United States policy toward Mexican immigration is suddenly on the table again. The Bush administration is talking about a new amnesty for undocumented Mexicans living in this country, as well as an expanded guest-worker program. One suspects that these initiatives have more to do with wooing Latino voters than with any recognition that the current U.S. approach to immigration is fundamentally flawed. Nonetheless, the administration's tactic creates an opportunity to reconsider the issue, and Congress should seize it.

By militarizing our southern border, penalizing employers who hire undocumented workers, barring immigrants from entitlement programs, and generally making life unpleasant for foreigners in this country, U.S. policy makers have sought to discourage immigration. But nothing of the sort has happened. Instead, these policies have resulted in the deaths of hundreds of Mexicans and the exploitation of thousands more, while undermining the wages and working conditions of U.S. workers. They have transformed a seasonal movement of migrant labor in three states into a settled population of immigrant families dispersed throughout the country. They have pushed half a million Mexicans to become naturalized U.S. citizens, setting the stage for even larger migrant flows in the future, as each new citizen creates a host of entitlements for the entry of relatives. And they have wasted at least $3 billion a year on border-control efforts that have achieved no discernible reduction in the flow of undocumented immigrants. In short, current U.S. policy has produced the worst of all possible worlds: continued, growing Mexican immigration under conditions detrimental to the United States, Mexico, and the immigrants themselves.

This mess is what comes of basing policy on the Cold War hysteria and economic panic of the early 1980s rather than hard facts. The fantasy is a tight border; the reality is North American integration. In the North American Free Trade Agreement (NAFTA), the United States committed itself to the continent-wide integration of markets for goods, capital, information, commodities, and services. But policy makers have refused to recognize that inevitably labor markets will also merge in an integrated economy. In reality, Mexican immigration cannot be stamped out. But it can and should be cooperatively managed, just as the flow of goods or capital across the border is managed, to maximize its benefits and minimize its costs on both sides. However unwittingly, the Bush administration is giving Congress a chance to finally get that right. We would like to suggest how it can.

Empirical research over the last several decades has provided the facts that policy makers need to face. First among them is this: Contrary to common belief, international migration does not arise from a lack of economic development, but from development itself. As industrialization spread across Europe from 1800 to 1929, it triggered waves of emigration in country after country. Likewise, the poorest and least-developed nations do not send the most international migrants today. The fact of the matter is that no nation has undergone industrialization without a massive displacement of people from traditional livelihoods, which are mainly located in the countryside; and in most cases a large fraction of these people have ended up migrating abroad.

A second basic truth is that immigration naturally follows other international links. When the upheavals of market creation and industrialization occur, those who emigrate do not scatter randomly, nor do they necessarily head for the nearest wealthy society. Rather, they tend to go to places to which they are already connected--either economically, through the trade and investment relations between countries; politically, as a result of treaties, colonial administration, or military occupation; or socially, through the sustained contacts created by overseas military deployments, student-exchange programs, diplomatic missions, tourism, trade, and so on.

Third fact: low-skilled immigrants generally enter a developed country like the United States in response to the strong and persistent demand that is built into the structure of postindustrial economies. Owing to shifts in the technology of production, the emergence of the welfare state, and the pervasiveness of social hierarchies, labor markets in developed nations have become increasingly segmented. There are "good" jobs attractive to natives, and there's a separate realm of poorly paid, "bad" jobs that natives shun. To fill the latter, employers turn to immigrants; in fact, they often directly recruit them.

As far as immigration from Mexico is concerned, these three realities point to only one conclusion: It is going to continue. We are talking, after all, about two countries with more than just a 2,000-mile border in common. Until 1848 the states of California, Arizona, New Mexico, Nevada, Utah, and Texas, as well as portions of Wyoming and Colorado, were part of Mexico. In 1853 the U.S. annexed additional Mexican territory, after which it invaded the country three times, most recently in 1917. Since then, bilateral trade has grown to $200 billion per year, and the border is now marked by a string of rapidly growing twin cities and a booming manufacturing sector. The two nations also share a 60-year history of uninterrupted migration. The Mexican-born population of the United States now stands at eight million. Annual legal immigration exceeds 130,000 persons per year (more than 110,000 of whom are immediate relatives of people already here). Another 3.4 million Mexicans legally enter the country each year for long visits, as tourists, exchange students, businesspeople, and such; and 213 million drive back and forth across the border each year, also entirely legally, for short visits. In addition, the Immigration and Naturalization Service (INS) estimates that net annual migration of undocumented immigrants is around 150,000 persons, with a total of three million undocumented Mexicans now living in the United States.

But if the international flow of workers is inevitable, other empirical data suggest how it might be better managed. Our own research, for instance, which involved interviews with close to 5,000 immigrants over the last 15 years, shows that when migrants first enter a developed country, most are not motivated, as Americans tend to imagine, by a desire to live where they can maximize earnings. In reality, most migrants move in an attempt to solve economic problems at home. They are leaving countries with inadequate capital, credit, and mortgage markets--countries also without unemployment insurance, crop insurance, futures markets, and other protections against sudden losses of family income. Initially, most migrants seek to work abroad only temporarily, and mainly in order to diversify family-income risks, to accumulate cash unavailable at home (most often so they can build a house), or to finance a small business or a consumer purchase (such as a car).

In other words, international migration is often less influenced by conditions in labor markets than conditions in other kinds of markets. So U.S. policies, which to date have chiefly aimed to make the American labor market less attractive to Mexican migrants, have been of little use. Programs designed to improve the performance and accessibility of a sending country's credit and insurance markets are far more likely to affect migration decisions.

Policy, however, is a matter of responding not simply to facts but to political interests, and the politics of immigration is complicated. Within the United States, owners of capital assets (including such human capital as high skills and education) generally benefit from international migration, whereas ordinary workers benefit only to the extent that the immigrants themselves are highly skilled and their participation in the economy creates job opportunities for less-skilled Americans. To the extent that immigrants are unskilled--like the overwhelming majority of Mexican immigrants--their chief effect on ordinary U.S. workers is a downward pressure on wages and employment rates.

Thus, the country's elected leaders--who in theory should seek to optimize the welfare of the largest number of citizens but in practice mostly seek to satisfy, or at least not antagonize, the citizens most important to their re-election--are left weighing the interests of workers against those of asset owners. And while the latter may be more economically and politically powerful, the former are more numerous. This yields a politics of immigration that varies strongly with the business cycle.

In comparison, the political interests of Mexico are more straightforward. There, the clearest beneficiaries of emigration are the migrant households themselves, and the principal interests of the government are to protect citizens working abroad, to maximize the flow of money repatriated, and to use those funds optimally to promote economic development. Still, these interests are not always easy to reconcile with shifting U.S. considerations.

Moreover, whatever policies are ultimately adopted should be morally defendable, which in our view means that they must be consistent with five basic principles.

  • People should be free to leave their country of citizenship for whatever reason without state interference.
  • Once migrants are accepted into a receiving nation, they should be given full legal rights.
  • Although clandestine migration cannot be eliminated, reasonable steps should be taken to minimize the number of people living and working in irregular statuses.
  • Sending nations should guarantee reciprocal rights of entry and legal protection to citizens of receiving societies.
  • And finally, immigration policies should not be imposed unilaterally but developed cooperatively through multilateral agreements.

    The challenge now before U.S. policy makers is not just to take account of long-ignored realities but also to address political interests and respect ethical imperatives. We have tried to reflect all of that in the following proposals.

    Our first proposal is to change the U.S. immigration quota for Mexico. When immigration is inevitable and necessary, it makes no sense to let Mexican demand for entry so far outstrip the supply of immigrant visas. Currently, just 20,000 visas for persons without immediate family here are allocated to Mexico each year, though the country has a population of 100 million. This is the same number of visas as for the Dominican Republic, with its population of only 8.2 million. In other words, the Dominican Republic is allowed a much higher rate of legal emigration, even though Mexico, as a partner in NAFTA, is far more closely linked to the United States and more vital to its interests. The time has come to increase Mexico's absurdly low quota to a figure closer to what the immigrants actually need.

    But our research shows that not all immigrants do want to move here permanently. To a great extent, Mexicans seek permanent-resident visas because that is the only door open to them. If another option were available--such as a temporary-worker program--more people would opt for it. Indeed, an enlightened policy, rather than making it difficult for migrants to come and go, would support their evident desire to return home by granting them temporary-work visas that allow them to enter, leave, live, and work in the United States without restriction for a period of, say, two years. We propose such visas as the mainstay of U.S.Mexican-immigration management. They would be renewable once in the lifetime of the migrant, but only after the worker returned home for at least a year. A binational agency managed by the U.S. and Mexican governments, to which aspiring migrants would apply directly, would distribute the temporary-worker visas, thus getting employers out of the corruption-prone business of labor recruitment. And if these visas were generously available, they would go a long way toward reducing undocumented migration and the ills that accompany it.

    In order to guarantee labor rights, these visas would not be tied to specific employers or jobs but issued directly to the migrants themselves. Whenever a work visa is tied to a particular job, as in the current guest-worker program, the migrant is left vulnerable to exploitation, unable to exercise the most fundamental right a worker can have: the right to withhold his or her labor. Granting migrants the right to change jobs would free them to participate in unions and to report violations of their labor rights. It would also make it more difficult for unscrupulous employers to use immigrants to lower the wages of native workers or cut corners on ensuring their health and safety.

    If something on the order of 300,000 two-year visas were issued annually, the temporary migrants working in the United States at any time would constitute only a small share of the U.S. workforce but a large fraction of those who presently migrate with permanent visas they don't really want or with no papers at all. What's more, we have historical evidence of the efficacy of such programs. In the 1950s, the United States issued 450,000 temporary visas annually to Mexicans--and reduced undocumented migration to near zero.

    With this kind of temporary-worker program in place, there would be no need for employer sanctions, which in practice have done nothing to reduce undocumented hiring and much to drive down wages and erode working conditions for immigrants and natives alike. We therefore propose that Congress repeal the sanctions and, as a final step in eliminating this underground economy, regularize the status of undocumented migrants who have a record of peaceful, long-term residence in the United States. These people are already present, so "legalizing" them would have no effect on population growth; it would simply facilitate the assimilation of these immigrants and their citizen children into U.S. society.

    The Bush administration presumably has nothing so far-reaching in mind when it talks about a new amnesty and an enlarged guest-worker program, but why shouldn't both be designed to accomplish what an integrating continent needs? Enacting our proposals would even provide U.S. authorities with an opportunity to raise funds that might be used to offset the costs of international migration and to help Mexico on its road to development. Temporary migrants could be charged a $300 fee for each visa, to be paid to the U.S. Treasury at once or in installments. We already know that Mexicans are more than willing to pay this amount to gain access to the United States; it is less than the current cost of renting a coyote to smuggle them across the border. And a $300 fee paid by 300,000 temporary migrants would yield both greater safety and annual revenues of $90 million.

    Another potential source of revenue is the federal taxes withheld from the paychecks of temporary workers. If we assume, conservatively, that the 600,000 temporary migrants in the country at any one time earned annual incomes of $15,000 and had payroll taxes withheld at a rate of about 15 percent, the resulting revenue stream would be $1.35 billion per year.

    Further revenues would be freed by a drastic reduction in the personnel and resources devoted to border enforcement. And this we highly recommend, regardless of what other changes are adopted. In a forthcoming book, we examine the declining efficiency of INS border-control operations: As they've expanded, they've driven migrants into parts of the desert where they are essentially uncatchable (as well as greatly endangered). We estimate that a much smaller investment in INS operations would yield the same degree of deterrence as is currently achieved, saving taxpayers at least $3 billion annually.

    In total, then, about $4.4 billion a year could become available. The bulk of these funds, we suggest, should be earmarked for states with large immigrant populations, because immigration does impose substantial costs on them. Indeed, the number of foreign-born residents in each state should be included in the basic formula for federal revenue sharing. By transferring sufficient resources to the states, the federal government could do much to assuage the sort of anger and resentment against immigrants that surged in California in the late 1980s and early 1990s.

    The remainder of the funds, however, should be used to facilitate the improvement of markets and the social infrastructure in Mexico. The integration of Spain and Portugal into the European Union offers a successful model. During the 1960s and 1970s, hundreds of thousands of emigrants left those two nations for work in the wealthier countries of northern Europe (particularly Germany), and northern officials worried that admitting Spain and Portugal into the European labor market would unleash even larger waves of emigration. But in preparation for their integration into the union, EU members invested substantial funds in Spain and Portugal. Transportation, communication, banking, and social-welfare networks were all upgraded. As a result, when unification finally occurred, in 1986, there was no rise in emigration. On the contrary, Spain and Portugal immediately began to experience large net return migrations, despite the fact that per capita income in both countries was (and still is) well below that in Germany.

    Similar investments are likely to have similar effects in Mexico. For example, a binational insurance provider might be established, through which Mexicans could purchase low-cost protection against income losses of various kinds, giving them a means of risk management other than migration. A binational development bank might offer matching grants to Mexican communities for the construction or improvement of local infrastructure, thus multiplying the benefits of the $3.6 billion in "migradollars" that Mexican workers now send home from the United States annually.

    Particularly useful would be an extension of banking services to Mexicans of modest means. This would solve two persistent problems that have plagued Mexican immigrants--the high cost of sending money home (the few firms that dominate the market gouge their customers for as much as 20 percent of the value of their international transfers) and the unfavorable rates of exchange they receive in Mexico. And expanded banking services would also go a long way toward improving Mexicans' access to capital and credit at home, so they wouldn't have to resort to international migration just to finance a house or an education--or a refrigerator. A poor people's bank could establish dollar-denominated accounts to protect migrants against instability in exchange rates, and peso accounts could offer above-market rates to attract funds. The bank could use these deposits to make loans to families seeking to finance homes, businesses, or schooling. Low-rate, low-balance credit cards could be issued so people could finance big-ticket consumer purchases.

    Even taken together, the measures we've proposed here would not eliminate undocumented migration or solve all the problems of an integrated labor market. They would, however, reverse the deleterious consequences of current U.S. policies. They would also promote the long-term economic growth and political stability of Mexico--and that must be our overriding goal in managing U.S.-Mexican immigration. For there's one last fact that should inform U.S. policy: Immigrant flows do not last forever. Historical data indicate that most European nations in the course of economic development underwent an "emigration transition" from low- to high- to low-emigration rates. In Europe this process took eight or nine decades, but recent experience suggests that the transition time is now considerably shorter. Take South Korea. In 1965 it ranked among the world's poorest nations; by 1998 it was one of the wealthy industrialized nations. In the interim, some 780,000 Koreans emigrated to the United States. Yet by the end of South Korea's transition, gross emigration was only 13,000 a year and net migration was near zero. The goal of U.S. policy should be to help Mexico move as quickly as possible through its own emigration transition and assume its place as a full and equal partner in the North American market.