizens and the spouses and minor children of legal permanent residents. This would eliminate the preferences now in the law for the brothers, sisters, and adult children of citizens and the adult children of legal permanent residents. A strong case can also be made for eliminating the current preferences for the spouses and minor children of non-citizens, since these provisions apply to family members acquired after the alien has received a green card, but before he has become a citizen. Reforming family immigration in this way would not only reduce unskilled immigration immediately, but would also help cut off the chain migration that occurs as the spouses and siblings of those admitted in these categories petition to bring in their relatives. The Commission also wisely suggested eliminating the visa lottery and tightening up the requirements for employment and humanitarian-based immigration. If all of the above measures were implemented, the overall level of legal immigration would likely fall to less than half its current level within a few years.
But, even with these changes, the United States would still continue to admit far more immigrants than any other country. We should also do more to fight illegal immigration. This is undoubtedly the lowest-skilled flow of immigrants, with at least 75 percent lacking a high school degree. It is estimated that one-third of the immigrant high school dropouts holding jobs in the United States are illegal aliens. While border control is important, there is broad agreement among those who study illegal immigration that denying illegal aliens access to jobs offers the best hope of stemming illegal immigration. Therefore, implementing a system of employment verification and hiring more inspectors to ensure that the law against hiring illegal aliens is enforced should be made a top priority.
Cuts in legal immigration would also be helpful in reducing illegal immigration because the two are so closely linked. As the annual level of legal immigration has grown, illegal immigration has grown proportionally. Communities of recent legal immigrants serve as magnets for illegal immigration by providing housing, jobs and entree to America. In fact, 25 percent of legal immigrants admitted each year are illegal aliens using a provision in the law which allows them to "adjust" their status. Any serious attempt to control illegal immigration will have to deal with this fundamental reality.
Finally, in formulating an immigration policy that makes sense, it is important that it not be confused with immigrant policy. Immigrant policy is concerned with how we treat foreigners living in the United States. Barring legal immigrants from using some federally funded welfare programs, as Congress did in 1996, is an example of a change in immigrant policy - a change that was neither effective nor fair. Immigration policy, in contrast, is concerned with the question of who may come and how many. And it is here that the changes are needed. Polls repeatedly find large majorities of Americans in almost every demographic group in favor of significant cuts in legal and illegal immigration. Tapping into the broad public anxiety over immigration would likely be politically popular, as long it did not scapegoat immigrants or confuse immigrant and immigration policy.
Unfortunately, the failure to make such basic distinctions has allowed an odd, but formidable, coalition of pro-immigration interest groups - including ethnic advocates, elements of the business community, libertarians, and the far left of the Democratic Party - to dominate the immigration debate. They have been very effective in preventing changes in immigration policy which would better serve the country's interests. If we are unwilling to change immigration policy, then we ought seriously to consider doing more to help those harmed by it. Anything less would be unjust and unwise.