Heaven's Door: Immigration Policy and the American Economy, by George Borjas, Princeton: Princeton University Press, 263 pages, $27.95
Throughout the history of American immigration, critics have viewed each new generation of arrivals as a threat to the nation's prosperity and culture. George Borjas, an economist and professor of public policy at the John F. Kennedy School of Government at Harvard University, sounds that warning yet again in his new study, Heaven's Door: Immigration Policy and the American Economy. Borjas is no nativist know-nothing: He is the author or co-author of many peer-reviewed articles on immigration and a major voice in the ongoing debate on the subject, so his book demands careful attention. So, too, does his life: Borjas is a Cuban-born refugee, yet the immigration policies he proposes would have precluded his own family's arrival in this country. That irony undercuts his anti-immigrant prescriptions as fully as the inconvenient facts he ignores.
The central argument of Heaven's Door is that America is being harmed economically because it allows citizens and permanent residents (green card holders) to sponsor close family members for immigration. Borjas writes that such immigration patterns are a direct result of the 1965 amendments to the Immigration and Nationality Act, and that we should change the law.
His argument is flawed. First, the book is ahistorical, suffering from a lack of understanding of–or an unwillingness to acknowledge–the discriminatory treatment that the 1965 legislation sought to address. Second, Borjas uses inappropriate data to support his conclusions, failing to distinguish between legal and illegal immigrants, and using information on illegal immigrants to advocate reducing legal immigration. Third, he assumes immigrants remain economically static, ignoring much evidence that they adapt to America by acquiring skills or training unavailable in their home countries. Finally, Borjas writes as if his conclusions are generally accepted –which they are not–and neither engages nor refutes criticisms of his earlier work.
Some history is necessary to understand the changes in immigration policy that occurred in 1965. In the 1920s, Congress replaced what had been a policy of nearly open immigration with "national origins" quotas that, in effect, barred Asians, Italians, Greeks, and Jews. These quota laws, passed after lobbying by the Ku Klux Klan and others, codified the eugenics theories of Madison Grant, whose work focused on the supposedly inferior skull sizes of Jews and other immigrants.
The 1965 amendments eliminated these quotas and carried forward almost intact the family immigration categories put in place by the 1959 amendments to the Immigration and Nationality Act. Although on paper half of the available immigration slots under the quotas were reserved for skill-based immigration, 86 percent of the visas issued between 1952 and 1965 went to family immigration, according to the Immigration and Naturalization Service.
Borjas is simply wrong when he states that the 1965 act "enshrined a new objective for awarding visas…the reunification of families." The historical records at Ellis Island make clear that most immigration prior to the 1920s was also family-based, and such unification never entirely lost its role. In fact, a report of the House Judiciary Committee on the 1959 legislation states, "The recognized principle of avoiding separation of families could be furthered if certain categories of such relatives were reclassified in the various preference portions of the immigration quotas." Joyce Vialet of the Congressional Research Service analyzed the 1965 Immigration Act and concluded, "In response to the demand for admission of family members, Congress enacted a series of amendments to the Immigration and Nationality Act (INA), beginning in 1957, which gave increasing priority to family relationship. The family preference categories included in the 1965 Act evolved directly from this series of amendments. Arguably, the 1965 Act represented an acceptance of the status quo rather than a shift to a new policy of favoring family members."
Under current law, a U.S. citizen can sponsor a spouse, child, parent, brother, or sister, while a green card holder can sponsor a spouse or child. We also accept refugees and employer-sponsored immigrants. Approximately 75 percent of family immigrants are spouses or children. Borjas argues that such immigration practices have caused numerous problems:
Education. Using U.S. Census data, Borjas concludes that immigrants today are less skilled than natives. But the Census includes data about many illegal immigrants. If one measures only legal immigrants, as the New Immigrant Survey (1998) does in a sophisticated research project led by RAND economist Jim Smith, one arrives at quite different conclusions. The New Immigrant Survey's findings directly contradict Borjas. "The median years of schooling for the legal immigrants, 13 years, is a full one year higher than that of the U.S. native-born," the survey concludes. It would seem that, on balance, legal immigrants are not less skilled than natives. Legal immigrants do congregate at the top and bottom of the education scale, but less so than Census data imply. Besides, economists agree that immigrants increase America's labor productivity most when they fill niches at the top and bottom.
Earnings Growth. Based on his study of the Census data, Borjas finds that the earnings of immigrants never catch up with those of natives. But other researchers using the same Census data, including refugees and illegal immigrants, have questioned that finding. Writing in the May 1999 American Economic Review, economists Harriet Duleep, a senior research associate at the Urban Institute, and Mark Regets, a senior analyst at the National Science Foundation, found that the gap in earnings between new immigrants and natives largely disappears after 10 years in the United States, with immigrant wage growth faster than native (6.7 percent vs. 4.4 percent).
Borjas misses this point because he excludes the self-employed, a major statistical blind spot when looking at immigrants. If, out of 100,000 immigrants, 60,000 started restaurants and software firms, and 40,000 worked as waiters, Borjas would count only the earnings of the 40,000 waiters. Moreover, if 10,000 of the waiters later started their own successful restaurants, Borjas would remove them from calculations of immigrant earnings growth, thus further biasing the results downward.
A July 1998 study by Stephen Moore of the Cato Institute supports the Duleep-Regets findings. It shows a direct correlation between time spent in America and economic well-being, finding that after two decades immigrants' home ownership rates exceed 60 percent, while their poverty rates fall below those of natives.
Fiscal Impact. Borjas portrays immigrants as fiscal liabilities by citing information on the annual costs of immigrant households from a 1997 National Academy of Sciences (NAS) study. Yet economist Ronald Lee of the University of California, Berkeley, who performed the key fiscal analysis in the NAS report, testified before the Senate Immigration Subcommittee in 1997 that such data are highly misleading. "These numbers do not best represent the panel's findings and should not be used for assessing the consequences of immigration policies," Lee testified. The problem, Lee found, was that calculating annual numbers requires using a model that counts the native-born children of immigrants as "costs" created by immigrant households when those children are in school but fails to include the taxes those children pay once they grow up and enter the work force.
According to Lee, the entry of a typical immigrant to the U.S. has a positive cumulative impact within the immigrant's lifetime: Taxpayers save $80,000 as a result of that entry, most of it during the lifetime of the immigrant and his offspring. "Most immigrants arrive at young working ages, with their education already paid for," testified Lee, and they "help pay for government activities such as defense for which they impose no additional costs." Borjas asserts incorrectly that any benefits come only after the passage of hundreds of years. As for the fiscal impact of legal immigration on the states, Berkeley's Lee said that, with the appropriate assumptions, a dynamic analysis would likely show 49 of the states coming out ahead, with the 50th, California, a close call.
Overall Immigrant Contributions. Borjas argues that the economic contribution of immigrants to the economy is only $8 billion a year. Cato's Moore and others have called that assertion absurd, noting that Borjas' methodology depends on immigrants being identical to natives and with no benefits accruing to Americans from immigrant entrepreneurs, immigrants with abilities different from natives, a larger economic pie, or a greater selection of goods and services. Borjas asserts immigrants save Americans money–and thereby benefit the economy–only if they lower native wages. That would mean
that Hungarian-born Intel founder Andy Grove has provided no economic benefit to America, even though Intel today employs over 65,000 people and has net revenues of $26 billion a year.
In the San Francisco area alone, 2,775 companies led by Chinese and Indian immigrants have annual sales of nearly $17 billion and employ more than 58,000 people, according to Anna Lee Saxenian of the Public Policy Institute of California. Economies of scale, immigrant-induced productivity improvements, and other factors that Borjas concedes would exponentially increase the immigrant benefit to the economy are not included in his book because, he says, they are difficult to quantify. Even if that's the case, he should at least concede that his $8 billion figure tells us very little.
Borjas further asserts that immigrants who succeed in the U.S. would likely succeed in their own country. But if this is true, then Borjas should provide a list of all the great semiconductor firms started in communist Hungary after Andy Grove fled in 1956. In fact, immigrants come here precisely because oppressive political or economic policies block them from succeeding in their own countries.
Labor Market Impact. Borjas argues that native high school dropouts nationally experience lower wages because of immigrants. This rests on the assumption that low-skill natives leave states in response to increased immigration, thus perhaps explaining why numerous studies have not detected negative wage effects from immigrants. However, 1997 research by Columbia University economist Francisco L. Rivera-Batiz demonstrates that Borjas' theory cannot be correct. To the extent that any native out-migration is measurable in states that receive a lot of immigrants, it's actually college-educated natives who have left (and possibly for reasons that have nothing to do with immigration).
Borjas does not appear to have researched the out-migration question himself, nor does he refute Rivera-Batiz's conclusions. Instead, he writes, "The few studies that attempt to determine if native migration decisions are correlated with immigration have yielded a confusing set of results."
Borjas' labor-market analysis has other problems. For example, he sets up a model that guarantees immigrants have a negative effect by assuming they fill all the same jobs as natives, rather than allowing for the more likely scenario found by many other economists that immigrants are complements to natives in the labor market.
Also, Borjas' assertion that we have too many workers appears mistaken to anyone monitoring today's economy or long-term labor trends. Under current immigration levels, the U.S. labor force will grow by a (possibly inadequate) 40 percent between 1995 and 2050, according to the National Academy of Sciences. Under Borjas' immigration proposals, the U.S. labor force would grow by perhaps half that much.
The restrictionist argument that lower-skilled jobs have almost all disappeared is contradicted by an August 1997 study by economist Linda Levine of the Congressional Research Service. Levine concludes that "many occupations with limited educational requirements are experiencing above-average rates of job growth or substantial increases in employment levels." In 2005, writes Levine, "about one-half of all jobs" in America will require "no more than a high school diploma."
To solve problems that better research indicates do not exist, Borjas recommends the adoption of a Canadian-style point system, in which a government body assigns points to such characteristics as education level and admits only those who achieve a designated score. In practice, Borjas' plan would transfer power to federal bureaucrats at the expense of individuals, families, and employers. "A point system has many imperfections," concedes Borjas. "A few hapless government bureaucrats have to sit down and decide which characteristics will enter the admissions formula, which occupations are the ones that are most beneficial, which age groups are to be favored, how many points to grant each desired characteristic and so on."
After noting that the list of occupations, each assigned points, takes up 10 pages in the Canadian system, Borjas writes, "Most of these decisions are bound to be arbitrary and clearly stretch the ability of bureaucrats to determine labor market needs well beyond their limit." As if bureaucrats are well suited to handle any labor market decisions. In any case, it's clear that no government test can ever measure life's most important intangibles: drive, individual initiative, and a commitment to family.
Borjas concedes that keeping out Mexicans is a goal of the point system. "Most likely," he writes, "the predominance of Mexican immigrants and of immigrants from some other developing countries will decline substantially." In other words it's not just bad policy but bad politics.
It's important to understand that those who advocate a more "skill-based" immigration system are also among the most vociferous opponents of skilled immigrants. In 1998, anti-immigrant groups and their congressional allies fought the expansion of "H-1B" temporary visas for high-skilled, foreign-born engineers, computer scientists, and others, as if adding 50,000 more professionals to a 130 million-person work force would mean the end of Western civilization. Borjas himself derisively refers to these scientists and engineers as "high tech braceros," equating them with migrant farm workers.
Borjas undermines any pretense of rigorous analysis when he writes, "I suspect that an annual flow of one million immigrants is probably too large." We should reduce this number, he thinks, to 500,000 (the average annual immigration level in the 1970s, he later notes). This is a breathtaking denial of opportunity to a half million people on the basis of unsupported suspicions.
But that's not the end of Borjas' proposal. Noting that establishing a point system would be useless if those excluded entered illegally, he argues that we should subtract the number of illegal immigrants each year from the number of legal immigrants. Based on INS estimates, that would further reduce legal immigration to 200,000 persons a year.
What would such a limit mean? If so restrictive a proposal had been the law in 1997, nearly 50,000 citizens would have been unable to sponsor their own spouse or minor children for immigration (there were 248,326 spouses and minor children of U.S. citizens who immigrated legally in 1997). In addition, the government would have prohibited citizens from sponsoring parents, adult children, or siblings, while green card holders would have been forbidden from sponsoring spouses or children. No refugees would have entered. And not one "skilled" immigrant would have been admitted, either through a point system or through high-tech company sponsorship, because no slots would have been available. Thankfully, there's little chance Congress would adopt such a proposal. In 1996, it turned back, by decisive margins, efforts to reduce legal immigration.
As for what to do about illegal immigration, Borjas advocates stiffer employer sanctions and improved identification documents, though he does not appear convinced those will do the trick. Illegal immigration is a legitimate concern. In places such as Douglas, Arizona, where illegal immigrants are literally trampling on ranchers' property rights, the interesting policy question raised by Gov. Jane Hull is whether illegal entry can be curtailed or prevented by a combination of law enforcement and market forces, such as by providing temporary visas to willing workers in hospitality and other service industries. But Borjas merely adopts the view of many restrictionists: We're not sure we can stop people who are entering illegally, so let's go after the people who immigrate legally. That's not an acceptable policy.
In 1962, George Borjas' widowed mother, with the help of Catholic priests, came with George to America on one of the last "freedom flights" from Cuba. With little money or education, his mother worked hard to raise George, and he obviously worked hard, too. Borjas writes, "Although my family and I entered the country as refugees, my family would have been unable to `pass the test' implicit" in the policy he now advocates.
Critics of immigration can't explain why, if legal immigration is supposed to be so harmful, America remains a society rich in opportunities. George Borjas can't explain it either. "To this day, I continue to be amazed by the courage and boldness that my mother and millions of others exhibited in picking up the little they had, and starting life again in a foreign country–without knowing the language, the culture, or almost anything about it," he writes. "They all relied on their unshakable belief that the United States was a far better place, and that even if they themselves could not share in those opportunities, their children surely would." This is the same dream that brings so many to America. We benefit and they benefit.
We should admire his mother's courage and all that Borjas has accomplished. And by all means, we should read and ponder his book. But in the end it is George Borjas' life, not his book, that teaches us the most about the right immigration policy for America.