Agenda for Shared Prosperity

March 15, 2007 | EPI Briefing Paper #186

Getting Immigration Reform Right

by Ray Marshall

Download print-friendly PDF version Acrobat (PDF)

See also Immigration for Shared Prosperity (an EPI book, April 2009)

Congress' difficulty in passing immigration reform legislation comes as no surprise to those who have followed this issue over the years, especially the debates that led to the seriously flawed Immigration Reform and Control Act (IRCA) of 1986. Many of the factors that caused IRCA to fail are as prevalent now as they were in 1986. Diverse economic interests, personal biases, and political ideologies make it hard to build consensus for effective immigration policies. These complications are exacerbated by the absence of reliable information about the magnitude of unauthorized immigration and its impact on the American economy and society. Unlike many other policy issues, there are no clear political alignments on immigration, making it difficult to build the coalitions needed to align the complex components of a successful immigration policy.

By the time IRCA was amended enough to pass the Congress, it became very clear to immigration experts that, instead of restricting their entry, IRCA would accelerate the flow of unauthorized immigrants into the United States, which is exactly what happened. Common estimates of the number of unauthorized immigrants in 1986 were between 3 and 6 million; today, estimates range from 10 to 20 million. The networks that give employers a dependable supply of unauthorized immigrant labor are much more institutionalized and difficult to control. If the United States does not get policy right this time, 20 years from now the number of unauthorized immigrants probably will have at least doubled and be even more difficult—if not impossible—to control.

That said, however, immigration is not the problem: the United States is and will remain a nation of immigrants, who have contributed greatly to the vitality, diversity, and creativity of American life. Immigrants are particularly important to the U.S. economy, accounting for over half of the workforce growth during the 1990s and 86% of the increase in employment between 2000 and 2005. Because there will be no net increase in the number of prime-working-age natives (aged 25 to 54) for the next 20 years, the strength of the American economy could depend heavily on how the nation relates immigration to economic and social policy.

Unauthorized immigration, on the other hand, subjects migrants to grave dangers and exploitation, suppresses domestic workers' wages and working conditions, makes it difficult to adjust immigration to labor market needs, perpetuates marginal low-wage industries addicted to a steady flow of unauthorized immigrants, is unfair to people waiting to enter the United States legally, and undermines the rule of law. The issue is not immigrants, but their legal status, characteristics, and integration into American life.

Because of its importance to America's diverse and rapidly growing Hispanic population, immigration also has significant political implications. Hispanics' political power is enhanced by their geographic concentration in areas where Democrats and Republicans must contend for national dominance, especially in the Southwest and Rocky Mountain West. This reality was an important component of the political strategy fashioned by George W. Bush and Karl Rove. During his first term, President Bush courted Latinos with a strategy that included speaking Spanish, Hispanic appointments to prominent positions in his administration, and an immigration policy that included a guest worker program championed by Mexican president Vicente Fox. The Bush-Rove strategy was derailed by nativist Congressional Republicans, who adamantly opposed comprehensive immigration reform in favor of exclusive reliance on border security. As Bush and Rove feared, nativist elements in their party provided strong Hispanic support for Democrats in the 2006 elections, as they did in California under Republican governor Pete Wilson. Indeed, resentment toward the nativist pronouncements of anti-immigrant groups is one of the few unifying issues for America's diverse Latino population.

Because of deep international economic and demographic integration, immigration has important foreign policy implications, especially for U.S. relations with Mexico, the source of most unauthorized migrants to the United States. In fact, for many years, Mexican policy has been based on the expectation of heavy migration to the United States. In the 1970s, for example, Mexican foreign minister Jorge Castaneda (the father of former President Vicente Fox's first foreign minister) told us that, whatever we did, the United States would absorb a large part of Mexico's population growth. Many of us who were attempting to formulate policy for the United States did not want to believe that we would have so little control of immigration, but he was right.

Migration clearly is very important to Mexico: it provides a safety valve to compensate for that country's failure to provide adequate domestic jobs for most of its workforce growth, and remittances from the 20 to 25 million Mexicans living in the United States have become second only to oil exports as a source of Mexican foreign exchange. Remittances also are the lifeblood of many rural communities and supplement that country's weak social safety nets. Given Mexico's slow growth and serious structural problems (poverty and inequality; corruption; low tax collections; poor education system; ineffective political checks and balances; inadequate infrastructure development; restrictive business regulations; rigid, antiquated, and inefficient labor market policies and institutions; and the limited capacities of governments at every level), it is unlikely that its citizens will have adequate job opportunities at home anytime soon. What the United States does about immigration, therefore, has important implications for Mexican economic and political developments, with significant positive or negative spillover effects for America.

Since past mistakes can provide lessons for more effective future policies, this report will first explore the reasons for IRCA's failure, including some common myths about unauthorized immigration. This report concludes with an analysis of a comprehensive mix of policies that could serve the best interests of the United States and other countries, especially Mexico.

IRCA's defects

IRCA's main technical defect was that it did not include a secure worker identity or work authorization system, without which all other control measures were less effective and often counterproductive. This reality was well known to participants in the immigration policy debates—both those who wanted tighter controls (who lost the legislative contest) and those who favored relatively open migration (who won). In connection with their work for the 1979-81 Select Commission on Immigration and Refugee Policies (SCIRP), Labor Department experts developed a work authorization process for new hires and job changers that would have made a federal agency, not employers, responsible for verification; the employer's only obligation would have been to verify an identification number the applicant obtained from the federal work authorization agency. Because of opposition from an alliance of open immigration advocates and civil libertarians worried about a national identity card, IRCA opted for an array of easily counterfeited identifiers, permitting a fair amount of fraud, especially in the Act's employment and adjustment-of-status programs, thus accelerating the flow of unauthorized immigrants. IRCA also gave employers responsibility for verifying work authorization documents, a task they had neither the ability nor the will to perform.

To understand why employers lacked the will to screen unauthorized applicants, it is necessary to examine the magnetic relationships between them and unauthorized immigrants. For hard-to-fill jobs, employers often prefer unauthorized immigrants to authorized residents. This preference is due not only to immigrants' willingness to accept lower wages, but also because they are a more dependable supply of labor for these jobs and, because of their limited options, are less likely either to leave or complain to government officials about abuses. Very effective informal immigrant information and support networks give employers a dependable supply of labor. Since 1986, these networks have been strengthened by the spread of relatively inexpensive information technology, especially cell phones and radios. On the workers' side of the employment relationship, jobs which are unattractive to natives not only are much better than those available in their home countries, but also provide a measure of security for immigrants and their families, despite their unauthorized status.

These networks are strengthened and perpetuated by community support groups, home country officials, and employers' investment decisions. Once institutionalized, these bonds are very hard to break and tend to exclude natives from the process.

Myths strengthen the networks

These tight employer-immigrant relationships are reinforced by public attitudes and myths, the most prominent of which is that immigrants only fill jobs Americans won't take, an attitude encouraged by employers, immigrants, and their foreign and domestic supporters to justify unauthorized immigration. The truth is that there are no such occupations: according to the Center for Immigration Studies (CIS), of 473 occupational titles, only four (stucco masons, tailors, produce sorters, and beauty salon workers) have immigrant (authorized and unauthorized) majorities, and natives hold over 40% of the jobs in these occupations. Like most enduring myths, this one has an element of truth: natives with more options are less willing to take these jobs. But, as noted, once the strong employer-immigrant bonds are established, it is hard for even willing natives to compete for these jobs, thus appearing to confirm the myth.

Those who perpetuate this myth ignore other options that can, and have been used as an alternative to the employment of unauthorized immigrants, including actively recruiting authorized residents; improving management (which often is very bad in low-wage occupations, where the costs of inefficiency are transferred to workers through such practices as piece rates); introducing technology to improve productivity, as was done in California agriculture after the end of the bracero program in 1964; or, obviously, improving wages, benefits, and working conditions.

Another popular misconception is that unauthorized immigration is really not so bad because its negative impacts on natives are small and it improves the competitiveness of the American economy. Again, there is enough truth to this argument to give it superficial plausibility. There are, however, several problems with equating the economic effects of unauthorized and authorized immigration, as some analysts do. For example, studies of the impact of refugees—who are authorized residents, usually with more human and financial capital—have been cited as evidence of the beneficial effects of unauthorized immigration. Similarly, authorized immigrants, who tend to have both lower and higher levels of schooling than natives, cannot be equated to unauthorized immigrants with little or no formal education. It is significant that, controlling for other things, authorization improves immigrants' wages.

Economists disagree about the impact of unauthorized immigration on American workers. Some find little or no negative impact,1 while others report large and significant effects. For example, one widely cited study found that for the nation as a whole, between 1980 and 2000 immigrants (authorized and unauthorized) reduced the wages of high school graduates by over 8%, college graduates by almost 4%, and all workers by over 3%.2

A resolution of this controversy is beyond the scope of this paper, but my experience, as well as my studies of the impact of immigration on labor markets, lead me to several conclusions:3

1. Much of the controversy among economists is over data and methods. Although there have been improvements, there are no accurate data on unauthorized immigration. There are, in particular, no longitudinal data that follow the same workers through time. Analysts therefore make mistakes when they attempt to reach longitudinal inferences from cross-sectional data. For example, data comparing the impact of immigrants on native wages in metropolitan areas at different dates must account for inter-area migration. This is because competing low-wage authorized residents tend to avoid areas with heavy influxes of unauthorized immigration, while higher wage authorized residents tend to move into those areas. Any inter-city study that did not account for these migrations could conclude, erroneously, that unauthorized immigrants had no negative—or even positive—effects on native workers.

2. As noted, studies that do not distinguish unauthorized from authorized migration are likely to reach erroneous conclusions about the impact of unauthorized immigration.

3. Labor market conditions clearly make a difference. The negative immigration effects for low-wage natives will be greater if there is widespread joblessness among native workers who, for reasons noted earlier, could not compete with the unauthorized immigrants even if they wanted to. The magnetic relations between employers and unauthorized immigrants are not likely to be detected by quantitative analyses.

4. Whatever the limitations of empirical research, economic theory predicts that natives whose work is complementary to that of immigrants (e.g., managers or skilled workers) will benefit from immigration, but that the wages of those workers who compete directly with immigrants will be reduced. Therefore, public policy should minimize low-wage competition and maximize complementarity.

5. Although the magnitude can be debated, there is little question that unauthorized immigration reduces the wages and dilutes the quality of jobs for low-wage domestic workers. It is true, of course, that immigration is not the only factor depressing these wages, but it is a significant one, especially for high school dropouts, whose real wages have fallen by over 16% since 1979 because of immigration, globalization, technological changes, and weaker worker protections. Public policy makers should develop immigration, social, and high-value-added economic policies to enable these workers to maintain and improve their conditions.

Since workers tend to be segmented into non-competing groups, it is useful to examine the impact of immigrants on young and minority workers who compete most directly with them. In a careful assessment of these effects, three Northeastern University labor market researchers permit an assessment of these effects: these analysts found that immigrants who arrived in the United States between 2000 and 2005 (over half—56%—of whom were unauthorized) accounted for an unprecedented 86% of the net increase in the number of employed persons, displaced native-born workers, and weakened the structure of labor markets.4 The impact was particularly large for young native-born males (16 to 34), whose employment fell by 1.7 million between 2000 and 2005, while the number of young immigrant males increased by 1.9 million. The negative impact was greater for young blacks and Hispanics. These researchers also found that the employment of immigrants was accompanied by a shift in the structure of private labor markets toward more informal employment not covered by unemployment insurance, health benefits, and worker protections.5

The argument that immigration strengthens the competitiveness of the American economy depends on how competitiveness is defined. What many economists mean is that lower wages improve competitiveness because they reduce the price of American products. But, while this is an easy option for employers, wage competition is a losing strategy for workers, communities, and nations: there are always countries with lower wages. For example, the United States is losing jobs to Mexico, which, in turn, is losing jobs to China and other countries where wages are much lower than Mexico's. Moreover, in a high-wage country, wage competition implies lower and more unequal wages, which is exactly what has been happening in the United States since the 1970s. There can be little doubt that growing inequality will weaken democratic institutions, economic performance, and national unity.

It is true, of course, that in a competitive global economy, earnings for similar workers tend to converge. The policy issue, however, is whether convergence takes the form of more rapidly rising wages in developing countries, which would be better for people everywhere, or lower wages in high-wage countries, which will increase inequality and reduce wages for many workers, as well as aggravate national and international tensions.

A better alternative, suggested by the experiences of some East Asian countries, would be for all nations to adopt value-added strategies to compete by improving productivity, quality, flexibility, and innovation. Given this definition, immigration that reduces American wages and perpetuates marginal, low-wage industries does not improve the kind of competitiveness we should encourage.

Immigration policy, therefore, should be designed to give greater attention to increasing the flow of workers whose skills and education are in short supply in the United States. This will not be done by unauthorized immigrants, who are predominantly workers with little formal education and limited English language skills. For example, according to The Instituto Tecnológico de México, between 1992 and 2002 over three-fourths of unauthorized Mexican immigrants had less than eight years of formal education; 11% had no formal education at all; and one-third had less than four years.6 This is a problem because even low-wage jobs have increasing education requirements.

During the 1970s it was often argued that unauthorized immigrants had positive fiscal impacts because they paid more taxes than the cost of public services they used. This might have been true when most immigrants were mainly young adults without families, but that is no longer the case as immigrants settle into the United States and form or unite families. Since most unauthorized immigrants have low incomes, it is not surprising that the taxes they pay do not cover the cost of the public services they receive. In a 2004 study, Gordon Hanson reported that 25% of unauthorized immigrants from Mexico, and 15% of natives, received some kind of welfare.7

In studies of New Jersey and California—two states with large immigrant populations—the National Research Council (NRC) found, on the basis of 1989-90 data, that the average immigrant household in New Jersey received a net fiscal transfer of $1,500 from natives, or 3% of average immigrant household income; in 1994-95, the average fiscal transfer from native to immigrant households in California was $3,500, 9% of average immigrant household income. For the United States, however, the NRC estimated that the short-run immigrant fiscal burden on native households was $200 or 0.2% of GDP.8 Thus, while the fiscal immigration burden for the whole country is relatively small, it is larger in states with relatively generous welfare benefits and higher percentages of immigrants with low incomes and more children.

What should we do?

The foundation for an effective immigration policy is to recognize the power of the forces perpetuating unauthorized immigration and find ways to authorize the flows and make immigration an integral component of economic and social policies to promote broadly shared prosperity in the United States, Mexico, and other countries.

Effective immigration policy must contain a comprehensive mix of measures, including stronger border controls and internal enforcement processes; a secure work authorization system with strong penalties against employer and immigrant violators; a means to adjust the status of people who have lived and worked satisfactorily in the United States for some years, accompanied by a credible signal that there are unlikely to be future status adjustments; an immigration standard that gives greater weight to the country's labor requirements; and cooperation with Mexico and other countries to encourage economic development in immigrant-exporting areas through cooperative investment, trade, and aid measures; and strengthening NAFTA's labor agreements to limit wage-suppressing competition and give workers stronger voices in the work place and in national policy decisions. Following are seven specific proposals for immigration reform.

1. Secure identifiers. The first priority should be to devise a secure work authorization system along the lines of the one we developed for SCIRP. Heightened concern about national security and advances in identification technology probably make a secure identifier more acceptable today than it was in the 1970s and 1980s.

2. Strong border controls and visa enforcement. The United States needs strong border and internal enforcement systems to prevent unauthorized immigrants from entering the United States or remaining after visas expire. Border security is clearly very important, but by itself will not be adequate since over one-third of unauthorized immigrants have overstayed visas. Visa violations will undoubtedly increase with tighter border controls.

3. Adjustment of status. The status of unauthorized immigrants who have satisfactorily settled in the United States should be adjusted. The Senate proposal on this matter would allow workers who have lived in the United States for five or more years to remain and earn the right to become citizens. Those who have been here for two to five years would be given guest worker status, but must leave the United States and reenter to keep their status. Those who have been in the United States for less than two years would be required to leave.

The adjustment of status is one of the trickiest and most controversial immigration proposals. If it is not done right, then immigrants will not come out of the shadows to authorize their status. Authorization also could accelerate the future flow of unauthorized immigrants, as IRCA did. The provision that workers who have been in the United States for two to five years must leave the country and apply for authorized guest worker status is problematic, as is the requirement that those here for less than two years leave the country. Unless a credible enforcement strategy is implemented, these provisions are not likely to be very effective. Immigrants know that their chances of being apprehended and removed are smaller than the probability that they can either remain in the United States and work or ultimately acquire lawful status through various authorized means. During the 1990s, for example, about 1.5 million unauthorized immigrants gained authorized status and only about 412,000 were removed. This was in addition to IRCA's amnesty provision, which authorized the status of 2.7 million immigrants—2 million of them from Mexico. But authorization did not slow the influx of unauthorized immigrants, partly because the seriously flawed identification system invited fraud and partly because those whose status was authorized were not allowed to bring their families.

The most common objection to allowing long-time unauthorized immigrants to become authorized residents and earn citizenship is that it rewards unauthorized behavior. It is true, of course, that the immigrants' behavior was unauthorized, but the law was so poorly constructed and haphazardly enforced that unauthorized immigrants have many co-conspirators. These include Congress, which passed a seriously flawed law and failed to adequately fund an effective enforcement strategy; businesses that hired workers with clearly fraudulent documents and, along with members of Congress, pressured officials not to enforce immigration laws; banks that issued credit cards to unauthorized immigrants; the IRS, which gave them taxpayer ID numbers; the public, which sympathized with hardworking immigrants who seemingly did little, if any, harm and purportedly only took jobs natives wouldn't take; various sympathetic support groups, who thought immigrants deserved the right to seek the American dream; labor unions—formerly among the staunchest opponents of immigrant worker programs—who now actively organize and protect unauthorized workers; and public officials in Mexico, who emphasize both the immigrants' constitutional right to migrate and America's dependence on unauthorized immigrants and have adopted measures, like photo ID cards, that facilitate unauthorized immigrants' ability to work and live in the United States. Given numerous co-conspirators, it would be hard to assign culpability only to the immigrants. We would be more justified in condemning unauthorized immigration if we had a law that, instead of being a confusing fiction, met the standards of a good law, i.e., was fair, transparent, and enforceable.

Of course, another reason to adjust the status of these immigrants is that the alternative of a massive roundup, modeled after the 1950s Operation Wetback, is unthinkable.

4. Foreign worker adjustment board. The composition and size of economic immigration should be calculated by an independent foreign worker adjustment board. And labor market needs should become a more important component of immigration policy, as they are in countries like Canada and Australia.

An adequately staffed independent board could make technical projections of labor market needs and balance the interests of employers, workers, and the public. Immigration is too technical and political to be left entirely to the Congress, especially where employers have inordinate power to import labor surpluses to keep wages down. In their immigration reform lobbying, for example, business groups make it clear that any outcome that gives them fewer foreign workers (authorized and unauthorized) is unacceptable. Lindsey Lowell, of the Institute for the Study of Immigration at Georgetown University, has estimated that the allowable number of foreign computer and engineering workers admitted under the Senate's immigration proposals would increase by a factor of five by 2017 and would cause foreign workers to be about 19% more than BLS's total projected employment in these occupations.9

5. Improve temporary worker programs. The United States should improve the administration of existing temporary worker programs, but should not adopt a large new guest worker initiative.10 Experience in the United States and Europe shows that the short-run economic benefits of guest worker programs are more than offset by long-run social, political, and economic problems. It is not good policy for a democracy to admit large numbers of workers with limited civil and employment rights. Because their frame of reference is conditions in their home countries, guest workers are willing to accept second-class status, at least for a while, but their children compare their conditions with those of natives and are likely to resent their inferior status. Indeed, much civil unrest in Europe has originated from the children of guest workers, who are citizens but still disadvantaged because of their parents' conditions.

In the late 1970s, an informal poll of a group of my fellow OECD labor ministers found that none of them would adopt a guest worker program if they had it to do over again. Moreover, all of them found that it was difficult to terminate these programs once they became institutionalized. Immigration experts have uniformly found that nothing is more permanent than a temporary guest worker program. For these reasons, while every major U.S. immigration study commission, including SCIRP and the 1995 U.S. Commission on Immigration Reform, started with the idea that a guest worker initiative might be desirable, after careful examination rejected such a program as bad policy.

Guest worker advocates usually contend that these programs will stem the flow of unauthorized immigration, citing the 1942-64 U.S.-Mexico bracero experience as justification for this conclusion. It is true that in the 1950s, when the bracero program—ostensibly a temporary wartime agricultural worker initiative—more than doubled, peaking at over 450,000 workers, border apprehensions declined. However, the number of unauthorized immigrants probably remained much higher than the number of braceros. Between 1942 and 1946, about 4.6 million braceros were admitted and 5.2 million unauthorized immigrants were apprehended. The number of unauthorized immigrants is estimated to be three times the number of apprehensions.

Advocates cite the 1980 Congressional Research Service (CRS) study prepared for SCIRP to justify these programs, although they rarely give the full quote, which is: 11

...the bracero program by itself did not prove to be a solution to the problem of large-scale illegal entry from Mexico. On the contrary, as it was administered during the early stages...the existence of the bracero program appeared to make the problem worse. It was not until sharply increased enforcement measures were combined with greatly expanded programs that it was possible to divert most of the illegal flow into legal channels. However...both these measures were effected at a considerable price, in terms of the apparent adverse effect on domestic agricultural workers...and the ill will created, particularly in the Mexican-American community by Operation Wetback.

The bracero program also caused considerable friction between the U.S. and Mexican governments, as well as corruption by program administrators and abuse of braceros in the United States and Mexico. As the program became institutionalized, it was very hard to terminate. It nevertheless resulted in strong immigrant-employer bonds, which remained and spread to other industries and geographic areas on an unauthorized basis after the program ended. The bracero program thus sowed the seeds of increased unauthorized immigration. It is therefore a real stretch to argue that this experience supports the need to revive a large-scale guest worker program as an immigration control measure.

New guest worker programs are not only unwise, but also unnecessary. If an independent entity concludes that more foreign workers are needed, they should be admitted as immigrants with full legal rights, including the right to earn citizenship. And, if qualified unauthorized immigrants' status is authorized and they unite their immediate families, there will automatically be a continuing flow of workers from Mexico and other source countries.

If it is concluded that more truly temporary foreign workers are needed, this should be achieved by improving the administration and strengthening the foreign and domestic worker protection of current programs. Employers complain that these programs are too cumbersome and litigious, at least partly because they do not like the provisions protecting the interests of foreign and domestic workers. Employers have been able to "game" the system to get the foreign workers they prefer and want the recruiting standard to be predicated on finding U.S. workers who are as good as the highly screened foreign workers, not the legal requirement that we recruit domestic workers who meet reasonable minimum standards.

It is particularly important to strengthen the worker protections in present temporary worker programs. There is abundant evidence that disparate foreign workers are subjected to appalling abuses in the United States and their home countries; these include fraudulent claims by recruiters and contractors about the quality and amount of work in the United States, deplorable living conditions, and failing to pay for work done. The practice of seizing foreign workers' passports and their heavy dependence on particular employers often subject these workers to near-peonage conditions.12 My experience suggests that employers' desire for low-wage compliant labor and guest workers' limited options make it difficult—but not impossible—to protect these workers. However, these protections must be included in a comprehensive immigration reform program.

6. Protective labor legislation. The rigorous enforcement of protective labor legislation, especially a higher minimum wage, also would make many jobs held by unauthorized immigrants more attractive to domestic workers. Some experts believe that a higher minimum wage would be sufficient to stem unauthorized immigration.13 While a higher minimum wage and strengthened enforcement of protective labor laws are desirable, they would not be adequate immigration control measures. Many employers would prefer unauthorized immigrants to natives even at the minimum wage: many unauthorized immigrants are in informal and exempt sectors of labor markets not affected by the minimum wage, and a higher minimum wage would attract more unauthorized immigrants. The tight bond between immigrants and employers is not likely to be altered very much by minimum wage enforcement alone.

7. Trade, investment, and aid programs. The ultimate solution to the unauthorized immigrant problem will be sufficient growth in Mexico and other source countries to provide suitable employment for their citizens. Unfortunately, Mexico's growth is unlikely to provide acceptable jobs for most of its new workers anytime soon. Mexicans migrate because of the low quality of jobs, not just the number; about 90% of all migrants have jobs when they migrate. Mexico's daily minimum wage is less than the U.S. hourly wage; the average Mexican wage is 10% to 20% of the U.S. wage, about half of its population lives below poverty levels, and half of its economically active population is in the informal sector. Although Mexico's official labor standards are high, and opportunities have improved for some workers, actual conditions for most workers are very poor and independent labor unions have great difficulty operating because they are subordinated to undemocratic, often corrupt government-controlled organizations. Hence, it is not surprising that polls show almost half (49%) of Mexico's adults say they would like to move to the United States.

Although there have been important improvements in political institutions and macroeconomic performance, Mexico's low-wage development policies are unlikely to stem the flow of unauthorized migration. As noted, Mexico is losing jobs to other countries, especially China, where average wages are less than half of Mexico's.

Mexican authorities, therefore, would like to have a bilateral agreement with the United States modeled after the bracero program. And it looked for a while like such a program, supported by President Bush and given high priority by former Mexican President Vicente Fox, would be adopted. However, this proposal was derailed in 2001 by the terrorist attacks of September 11, and has not gotten back on track, much to the chagrin of President Fox and other Mexican leaders.

However, it would not be advisable to revive a bracero-like program, which, in addition to reasons presented earlier, would be a source of friction because Mexico is unlikely to prevent its citizens from exercising their constitutional right to migrate and the United States is unlikely to surrender its sovereign right to control immigration. On the other hand, Mexican leaders believe Mexicans should have the right to migrate to the United States, and have echoed the common myth that immigrants only take jobs Americans won't fill. Before September 11, they thought U.S. employers' dependence on migrants was sufficiently strong and U.S. immigration controls were sufficiently weak that the United States was unlikely to do much to stem the flow of unauthorized workers.

The United States should, however, help Mexico promote job growth in its primary migrant-exporting areas. Desirable activities include trade and investment policies focused on those areas, infrastructure development, and strengthening NAFTA's weak labor side agreement, especially to give workers greater control of their own unions, thus strengthening their voice at work and in national policy making. The United States should make a major effort to help Mexico improve basic education, especially for low-income students and girls. There is considerable evidence that the education of girls is a very significant way to break intergenerational poverty cycles.14

The United States also should consider Mexico's proposal to create a joint Canadian-Mexican-U.S. development fund modeled after the very successful European Union experience, which did much to improve conditions in poorer European countries and stem the flow of migrants to the richer countries expected with economic integration. Of course, the rich-poor gaps in Europe were much smaller than in North America. Trade, investment, and aid programs should be used to leverage the structural reforms necessary for faster economic growth in Mexico.


Immigration is one of the most difficult issues policy makers face, mainly because of complex contending political, ethnic, and economic interests. Effective immigration policies clearly will affect the strength of our economy, social and ethnic stability, and our relations with Mexico and other countries. The United States must get it right this time.

Ray Marshall was Secretary of Labor in the Carter administration. He is Professor Emeritus and holder of the Audre and Bernard Rapoport Centennial Chair in Economics and Public Affairs of the LBJ School of Public Affairs at the University of Texas. He is author of more than 30 books and monographs, including Thinking for a Living: Education and the Wealth of Nations, and Back to Shared Prosperity, and a member of the board at EPI.


1. David Card, Is the New Immigration Really so Bad? National Bureau of Economic Research, August 2005.

2. George B. Borjas and Lawrence F. Katz, "The evolution of the Mexican-born workforce in the United States." National Bureau of Economic Research, April 2005.

3. Ray Marshall, "Immigration and human capital," in David Hornbeck and Lester M. Salamon (eds.), An Economic Strategy for the Nineties (Baltimore, Md.: Johns Hopkins Press, 1991); and "Immigration: An international economic perspective." International Migration Review, 18, Fall 1984.

4. Andrew Sum, Paul Harrington and Ishwar Khatiwada, "The impact of new immigrants on young native-born workers, 2000-05," Center for Immigration Studies, September 2006.

5. Ibid.

6. William P. Kocewicz, "What ails Mexico's economy," NRO Financial, May 3, 2006.

7. Gordon H. Hanson, Immigration Policy, National Bureau of Economic Research, July 2004.

8. James P. Smith and Barry Edmonson (eds.), The New Americans: Economic, Demographic and Fiscal Effects of Immigration. Washington, D.C.: Academy Press, 1997.

9. B. Lindsey Lowell, Projected Numbers of Foreign Computer and Engineering Workers Under the Senate's Comprehensive Immigrants Reform Act (S.2611), Institute for the Study of International Migration, Georgetown University, August 2006.

10. The Citizenship and Immigration Service issues 70 different types of visas that increase the number of foreign workers. In 2005, the types and numbers of non-immigrant admissions were:

Type of visa


F-1 (student)


H-1B (high-skilled workers and fashion models sponsored by employers)


H-2B (seasonal nonagricultural workers)


L-1 (executives of foreign companies with U.S. offices)


B-1 (temporary visitors for business)


H-2A (temporary agricultural workers)


P-1 (internationally recognized athletes or entertainers)


Source: Elizabeth M. Grieco, "Temporary admissions of nonimmigrants to the United States," Annual Flow Report, July 2006, p. 3.

Individuals admitted temporarily for specific purposes are expected to leave after six years (H-1B visas) and students are supposed to leave after completing their studies. However, Lindsey Lowell estimates that most "temporary" visa holders become permanent residents, including two-thirds of students and half of workers.

11. The Congressional Research Service, Temporary Worker Programs: Background and Issues, A report for the use of the SCIRP, 90th Cong., 2nd Session, February 1980, p. 22.

12. See Steven Greenhouse, "Low pay and broken promises great guest workers." The New York Times, February 28, 2007.

13. See Michael S. Dukakis and J.B. Mitchell, "Raise wages, not walls," The New York Times, July 25, 2006, p. A23.

14. Nancy Birdsall, Education: The People's Asset, CSED Worker Paper No. 5, September 1999; Nancy Birdsall, David Russ, and Richard Sabot, "Inequality and growth reconsidered: Lessons form East Asia," Economic Review, Vol. 8, No. 3, 1995.

Download print-friendly PDF version Acrobat (PDF)

See also Immigration for Shared Prosperity (an EPI book, April 2009)

Top of page



Copyright © 2010 Economic Policy Institute. All rights reserved. Contact