The Great Immigration Debate

Like a great glacier carving valleys, feeding rivers, and depositing soil, immigration is reshaping America's character and future -- her economy, workforce, family structures, demography, culture, cuisines, languages, and politics. Yet of all the first-order policy issues facing the nation, this may well be the hardest one for us to approach rationally.

We are hardly alone, of course. Immigration also arouses fierce passions in France, Spain, Japan, and most other nations. But Americans feel a special skittishness and ambivalence about the subject. Our self-contradictions abound. Defining ourselves as a nation of immigrants, we also view immigration as a threat. Defending our autonomy, we also invite millions of strangers to come here, transforming our society Demanding secure borders against illegal workers, we also advocate the free movement of goods, technology, and capital. Growing global interdependencies further muddy the debate, making it harder to know who "we" are and what "our" national interests really are.

In one sense, immigration policies are highly constrained; in another, they are up for grabs. Any sovereign nation must decide who may enter, who may become a member, and on what terms. In the American welfare state, completely open borders would be a political impossibility, engendering a harsh backlash against immigrants that could make earlier nativism pale by comparison. Within that constraint, however, there are any number of ways to combine and weight our multiple goals in admitting immigrants. We let in some immigrants to reunite families; others because of the labor, skills, or other resources they bring; still others for a mixture of humanitarian and geopolitical purposes -- providing a haven for refugees and embarrassing regimes we oppose. This diversity of goals, coupled with an intense demand for U.S. immigration slots, gives immigration policymakers some latitude.

Congress usually manages to keep this incendiary subject off its agenda; in the past, major policy shifts occurred infrequently (1924, 1952, 1965). In the 1970s and '80s, protracted legislative wars produced three far-reaching new laws, leading a battle-scarred Congress to hope that it might move on to less painful subjects. But it has not turned out that way Now only a few years after Congress passed the omnibus Immigration Reform and Control Act of 1986 (IRCA), it is about to revamp the entire structure of legal immigration. Once again it must grapple with the complex politics of immigration.

The ideological poles of the current debate are easy to categorize. At one end are libertarians and free market purists. They favor not just expansion but essentially open borders; limits on movement are acceptable only as a concession to misguided public demands. Opposed to an active role for government, they fear welfare-state redistributions of wealth and status to immigrants and citizens alike. Political freedom to migrate in search of liberty and economic freedom to migrate in search of high rewards are, for the expansionist, two sides of the same coin.

At the restrictionist end is a melange of groups animated by anxieties about migrants' effects on the environment, labor market competition, population growth, and public services. These groups, which include some labor union and civil rights liberals as well as environmental activists, defy conventional partisan or ideological classifications. Some restrictionists also harbor nativist or racist animus toward immigrants. These feelings, however, are stigmatized at the level of national debate; individuals may voice them but organized political groups do not.

Most Americans fall somewhere between these two poles. Many with little sympathy for libertarian or free market orthodoxies view the present levels of immigration as sufficient. Some favor more of it. Americans generally admire immigrants, value ethnic diversity and take pride in our far-flung origins. Those animated by a vision of cultural assimilation seek a melting pot, while the more pluralistic prefer a cultural stew. Still others are stirred by humanitarian ideals. But despite their pro-immigrant sympathies, they all share a concern about how immigration affects one or another aspect of American life; their particular concerns tend to reflect their class, locality, ethnicity, and other factors. In this vast middle range of opinion, the liberal-conservative axis is a poor guide to attitudes toward immigration.

Economists are prominent in the current policy debate. They emphasize that people cross borders for much the same reason that Toyotas, computer programs, and Eurodollars do: their expected economic value will be greater at their destination. To economists, the policy problem is to decide what kinds of skills we need and then to devise ways to induce workers who possess them to come. Correct criteria will make us a richer country; poorly designed ones will leave us poorer. So long as newcomers produce more wealth than they consume, the more the merrier.

Most citizens and elected leaders have a very different view of immigration. They see immigrants less as "human capital" than as bearers of alien cultures with distinctive values, language, interests, and claims. Newcomers will have children, take up space, use public facilities, compete for jobs, housing and preference, eventually vote, and seek to mold our society to their own visions. Natives may view these changes with pleasure or alarm, but they know that they are irreversible. To the lay mind, immigration is not just a wealth-enhancing transfer of resources; it is also an enormous social gamble. All societies are risk averse and hedge their bets. They impose the kinds of limits that economists find inefficient.

In this essay, I argue as an expansionist strongly opposed to open borders. Although our immigration numbers are higher than they have been since World War I and higher (in absolute terms, not per capita) than in other industrialized nations, they should be higher still. But expansionists must also persuade an anxious public to accept higher numbers. Unfortunately, many expansionists have little interest in how popular attitudes have shaped existing limits, while others disparage those attitudes. Not surprisingly, then, politicians do not take expansionist views, including those emphasizing the economic benefits of immigration, as seriously as they should. This is regrettable: Although economists' stories are woefully incomplete, they have much to teach us. They can explain why more immigrants would be good, why different immigrants might be even better, and which are the best ways to decide how many, and which, to admit.

The Unexpected Wave

The United States receives many more immigrants than any other country in the world. (The hemorrhage of East Germans moving west, 2,000 per day at its height, is presumably temporary.) The U.S. admitted about 6.3 million new immigrants as legal residents in the 1980S. As the decade ended, moreover, the trend was strongly upward; the 1989 figure soared to almost 1.1 million from the 1988 level of 643,000, which in turn was up from 601,000 in 1987. (Special amnesty programs were responsible for 492,000 of the 1989 admissions and 69,000 in 1988). Including 2.2 million illegal resident aliens being amnestied, legal admissions in the 1980s were 8.5 million, almost double the total in the 1970s and close to the almost 8.8 million record set during the century's first decade when immigration was essentially unrestricted. (An estimated 160,000 emigrate from the U.S. annually; 30 percent of immigrants eventually leave).

This total, of course, excludes new illegal migrants. They come primarily from two groups -- illegal entrants, and those who enter legally on student, tourist, or work visas and then "overstay" As of 1986, the best estimate of the number of resident illegals was 3 to 4 million, of which 2.7 million will soon be legalized. New illegals enter constantly, of course; a 1987 study estimated a net increase of 100,000 to 300,000 each year. A RAND-Urban Institute study of employer sanctions, a policy innovation introduced in the 1986 immigration law, found that they somewhat reduced the illegal flow, but it seems to be growing again. Although the 1980s migration was very large in both absolute and historical terms, it is less significant as a proportion of the U.S. population. The foreign-born constituted 14 percent in 1900; today's figure is under 7 percent, which is smaller than the proportion even in traditionally homogeneous countries like Great Britain and France. The historical comparison, however, tells only part of the story. It omits certain characteristics of the new flow that help explain its high political profile. According to Census estimates, immigration now accounts for 27 percent of growth in an otherwise slowly increasing population. The new immigrants, whether legal or illegal, are also highly visible because of their concentration. In 1989 fully 80 percent of the legals planned to live in six states; 43 percent planned to live in seven metropolitan areas in California. Illegals, who depend more on informal job and protection networks, congregate even more than that; over 52 percent of amnestied illegals will live in those seven California communities. In cities like Miami, Houston, Los Angeles, and New York where immigrants have quickly come to dominate important parts of the local economy, culture, and political system, their small share of the national population seems beside the point.

Unlike most earlier newcomers, moreover, these are racially and linguistically distinct from most of the native population. The top ten source countries for legal (including amnestied) admissions in 1989 were Mexico, El Salvador, the Philippines, Vietnam, Korea, China, India, the Dominican Republic, Jamaica, and Iran. Of these, only Filipinos, Jamaicans, and Indians are likely to be proficient in English. But the newcomers, as we shall see, also tend to have less education and fewer skills than either the native population or earlier waves of immigrants. 'These gaps (which would be wider if illegals were included) generate disadvantage and hostility.

These new migratory patterns are propelled by demographic and geopolitical "push" factors over which we have little control. But changes in U.S. law have also helped to "pull" new immigrants to our shores, usually with far-reaching consequences that Congress failed to foresee and would probably have opposed. The most important by far was the 1965 law. An often-overlooked achievement of the civil rights movement, it repealed the Europe-centered national origins quotas first adopted after World War 1. Those quotas had set ceilings, country by country, based on the national origins of the U.S. population as recorded in earlier censuses when southern and eastern European countries were a less significant source of immigrants. In the 1965 law, Congress placed a new emphasis on family unification, intending to favor those from southern and eastern Europe, who by then had grown numerous and politically influential. The legislation gave unlimited immigration slots to immediate kin of U.S. citizens and legal residents and 80 percent of the numerically limited slots to their other close relatives. But as it turned out, the 1965 law spawned migration from regions that had previously supplied relatively few immigrants: Asia, the Pacific, Latin America, and Africa. Kennedy administration officials predicted that only 5,000 Asians would migrate in the first year and virtually none thereafter, and that few Western Hemisphere immigrants would come. Those predictions could not have been more wrong. In 1976 Congress imposed the first ceilings on Western hemisphere immigrants.

The Refugee Act of 1980 is another case of unanticipated consequences and unrealized goals. It was intended to regularize the flow of humanitarian-based admissions by asserting more congressional control over an ideologically and geographically driven, ad hoc, discretionary process. Although the Refugee Act gives Congress as well as the President influence on the number and geographical origins of overseas refugee admissions, the actual patterns of refugee selection have changed little. They are still based on ideological affinity prior linkages to the U.S. government, foreign policy goals, and family ties. Admissions still benefit the same groups (especially Soviet Jews, Pentecostals, and Indochinese who supported U.S. policy in Southeast Asia) to the disadvantage of everyone else (especially Africans). The act's universalistic aspirations were further undermined by a recent amendment sponsored by Senator Frank Lautenberg, which relaxes the eligibility standard for particularly favored groups. Even the 1980 act's effort to limit the number of refugee admissions has failed; over 100,000 were authorized in 1989, more than twice the "normal flow" specified in the statute.

What has generated the most controversy, however, is the asylum provision, designed for refugees who reach U.S. territory rather than seeking resettlement from overseas. Although the asylum remedy is an essential part of any human rights policy, Congress added it almost as an afterthought. No one expected it to affect the immigration system, yet asylum is now a significant source of new admissions. It also seriously complicates immigration enforcement.

Only weeks after President Carter signed the law, about 125,000 Cubans fleeing the Castro regime reached the south Florida shores. The Carter administration initially welcomed the "Marielitos." But as their number mounted and many appeared to be criminals and mentally ill, hospitality quickly turned to hostility. Goaded by state and federal officials, the Immigration and Naturalization Service (INS) moved to exclude and deport them. At the same time, many Haitians, Salvadorans, Nicaraguans, and others migrated to the United States to escape political and economic turmoil. Although many agreed to return to their countries voluntarily (the most common immigration enforcement technique), thousands refused and applied for asylum under the new law.

The INS, laboring in quiet but convenient obscurity was wholly unprepared for this unprecedented flood. Its leadership, caught in a Presidential transition, was weak. Its highly decentralized structure invited inconsistent policy implementation. Its system of adjudication, designed for more traditional case flows, was simply overwhelmed.

Asylum claims were the new jokers in the deck. Aliens with no chance of gaining legal admission to the U.S. might get a foot in the door by filing them. Few could meet the law's requirement that they show a "well-founded fear of persecution on account of race, religion, nationality me be ship in a particular social group, or political opinion." But virtually anyone (especially with a lawyer) could hope to delay formal expulsion long enough -- certainly for months, often for years -- to gain some other form of relief or to melt into the large clandestine world of illegal aliens. Asylum claimants thus had everything to gain and nothing to lose. (Even being detained pending decision on the claim would leave them no worse off; they could always gain their freedom by accepting deportation. Marielitos were an exception here; Cuba, flouting international law, refused to take them back). The backlog of asylum claims steadily mounted, reaching almost 200,000 in the early 1980s.

The INS responded by adopting harsh, arbitrary rules. For example, it processed asylum claims wholesale rather than individually; coerced some new arrivals to waive their rights and leave the U.S.; detained others under poor conditions for long periods (sometimes indefinitely) rather than consider parole; and interdicted Haitians on the high seas, forcing them to return without a genuine hearing. Immigrants challenged each of these practices in court but without much hope for success. In the past, judges had upheld almost all federal immigration policies, viewing them as expressions of U.S. sovereignty to which ordinary constitutional principles did not apply. They accorded Congress (and its agent, the INS) "plenary power" to regulate immigration and protect national borders pretty much as they saw fit. Policies that in other contexts would have been flatly unconstitutional easily passed judicial muster.

By the early 1980s, however, something had changed. While the courts continued to cite the same old cases and acknowledge the government's s plenary power over immigration, the immigrants began to win important cases. The courts, drawing on constitutional and administrative law norms long established in other policy areas, invalidated key INS policies women and youths are entering, perhaps presaging longer periods of illegal residence and a greater dependence on public benefits, both matters of great political sensitivity. Finally, the law's initial effect appears to be waning. The amnesty has also generated some illegal migration by the families of newly-legalized aliens. Evasion is simple without an effective system of worker identification. Congress is being urged to repeal or dilute employer sanctions. Hopes for a new amnesty are rising among those ineligible for the first one; indeed, 4,000 new applications are still being filed each week under pending court orders. Most ominous, the Central American economies continue to deteriorate.

Certain developments could inhibit illegal migration. More liberal regimes in eastern Europe and Nicaragua should make it much harder for illegals from those countries to make non-frivolous asylum claims to gain a foothold in the United States. Economic growth, population control, and civil peace in major source countries like Mexico and Nicaragua could reduce migratory pressures (although most migration, like most revolution, is spurred by rising expectations, not hopelessness). More vigorous INS enforcement of employer sanctions may discourage illegals from remaining or coming here. Still, policymakers would be foolish to count on such changes to reduce illegal migration much below recent levels.

What the Economists Tell Us

During the 1980s, the U.S. invested heavily in a series of policy changes aimed at tighter control of our borders. This investment has yielded limited dividends and may even have aggravated the problem it was supposed to solve. Was the game worth the candle? Some economists argue that we have played the wrong game. Preoccupation with illegal aliens, they believe, has diverted us from the two most important issues: Which kinds of aliens should we seek, and how can we get them to come?

The first question naturally leads to an inquiry into how new immigrants perform economically and socially Two recent books, one by Julian Simon, The Economic Consequences of Immigration, and the other by George Borjas, Friends or Strangers, illustrate that the answer to this question depends on how it is posed. Where Simon tends to discuss how immigrants do on average, Borjas disaggregates the data in order to distinguish among immigrants according to when and from which countries they came. Thus Simon finds that legal immigrants to the United States are young and as well educated as natives. Borjas, however, says that the more recent waves of migrants (those that came in the late 1970s) are much older than natives -- their percentage of senior citizens is twice as high -- while their education levels are bifurcated; many are well-educated, but a disproportionate fraction have fewer than nine years of schooling. Simon argues that immigrants are no more welfare-prone than natives, yet Borjas notes that recent waves are more welfare-prone than earlier ones; participation levels among female-headed families in these groups sometimes exceed 30 percent. Simon celebrates immigrants' innovative and entrepreneurial skills. Borjas argues, however, that the progress of the most recent arrivals is being retarded by their dependency on the ethnic market.

Like some other immigration economists, Simon and Borjas find that legal and illegal immigrants do not reduce native wage and employment levels (although foreign-bom workers are more vulnerable), do not significantly burden the public coffers (although local governments are hardest hit), and generate less chain migration by family members than restrictionists claim. But Borjas' main concern is the effect of the 1965 reforms. Any hopes that the reforms would spur economic growth, he shows, have been disappointed. The immigrants whom they favored have less education and job skills than those who entered under the national origins quotas in the early 1960s. (One wonders how much of this difference is explained by the Cubans, disproportionately professionals, who came then). The more recent waves have experienced lower wages and income, more unemployment, greater welfare dependency, more insularity and slower mobility. (One also wonders whether their performance reflected the lower growth rates during the 1970s and 1980s.) Even worse, Canada and Australia, our main competitors for skilled immigrants, manage to attract more highly qualified ones than we do.

Borjas develops an economic model of immigration to explain this pattern. Would-be immigrants evaluate their prospects in the home country with those in possible receiving countries by comparing each country's "returns to skill" -- the premium that its labor market pays for superior job performance. Borjas posits that a country's return to skill depends on its income distribution; a more egalitarian society like Sweden gives a lower return to skill than a less egalitarian one like the United States. Thus low-skill workers from unequal countries would do better in more equal societies, while their high-skill compatriots will want to stay at home where they are at the top of the heap. By the same token, low-skill workers from egalitarian countries will not want to move to unequal ones, while their high-skill compatriots will. The word "want" is important here because no immigrant selection criterion compels anyone to migrate; workers will migrate only if the receiving country's "offer" -- its economic opportunity -- is better for them than alternative offers.

Borjas recognizes, of course, that although his model assumes other things are equal, they actually are not. For example, income levels may be so much higher in one country than another that this difference swamps the returns-to-skill factor. Thus, as he notes, few high-skill Swedes want to move to the U.S., and low-skill workers may still prefer a more unequal country even though they will be at the bottom of the income distribution there. His point is that for any given source country, the model predicts which group of workers -- low or high skill -- is more likely to migrate to the U.S. The Mexicans who come here are likely to have low skills relative to other Mexicans; the Swedes who come here are likely to have high skills relative to other Swedes.

This model enables Borjas to explain why the post-1965 immigrant cohorts perform worse economically than the pre-1965 cohorts, and why migrants to Canada and especially Australia perform better. In the United States, the 1965 law introduced two fundamental changes. First, it radically altered the mix of source countries in favor of relatively inegalitarian countries, which encouraged their lower-skill workers to seek entry. Second, for any given source country, the law made family relationship rather than skills the principal admission criterion, which magnified the first effect. In contrast, Canada and Australia adopted point systems that gave far greater weight to skills, with the result that a source country's migrants perform better in Canada and Australia than do their compatriots who go to the U.S.

Borjas is refreshingly candid in acknowledging that the tradeoffs underlying any politically acceptable immigration policy preclude any purely efficiency-oriented approach. He recognizes that unskilled labor keeps consumer prices for some products lower than otherwise, and that this gain must be weighed against the gains that a more highly skilled population would generate. This balancing would be hard enough if these different gains (and accompanying costs) all accrued to the same people, but of course they do not. The varying effects on different groups require the kind of distributional choices that fuel immigration politics, choices that economists have no special competence to evaluate. What he does claim is that the post-1965 system imposes significant social costs.

Attracting the Skilled

If we should seek more highly skilled immigrants, how can we identify and attract them? He first considers the kind of point system established by Canada and Australia. Under this system, which Congress will probably adopt in some form, applicants are ranked according to their education, labor skills, English fluency, and other factors conducive to successful assimilation. Although such a system would be better than the status quo, Borjas contends, it suffers from some endemic problems. No point system can discriminate carefully enough among the labor skills of demographically similar people. (Despite its point system, Canada has suffered a similar, though smaller, decline in immigrant skill levels). Nor can a point system induce skilled people to accept our "offer" to come here if they have better offers at home or elsewhere.

Economist Gary Becker has proposed that the U.S. auction off its immigrant visas to the highest bidder, an idea endorsed by some others including Simon. But Borjas worries (as Simon does not) that this approach, like current policy, may simply attract the unskilled who have more to gain (a claim that seems somewhat inconsistent with his earlier argument about differential returns to skills). For many non-economists, the auction idea is morally repugnant, one more intrusion of market values into the realm of public choice. These critics forget, however, that the current system exhibits many of the objectionable features of market allocations, plus some others. As Borjas reminds us, the government already sells visas; it just happens to use a pricing scheme that favors those with family connections and clever lawyers. And Congress will probably approve a new provision, common elsewhere, granting visas to foreigners who agree to invest $X million and employ Y number of people in the United States -- a kind of auction. Any non-lottery system has biases, and our present one is no exception.

The auction idea is still long on theory, short on details. Would there be one auction for all visas, requiring sugarcane workers to bid against engineers, or would the labor market be segmented on some basis through separate auctions? Would the government set a minimum or maximum price, or would it permit the market-clearing price to apply? Would the government (or would-be employers) assist poor migrants to finance their bids? How would repayment be enforced? How would the government prevent employers from cartelizing the bidding process? Would the auction be introduced at once, or phased in? Would the auction proceeds be used for retraining natives or simply go into general revenues? Questions like these must be answered before a visa auction could command much support outside the economists' seminar rooms.

An auction tailored to meet non-economic goals and satisfy practical and moral objections might well be an attractive policy option. Any auction should cover only skills-based admissions, not refugee or family-based ones. The ability to bid should not be limited by wealth disparities; adequate loan and "scholarship" funds from public and private sources must enable poor workers with good prospects to finance their bids. Even with these safeguards, however, supplying credit for human capital would involve special problems, as we have found with the financing of higher education: loans taking more than a generation to repay might have to be subsidized; once here, aliens would be hard to trace; poor people might be unable to satisfy lenders that they have qualities conducive to economic success. Any auction system would surely exclude some workers whom we would want -- poor people who would learn fast and rise, potential entrepreneurs, diamonds in the rough.

There is a risk that any scheme for attracting skilled foreign workers would ease pressures to invest in improving the systems for educating and training natives, but this fear seems exaggerated. Native workers' skills will always be crucial to our economic strength; long-term strategies for developing them should not be diverted by volatile short-term migratory patterns. Newly expanded state and local education programs attest to the powerful political incentives to invest in native skills. Competition from newcomers, moreover, should spur young Americans to stay in school and upgrade their skills so they can share in the economic rewards. Still public funds, employer fees, or some portion of auction proceeds should be earmarked for training native workers, as with adjustment assistance under trade laws.

An Emerging Consensus

For all the heavy normative loading of immigration policy, the current debate in Congress is remarkably pragmatic -- less a struggle of fundamental ideals than a dispute about how best to balance numbers, categories, and enforcement methods. In this debate, the growth goal stressed by economists, while undeniably important, is far from decisive. A broad political consensus has been reached on three admissions principles that were bitterly contested not so long ago: (1) some increase in immigration levels; (2) universalistic criteria with respect to source country and humanitarian admissions; and (3) more labor-oriented criteria. Expansionists as well as restrictionists firmly couple these principles (especially the first) to a demand for stepped up enforcement against illegal aliens.

Some Increase in Immigration. Expansionists, recalling our long history of discrimination against aliens, often imply that restrictionists and status quo advocates either have, or seek to exploit, racist attitudes. As evidence, they cite the popular support for official English and increased border enforcement, as well as the (largely unsuccessful) opposition to bilingual education and to welfare for illegal aliens.

These disputes, however, do not demonstrate a resurgent bigotry. Despite many incidents of ethnic tensions, racial bias has declined steadily over the past decades as educational levels have risen. Most Americans, surveys also reveal, admire immigrants whether they be Korean vegetable dealers, Indian physicians, Filipino computer specialists, Ethiopian cab drivers, or Dominican nurses. Americans plainly adhere to the melting pot myth, believe that the U.S. can absorb and uplift hard working, family-oriented immigrants struggling to get ahead, and ascribe these positive qualities to aliens, even illegal ones. Recently, Californians (along with voters in several other states) voted by overwhelming margins to amend their state constitution to make English the "official language," despite widespread sympathy for immigrants' efforts to retain their native languages and customs. Political scientist Jack Citrin, reviewing the controversy, finds that most of the support for official English cannot be attributed to economic anxiety or anti-minority bias. The amendment, Citrin points out, was favored by 62 percent of respondents who identified themselves as strong liberals and 65 percent of those who called themselves strong Democrats. (Whether the amendment represents sound policy is of course a different question. I have some doubts).

Given the strong nativist sentiment in the United States prior to World War II, one is struck by its absence from the policy debates today and during the decade before passage of the 1986 immigration law. The Federation for American Immigration Reform (FAIR), perhaps the most prominent restrictionist group in recent years, exemplifies this point. Led by long-time activists in the environmental, population control, and labor movements such as former Colorado governor Richard Lamm, FAIR's positions are animated by those traditional liberal concerns, not nativism. It does not oppose immigration in principle but fears instead that increased immigration will threaten our resource base and labor standards. It wants illegal migration reduced before legal admissions are expanded further.

The current efforts of Congress to reform legal immigration also confirm this robust pro-immigration consensus. The main dispute today is not about whether levels should be increased; it is about how much larger the number should be and which categories of immigrants should be favored. Each of the leading bills now pending in the House and Senate would raise the levels significantly, the House bill far more so. This consensus is most impressive when one recalls how high immigration levels have already risen in recent years. Indeed, the current level is already close to the historic peak set in the century's first decade, and it is almost 20 percent above the 1984 level, 40 percent above the pre-Mariel (1979) figure, and 60 percent above the levels that generally prevailed until 1977.

There are many reasons for this consensus. The mythology of the melting pot and the iconography of the Statue of Liberty seem to grow stronger over time. Immigrants' extraordinary contributions to American society are evident in almost every community. Americans recognize that many immigrants possess the strong family and entrepreneurial values that we admire. Their work and consumption help to sustain many industries and communities. American employers like them for their productivity, docility, and downward pressure on wage rates. Ethnic groups have become publicists in their own right, pressing effectively for more admissions.

Countervailing pressures exist, of course. In addition to the kinds of environmental and labor concerns that FAIR voices, some people fear that immigration is causing cultural fragmentation, urban crime, and demand for welfare and local public services. In some recent episodes, blacks and Puerto Ricans mired in poverty have lashed out at recent immigrants, sensing competition for scarce jobs, housing, and social services. Many more people presumably share these fears but will not speak of them in polite company Whether or not the fears are justified (Borjas refutes some of them) is less important than the fact that they are commonly felt.

These fears partly explain why strong public support for high levels of immigration is qualified by a demand for firm limits. Those limits, the public believes, should be legally defined rather than dictated by the vagaries of geopolitics or the incentives of outsiders. The support for immigration is only as strong as the public's perception that these limits are being enforced -- a political fact that expansionists ignore at their peril.

This tradeoff is relatively new. Until the end of the Bracero program in the mid-1960s, illegal employment in the U.S. was relatively low and largely confined to the rural Southwest. Since then, the growth of a large, urban illegal population has dominated immigration politics. Americans believe that those who work here in defiance of the law and who are impervious to public control pose special social problems, and they expect the government to do something about them. According to polling data, Hispanics and many blacks tend to share these views, while their leaders often do not. This schism has caused organizational conflict over whether to seek repeal of employer sanctions.

If nativism does not fuel this demand for limits, what does? This might seem an idle question. AU industrialized societies, after all, seek to control which strangers may enter, work, and live within their territory. Even African states, traditionally most hospitable to border crossers, have begun to crack down on them. Still, the U.S.'s proud self-conception as a nation of immigrants and its liberal commitment to inclusiveness impose a special burden to justify whatever limits it decides to impose.

Americans' demand for limits, I believe, springs from a deeper anxiety than the ones discussed earlier. The master theme of immigration politics is the fear that we are losing control of our way of life. We seek to relieve this anxiety by focusing on things, like immigration, we think we can control. Virtually all participants in the long debates about the 1986 reforms managed to agree on one slogan, continually recited as if it were an incantation: 'We must regain control of our borders." This slogan was intended to conjure up a memory of a past golden age when we actually exercised control, an age that of course never existed. It also exalted a compelling ideal according to which the nation deliberately chooses whom it wants, excludes those whom it does not want, and sanctions those who violate its rules. The Border Patrol's enormous recent growth testifies to the ideal's evocative force.

Expansionists should not dismiss this popular demand for border enforcement as a macho fantasy about decisive government, impregnable territory, and firm control. Individuals, families and tribes define and defend their turf, not just nations. They may believe, with Frost, that good fences make good neighbors but Michael Walzer stresses a more potent motive: the same fence that keeps most people out is necessary to enable those within to think of themselves as co-venturers and to flourish as a community.

Immigration threatens Americans' sense of control by seeming to jeopardize three fundamental values: national autonomy, economic security and the "social contract" that secures the welfare state. Because each of these concerns has some basis in fact, public demands for tougher laws limiting who can enter, work, and claim welfare benefits seem plausible. Past reforms have not succeeded in allaying the public's anxiety.

Traditionally, immigration policy was designed to enhance the sovereign autonomy of the United States at the expense of all other values, and the courts interpreted the Constitution accordingly. Today, our power to work our will in the world is eroding. Immigration exemplifies this erosion. When Fidel Castro decided to expel his opponents and empty his jails and hospitals, he transformed the face of south Florida forever. When Mexico devalued the peso, it propelled tens of thousands of its people north to Houston and Los Angeles in a desperate search for dollars. When gunfire erupted in Tiananmen Square, industry and engineering schools throughout the United States had to reassess their plans. When Violeta Chamorro won her election, thousands of Nicaraguans in the U.S. instantly became deportable. When Thailand threatened to close refugee camps, we had to accept more Kampucheans and Vietnamese for resettlement in the U.S.

These constraints on our autonomy are certain to increase. For much of its history -- and during the period when our great national myths were forged -- the U.S. not only lived in splendid isolation from the world's turbulence but ascribed moral virtue to that isolation. To lose this insularity and innocence is painful. Insofar as immigrants symbolize that loss, they are natural targets for our resentment.

Immigration also threatens Americans' sense of control over their economic security These impacts are more politically explosive despite (or perhaps because of) the difficulty of measuring them. That immigration on balance creates jobs and enhances economic growth can no longer can be seriously doubted. For politicians the hard issue is distributional: who gains and who loses? Some economists, like Simon, come close to saying that there are no losers at all, at least in the long run. Others see other immigrants as the major victims. David Card's study of how the large Marielito influx affected their native competitors in the Miami labor market suggests that the newcomers adversely affected the wages or jobs of other Marielitos, not low-income natives. Borjas takes a similar view.

These studies, however, examine aggregate, long-term, equilibrium effects. It is still possible that immigration leaves particular worker groups with big losses, and that immigrant concentrations in urban areas block natives from moving there for better jobs just as Southern blacks did, for example, when immigrant declined after World War 1. Local politicians (which in the U.S. means almost all) are understandably more concerned about short-run losses concentrated on their constituents than about long-run gains by larger political units or the nation as a whole.

Immigration also threatens the social contract, the communal commitment that supports a complex structure of welfare state entitlements. This commitment rests on a perception that those entitlements are not being abused, that they are kept "within the family," which in turn requires public confidence that national borders are under control. The ethos of social altruism undergirding the welfare state grows weaker as its beneficiaries become more numerous, alien, and unassimilated to social norms. Politicians in Los Angeles see the city's public hospitals being overwhelmed by aliens, many illegal. They are unlikely to be impressed by studies showing that the average immigrant family is only slightly more likely than its native counterpart to claim welfare benefits. Had the amnesty program enacted in 1986 not limited legalized aliens' access to welfare benefits for five years, the entire legislation might well have been defeated. Expansionists, in short, must respect the need for real limits on immigration unless they are prepared to support real limits on welfare state entitlements.

Universalistic Criteria. One might have expected the 1%5 law's transformation of immigrants' racial, ethnic, and national mix to arouse a powerful political backlash. Immigrants from the newly-favored source countries were not yet sufficiently numerous to secure the policy politically, while immigrants from those countries disadvantaged by the change were already here and presumably voting. Yet no serious effort to repeal the law's more universalistic system has been mounted. And as the new immigrants naturalize and then organize politically to defend their group interests in the status quo, it will become even more firmly entrenched.

This commitment to source country universalism has not gone unquestioned. Tip O'Neill and Ted Kennedy won a limited exception in the form of a special program making 20,000 additional visas outside the preference system on a lottery basis to people from countries disadvantaged by the 1965 reform, notably including Ireland, and pending legislation would make that program permanent. Similarly, the point system in pending bills would reward those who speak fluent English, thus helping would-be immigrants from the Commonwealth countries. Much the same story can be told about the 1980 Refugee Act. As we have seen, a refugee admission system explicitly favoring certain geographic areas and political ideologies was supplanted by a universalistic one, only to have administrative practice and statutory amendment restore some of the earlier parochial biases.

Although universalism has eroded in the areas of immigrant and refugee admissions, it still has important practical effects. Both the 1965 and 1980 laws contain a strong presumption of universal equal treatment that proposed exceptions must strain to overcome with strong, public justifications. Many courts, for example, have upheld this presumption on behalf of Salvadoran asylum claimants despite administration opposition. Although there are real risks that the system will unravel under special interest pressures, the exceptions so far have been narrow enough to preserve the normative vitality of universalism.

More Labor-Oriented Criteria. So long as the demand for immigrant slots greatly exceeds the supply, we must choose among many plausible admission criteria. Today the main ones are family relationships, employers' manpower needs, and refugee protection. Within each of these categories, moreover, subgroups must compete with one another. Tradeoffs between legal immigrant slots and the perceived volume of illegal entrants, and among overseas "Convention" refugees, other refugees, and domestic asylum-seekers are inescapable. Within the refugee category, different source regions compete for preference; within the family category, different relationships compete for priority.

The family's weight on the immigration scales reminds us that the economic models are radically incomplete. No Congress whose only or even main goal was economic growth would have passed (or retained) the 1965 law, the Refugee Act, or the 1986 law. Even now, Congress is likely to reduce only marginally the strong policy commitment to family and (to a lesser degree) humanitarian values over labor market goals. The reason seems clear: Congress represents social values, and society views certain non-economic values as trumps.

Although economics cannot prove that these values are wrong, it can help us estimate what they are costing. Then we can decide how much we want to continue adhering to them. Borjas calculates that U.S. income is at least $6 billion lower each year than it would be if those who immigrated here in the late 1970s were as skilled as those who came in the early 1960s. That is money worth worrying about, of course, but a society with a (slowly) growing $5.5 trillion economy can afford to sacrifice some wealth for other goals. Clearly, ours has chosen to do so.

To recognize the costs of our policy is certainly not to endorse the status quo. In a world where tourist visas for family members are plentiful and international travel cheaper, reducing family-based admissions in favor of skill-based ones seems justified. The labor preference system can surely be improved. Point systems and auctions designed to promote a mix of social values might help accomplish these goals. For example, the government might siphon off from employers some of the labor visas' scarcity value, using it to support job training programs for domestic workers. Visa lotteries, Congress's current policy fad, should be avoided. They simply abdicate our obligation to choose in favor of a spurious fairness.

Expansion would ease the choice among categories by increasing the number of family, humanitarian, and skill slots. We cannot accomplish this, however, unless we address the public's real anxieties about autonomy, jobs, and welfare. Long before economic or cultural limits to expanded immigration come into play, I suggest, these political limits win arouse opposition. The goal should be to extend these limits. In doing so, we must not capitulate to these fears. Instead, we must treat them seriously and with respect. Strategies to build support for expansionist policies should address some of the sources of concern.

Six broad themes should shape these strategies. First, we must recognize the interplay between higher legal immigration ceilings and lower illegal migration. Solutions here will be difficult, as most politically viable options have been tried without much success. Second, we must secure the legitimacy of asylum by making it a speedy, reliable remedy for genuine refugees, not a dodge for resourceful aliens and lawyers. Newly adopted procedures should improve accuracy and fairness but also neglect (and may worsen) asylum-related enforcement delays resulting from a cumbersome appeals process. Third, we must try to reduce welfare dependency among immigrants. Borjas' findings in this regard should trouble any friend of immigration, for it is bound to strengthen restrictionists' hands. Fourth, we should encourage easier, faster naturalization, which other nations such as Canada have achieved, to enhance migrants' political influence and assimilation. Fifth, we should encourage English language mastery. This must be done without either enforcing cultural conformity or acquiescing in further social fragmentation. Reasonable people differ about how this can best be achieved but there should be no doubt or apologies about its urgency. Finally, any expansion of immigration should be accomplished in incremental steps to provide ample opportunity to reconsider and fine-tune policy, if, as has often been true with immigration reforms, the unforeseen occurs.

Most of these themes are controversial; policies to implement them will be even more so. Between the twin follies of open borders and restricted or status quo immigration lies a sensible policy of moderate, staged expansion. The many Americans who will gain inestimably from this policy should carefully attend to the needs of the relatively few who will disproportionately bear its burdens.