CIAO DATE: 04/02

U.S. Immigration Policy: Unilateral and Cooperative Responses to Undocumented Immigration

Table of Contents

List of Tables and Figures

Introduction

The Current Undocumented Immigration Policy Regime

Summary of Current Policy

Effectiveness of Current Policy Regime

Implications of Current Policy Regime

Three Alternative Proposals

Expanded H-2A Guest Worker Program

Implications of S.1814 and S.1815 for Various Economic Groups

A Strict Enforcement Regime

Implications of a Strict Enforcement Regime for Various Economic Groups

A Market-based Solution to Undocumented Immigration

Implications of a Market-based Regime for Various Economic Groups

Conclusions: Recommendation and Feasibility

Effectiveness and Efficiency

The Politics of Immigration Reform

References

Table 1. Immigration Enforcement Statistics

Table 2. Budget Outlays for the INS in Millions of Dollars and as a Percentage of Total  Federal Budget

Figure 1. Agricultural and Overall Unemployment, 1950-1990

Research for this paper was supported by a grant from the Institute on Global Conflict and Cooperation. A previous version of this paper was presented at the annual meeting of the International Studies Association, 14?18 March 2000, in Los Angeles. I am thankful for the comments of John Torpey on that version of the paper.

Introduction

This paper addresses the problem of undocumented immigration to the UnitedStates from Mexico, and current and proposed policies designed to control these undocumented flows. Undocumented migration from Mexico is a subject that already receives disproportionate attention in the sense that many-and probably most-undocumented immigrants in the United States do not illegally cross the U.S.-Mexican border, yet INS enforcement efforts focus overwhelmingly on these border crossers. Although undocumented Mexican migration to the United States is disproportionately targeted, the subject merits analytical attention for three reasons. First, undocumented immigration from Mexico to the United States is the largest illicit migration flow in the world, at about one million crossings per year. Second, partly for this reason, U.S. enforcement efforts devoted to controlling Mexican immigration cost taxpayers billions of dollars, and have resulted in the transformation of the INS into the largest civilian gun-carrying force in the world. And third, immigration remains central to U.S.-Mexican bilateral relations (Binational Commission 1997, Rico 1992, Rosenblum 1998) as U.S. immigration policy-making takes on an increasingly transnational character (Rosenblum 1999 and forthcoming).

Yet despite the U.S. commitment to NAFTA and President Bill Clinton's commitment to expand free trade throughout the Western Hemisphere by the year 2005, there has been limited effort to explicitly address Mexico-U.S. immigration within the context of the deepening bilateral relationship. ₁ The failure to create a bilateral immigration policy is ironic in light of other linkages between economic integration and immigration in the EU and, to a limited extent, in the case of U.S.-Caribbean Basin relations (Rosenblum 1999, and forthcoming). The failure to pursue bilateralism is also ironic given its previous success, including during the early years of the U.S.-Mexico bracero program (Craig 1971, McCain 1970, Rosenblum forthcoming).

In this paper, I summarize current U.S. policy toward undocumented Mexican immigration, which has been an expensive failure. I then take up three competing policy proposals: one pending in the U.S. Senate (S.1814 and S.1815) to expand the H-2A guest-worker program; one to construct a strict enforcement regime; and one based on linking U.S.-Mexican free trade to a free flow of labor. For each alternative, I predict likely outcomes and distributional consequences for seven types of actors (U.S. workers, U.S. consumers, U.S. employers, other U.S. citizens, undocumented immigrants, legal immigrants, and other Mexicans). I conclude that a binational approach to immigration control (a North American Common Market) is the most promising option, and I discuss its political feasibility.

The Current Undocumented Immigration Policy Regime

Since the late 1970s, U.S. immigration policy has made the reduction of undocumented immigration across the U.S.-Mexican border its highest priority. The specific goals established by the 1981 U.S. Select Commission on Immigration and Refugee Policy (SCIRP) and reflected in congressional debate since then are to limit illegal inflows, protect U.S. jobs, and ensure that employers still have access to labor.

Summary of Current Policy

The current U.S. policy regime employs four approaches to deter undocumented immigration from Mexico: border enforcement,; employer sanctions; a limited agricultural guest worker program; and limits on social programs for undocumented migrants.

Border Enforcement and Removal of Immigrants

Since 1978, the United States has expanded border enforcement in three separate stages (see Dunn 1996). The first stage began when Congress responded to rising undocumented immigration (see Table 1) by increasing INS funding by 24 percent between 1978 and 1980 (see Table 2), with many of the new resources going to equipment and hardware. While President Jimmy Carter was reluctant to expand actual border enforcement, ₂ "the expansion of the INS was taken to an unprecedented level during the Reagan administration, as the urgency surrounding immigration and 'border security' topics became even greater during this period" (Dunn 1996, 41). Congress' passage of the 1986 Immigration Reform and Control Act (IRCA) initiated the second stage of enhanced border enforcement as funding was increased again by a third between 1986 and 1988, and as Congress strengthened penalties against migrant smugglers. The third stage began in 1994 when the Justice Department and the INS initiated a program of "prevention through deterrence" designed to "make it so difficult to enter this country illegally that fewer individuals would even try" (GAO 1999, 3). The strategy sought to close off the most commonly used routes (through border cities) and shift undocumented traffic into remote areas where enforcement agents would have an advantage. Border escalation was endorsed by Congress in 1996 with the passage of the Illegal Immigrant Responsibility and Immigration Reform Act (IIRIRA) which authorized funding for 1,000 new Border Patrol agents and 300 new support staff each year for five years. The IIRIRA also streamlined deportation procedures, making it easier for aliens to be deemed inadmissible and removed with a shorter hearing process or none at all.

Table 1. Immigration Enforcement Statistics
 

Year	Total Apprehensions	Total Border Patrol Apprehensions	Mexicans Apprehended by Border Patrol	Mexican Undocumented Immigrants (Estimated)
1970	345,353	231,100	219,300	345,543
1971	420,126	302,500	348,100	420,126
1972	505,949	369,500	430,200	505,949
1973	655,968	493,100	480,600	655,968
1974	788,145	634,800	616,600	788,145
1975	766,600	596,800	579,400	766,600
1976	875,915	696,000	678,400	875,915
1977	1,042,215	812,600	792,600	1,042,215
1978	1,057,977	862,300	841,500	1,057,977
1979	1,076,418	888,700	866,800	1,076,418
1980	910,361	759,400	734,200	910,361
1981	975,780	825,300	797,900	975,780
1982	970,246	819,900	795,400	970,246
1983	1,251,357	1,105,700	1,076,300	1,251,357
1984	1,246,981	1,138,600	1,102,600	1,241,489
1985	1,348,749	1,262,400	1,218,700	1,320,000
1986	1,767,400	1,692,500	1,635,700
1987	1,190,488	1,159,000	1,124,000
1988	1,008,145	971,000	929,800
1989	954,243	893,000	832,200
1990	1,169,939	1,103,400	1,105,400
1991	1,197,857	1,132,933	1,095,100
1992	1,258,481	1,199,560	1,168,900	1,320,000
1993	1,327,261	1,263,490	1,230,100
1994	1,094,719	1,031,668	999,900
1995	1,394,554	1,324,202	1,293,500
1996	1,649,986	1,549,876	2,700,000
1997	1,536,520	1,412,953

Sources: INS, Statistical Yearbook,various years; Lorey 1990, tables 910, 921, 1012; U.S. Commerce Department, Statistical Abstract of the United States, various years.

Table 2. Budget Outlays for the INS in Millions of Dollars and as a Percentage of Total Federal Budget
 

Year	Total  INS budget	Percent of U.S. budget	I.N.S Enforcementa	Percent of U.S. budget
1970	105,768	0.054	66,783	0.034
1971	121,930	0.058	76,624	0.036
1972	130,934	0.057	83,465	0.036
1973	137,468	0.056	88,372	0.036
1974	155,161	0.058	88,372	0.033
1975	181,230	0.055	102,170	0.031
1976	214,609	0.058	117,768	0.032
1977	244,515	0.255	172,062	0.179
1978	283,087	0.069	187,527	0.046
1979	309,285	0.067	190,840	0.042
1980	340,742	0.068	219,378	0.044
1981	366,017	0.062	229,091	0.039
1982	446,461	0.066	276,813	0.041
1983	500,972	0.067	286,969	0.038
1984	510,638	0.063	300,378	0.037
1985	591,617	0.069	339,587	0.040
1986	571,267	0.060	360,860	0.038
1987	745,590	0.075	390,673	0.039
1988	729,314	0.073	479,963	0.048
1989	822,023	0.077	574,011	0.054
1990	1,171,185	0.102	594,466	0.052
1991	1,317,790	0.105	638,267	0.051
1992	1,472,453	0.111	700,693	0.053
1993	1,551,194	0.112	739,813	0.054
1994	1,686,097	0.120	783,906.5	0.056
1995	1,821,000	0.125	828,000	0.057
1996	2,291,000	0.151	1,016,000	0.067
1997	3,136,000	0.201	1,537,000	0.098
1998	3,761,000	0.235	1,674,000	0.105
1999	4,095,000	0.246	1,925,000	0.115

^aEnforcement = Separate entries for inspection for admission, detention and deportation, and Border Patrol prior to 1982; separate entry for "enforcement" starting in 1982.
Sources: U. S. GPO, The Budget of the U.S. Government; and U.S. Commerce Department, Statistical Abstract of the United States, various years.

Employer Sanctions

The 1986 IRCA included a provision making it illegal to employ undocumented immigrants, establishing civil penalties for first-time offenders and criminal penalties for repeat offenders. ₃ While IRCA responded to long-standing calls for employer sanctions, the legislation was problematic, as several analysts have noted (for example, Calavita 1994, D. Martin 1994, Fix 1991). IRCA made it illegal to knowingly hire an undocumented immigrant, but failed to establish a reliable procedure for determining work eligibility. Instead, IRCA requires employers to fill out an "I-9" form which documents that prospective employees presented two or three of 25 possible forms of identification, making it difficult for well-intentioned employers to insure that the documents were legitimate and easy for others to plausibly deny knowledge of workers' undocumented status. IRCA also included several features which guaranteed that it would not be enforced aggressively, including the requirement that INS officials obtain search warrants before inspecting agricultural worksites, and the dedication of limited funds to worksite inspections. With the passage of the IIRIRA in 1996, some enforcement problems were addressed through the creation of pilot programs to allow employers to verify citizenship status; but these programs have received limited support.

Agricultural Guest Worker Program

The IRCA combined the "stick" of increased border enforcement and employer sanctions with the "carrot" of the H-2A agricultural guest worker program. ₄ The H-2A program is modeled on the bracero guest worker program, and is designed to insure that agricultural employers have adequate access to labor without damaging the employment prospects of U.S. workers. Under the program, employers apply to the Department of Labor (DOL) to receive guest workers. The DOL certifies that a labor shortage exists, that wages and working conditions for domestic farmworkers will not be "adversely affected" by importing guest workers, and that housing is available. Employers then apply to the INS for approval to recruit employees, a step that is usually pro forma, but time-consuming. Finally, 60 days after initial paperwork is filed, H-2A workers are made available to employers on three- to six-month visas.

Denying Access to Social Programs

Finally, in 1996, Congress passed the IIRIRA, the centerpiece of which was to deny immigrants-including legal immigrants-access to a variety of social welfare programs. Undocumented immigrants were denied access to Social Security benefits, subsidized housing assistance, and food stamps. Legal immigration was restricted by raising the minimum household income level required to sponsor immigrants and holding sponsors accountable before immigrants gain access to means-tested benefits.

Effectiveness of Current Policy Regime

The current policy regime fails to meet its stated goals in almost every way: It does not deter undocumented immigration, protect U.S. jobs, or guarantee an adequate supply of unskilled labor.

Deterrence of Undocumented Immigration

The enhancement and expansion of immigration enforcement since the 1970s has been accompanied by an increase in the estimated number of undocumented immigrants, from slightly over 1 million when the current policy regime was initiated, almost all of whom were from Mexico, to approximately 5 million as of 1995, with about 2.7 million from Mexico. By concentrating overwhelmingly on the U.S.-Mexican border, the current regime fails to address the underlying causes of migration or to provide a comprehensive deterrent to undocumented immigration. There is widespread agreement among academics that immigration (both legal and undocumented) is a function of three types of forces: economic "pushes" which encourage people to leave their home countries, economic "pulls" which attract them to a destination, and "social networks" which facilitate flows to a particular destination (see, for example, Binational Commission 1997, Massey et al. 1993 and 1994, Stalker 1994). Official U.S. policy raises the cost of migrating at the border and attempts-in theory-to address only one of the three motivations for immigration, by reducing pull factors through employer sanctions and the reduction of social benefits. These policies fail to actually deter migration for three reasons.

First, employer sanctions are ineffective because few resources are devoted to work-site enforcement. Between 1986 and 1997, 92 to 97 percent of all INS apprehensions occurred at the border, with 97 percent of these at the U.S.-Mexican border (GAO 1997). In 1997, 7,537 workplace inspections occurred, which was an all-time high (GAO 1997), but which still represented only about one for every seven hundred estimated undocumented immigrants. Since its peak in 1997, the INS has explicitly adopted a strategy of abandoning work-site enforcement: "It is just the market at work, drawing people to jobs, and the INS has chosen to concentrate its actions on aliens who are a danger to the community." ₅ Moreover, even when inspections are made, the structure of the I-9 form makes it difficult for the INS to prosecute employers, who therefore have no incentive to stop hiring undocumented workers. As a result, employer sanctions have had little effect on the importance of undocumented labor in the United States (Cornelius 1998, Fix 1991).

Second, recent efforts to eliminate immigration pulls have focused on a "social service magnet" which probably does not exist. There is no evidence that undocumented immigrants are motivated, even in part, by the existence of social benefits in the United States. On the contrary, immigrants consistently identify the desire to find employment as their motivation for migration (Cornelius 1998, Massey et al. 1993 and 1994, U.S. Commission 1990). Moreover, when controlling for race and class, undocumented immigrants are unlikely to take advantage of social services available to them (Marcelli and Heer 1998).

Preliminary data suggest that INS efforts to raise the cost of crossing the border have been somewhat successful, but that costs do not outweigh the expected benefits of migration. The most easily observed outcome of the build-up in enforcement activities has been a shift in the flow of illegal alien traffic from heavily-guarded urban sectors to less-guarded rural areas. According to the most recent figures, apprehensions in San Diego and El Paso, traditionally the sectors accounting for the greatest number of undocumented entries, dropped from 408,265 in FY1997 to 373,127 in FY1998. As a percentage of all apprehensions, these two sectors decreased from 68 percent in FY1993 to just 24 percent in FY1998 (GAO 1999, 20). A growing number of immigrants are also being apprehended at ports of entry using fraudulent documents. Third, there is anecdotal evidence that fees charged by immigrant smugglers, or coyotes, have increased since the onset of the enhanced border enforcement program (GAO 1999, 23; author interviews). Finally, the most basic measure of the cost of entering the United States is the jump in migrant deaths at the border, from almost none before 1995, to 324 in 1999.

Although these observations indicate that the cost of illegally crossing the U.S.-Mexican border has increased in response to enforcement efforts, there is no indication that costs outweigh expected benefits for most migrants. Indeed, the fact that total apprehensions along the southwest border remain near historic highs (see Table 1.) suggests that the overall expansion in enforcement has had little effect. Rather, it appears that immigrants have been "squeezed" to different parts of the border (GAO 1999, Dillon 2000), ₆ and that the deterrent effect of increased border enforcement anticipated by the INS has, thus far, failed to materialize. Moreover, according to the INS's own estimates, since half of all undocumented immigrants overstay their legal entry visas, and so never cross the U.S.-Mexican border illegally; the current enforcement regime makes no effort to deter these immigrants.

Effects on U.S. Jobs

The second goal of U.S. immigration policy is to protect U.S. workers from excessive competition. Assessing the effects of policies on U.S. labor implies two questions: first, to what extent do undocumented migrants take jobs that would otherwise be held by native workers, and second, to what extent do undocumented migrants depress wages? Both of these questions have been analyzed extensively (see, for example, Smith and Edmonston 1997), and most analysts agree that current policies have done little to limit the availability of immigrant labor or the ease of finding illegal work. Rather, consistent with dual labor market theories, the majority of recent immigrants, and especially the majority of undocumented immigrants, are employed in sectors of the economy that are shunned by native workers. ₇ Regarding wages, the conventional wisdom is that undocumented immigration has a negative effect on wages in immigrant-dominated industries, an effect felt most strongly by recent immigrants and U.S. minorities who compete with immigrants (Hamermesh and Bean 1998, Portes 1995). It is likely that negative effects are felt up the pay scale and in other industries to a limited degree. Perhaps the strongest evidence that the current regime hurts U.S. workers is fact that organized labor-the group which worked hardest to obtain employer sanctions-has recently come out strongly in opposition to the policy (see below).

Figure 1. Agricultural and Overall Unemployment, 1950-1990

Source: Data taken from Jacobs 1999.
 

Availability of Labor

The third goal of the current regime is to guarantee sufficient unskilled labor to U.S. employers. ₈ The partial failure (at least) of border enforcement and employer sanctions as a migration deterrent implies that this goal is met. Nonetheless, agricultural interest groups and others have complained of limited availability of low-skilled workers, so the question merits attention.

There is no evidence of a labor shortage in agriculture or in other unskilled sectors of the economy. On the contrary, agriculture, the service sector, and low-skilled manufacturing have consistently displayed higher unemployment rates than other sectors of the economy. A GAO analysis of 20 large agricultural counties found that 11 out of 20 had unemployment rates double the national average, and 15 out of 20 had an unemployment rate in 1997 greater than 7.2 percent. The GAO also concluded that agricultural areas have high unemployment all year, including during peak agricultural periods (GAO 1997, 27). Moreover, while agricultural unemployment has fallen from its high of sixteen percent in 1983, the ratio of agricultural to overall unemployment remains near its post-war high of two to one (see Figure 1); so there is no agricultural labor shortage relative to the overall state of the economy. Interviews in San Diego County also failed to find any evidence of a labor shortage among non-agricultural immigrant-dependent industries, including in the period after the initiation of San Diego's border build-up (Cornelius 1998).

There is also no evidence that immigration restrictions have caused wages to rise, as noted above. On the contrary, U.S. Federal Reserve Chairman Alan Greenspan has argued that wage-depression effects of legal and undocumented immigration are primary factors contributing to high growth without inflation in the United States. ₉ Regarding agriculture, the GAO cites declining agricultural wage rates, in real terms, in the 1990s despite rising wages in other industries (GAO 1997, 28). For this reason, labor-intensive agribusiness has been highly profitable, with production increases of 52 percent in the decade following IRCA passage, and exports quadrupling to $10.6 billion (Farmworker Justice Fund 2000).

Implications of Current Policy Regime for Various Economic Groups

U.S. Workers

As noted above, the failure to control undocumented immigration has had a mild downward effect on wages, hurting U.S. workers. Ironically, employer sanctions legislation hurts U.S. workers by making it illegal to work without documentation, but then failing to punish employers of undocumented immigrants. This policy combination places undocumented workers at a severe disadvantage in their negotiations with employers. In the extreme, employers take advantage of this asymmetry to report troublesome or pro-union employees to the INS or to avoid paying full wages. ₁₀ Even when employers do not resort to these methods, the threat of deportation limits undocumented immigrants' labor rights, allows employers to maintain substandard working conditions, and discourages unionization. Given that unionization remains one of the best predictors of wage increases, the existence of employer sanctions contributes to the negative effect of immigration on wages, with some ripple effects in other industries.

For these reasons, U.S. labor unions now oppose employer sanctions. Unions were at the forefront of calls for employer sanctions throughout the twentieth century (Mink 1986). But following the 1986 passage of IRCA, unions representing unskilled workers, including ACTWU (now UNITE), SEIU, and the ILGWU, changed their positions when sanctions failed to deter immigration but made organizing more difficult (Delgado 1993, Haus 1995). ₁₁ The campaign by immigrant-constituent unions to repeal sanctions expanded as efforts to organize undocumented immigrants also grew. ₁₂ In 1994, the California Labor Federation passed a resolution against sanctions; and, most significantly, in February, 2000, the AFL-CIO executive council supported a major overhaul of U.S. immigration policy: "Current efforts to improve immigration enforcement, while failing to stop the flow of undocumented people in to the United States, have resulted in a system that causes discrimination and leaves unpunished unscrupulous employers who exploit undocumented workers, thus denying labor rights for all workers. . . . We strongly believe employer sanctions, as a nationwide policy applied to all workplaces, has failed and should be eliminated." ₁₃

U.S. Employers

U.S. employers benefit from a large pool of labor and the ability to pay low wages, but the current system is bad for employers in at least one respect. The system encourages employers in immigrant-dependent industries to break the law because those who prefer not to hire undocumented immigrants face a collective action problem knowing that their competitors are likely to do so. As a result, even though most employers are never punished for hiring unauthorized workers, individual employers are vulnerable to prosecution if they happen to be targeted.

U.S. Consumers

U.S. consumers unambiguously benefit from the current immigration policy regime. By failing to meaningfully control the flow of undocumented immigrants at the border or the workplace, current policy insures that the pool of unskilled labor remains large, and that prices remain low. In effect, undocumented and legal immigrants subsidize low-priced consumption in the United States by accepting low wages; and U.S. consumers pay half as much for food and other basic goods than do Europeans.

Other U.S. Citizens

At the same time, the enforcement-oriented regime is costly in various ways. Since 1976, overall and enforcement-specific INS funding has grown from $214.6 and $117.7 million, respectively, to $4.1 and $1.93 billion as of 1999 (see Table 2). Even these high spending levels are not enough to fund the current approach to immigration control, however: INS detention centers are filled beyond capacity, ₁₄ and the agency suffers from major morale and recruitment problems. ₁₅ Legal immigrants and U.S. citizens of Hispanic descent and all who live near the U.S.-Mexican border face additional costs associated with the current enforcement regime. Hispanic-Americans are frequently stopped and harassed by Border Patrol agents, ₁₆ and border area residents complain that INS tactics are destructive and display a disregard for private property. ₁₇ As one Texas resident observed: "The question becomes whether the benefits of the Border Patrol doing this outweigh what I consider to be the Border Patrol violating our constitutional rights." ₁₈

Undocumented Immigrants

The current regime is both good and bad for undocumented Mexican immigrants. On the one hand, undocumented immigrants benefit from opportunities to work. On the other hand, to the extent that U.S. enforcement agents are de facto agents of anti-union employers, the current system depresses wages for undocumented immigrants below market levels. A third problem from the perspective of undocumented immigrants is the costs associated with crossing the border: the average price paid by undocumented immigrants to coyotes, or people smugglers, jumped from several hundred dollars in the early 1990s to over a thousand dollars by the end of the decade. This is not a subsidy of U.S. consumers or employers, but simply a transfer from poor Mexicans to criminal networks. Far more severe is the rising cost in deaths and injuries to immigrants trying to cross the border in remote areas.

Legal Immigrants

Recent immigrants also suffer under the current regime. On a general level, legal immigrants are frequently the victims of discrimination, a fact exacerbated by harsh anti-immigrant rhetoric among policy-makers. In 1996, second-class status was institutionalized with the passage of the IIRIRA, which established two tiers of legal residency, denying a range of social benefits to legal U.S. residents (Schuck 1998).

Current policy provides questionable benefits to the small number of immigrants who participate in the H-2A guest worker program. While the program benefits some 15,000 Mexicans each year by affording them temporary employment in the United States, the program's poorly designed enforcement mechanisms guarantee that many H-2A immigrants do not enjoy the wages and benefits promised them under the law. In contrast to the bracero guest worker program (1942?1964), Mexican consuls play no H-2A oversight role, so the program relies on H-2A workers themselves to know their rights under U.S. law and to initiate formal complaints should they have reason to do so. ₁₉ Moreover, any complaints must be filed during the actual contract period, since it is impossible for the DOL to interview workers after they have returned to Mexico. For these reasons, the current system does not work: in FY1996, the DOL received zero complaints from H-2A workers about wages and benefits (GAO 1997, 10).

Other Mexican Citizens

Finally, the current U.S. immigration enforcement regime has positive and negative consequences for Mexico on economic and political levels. On an economic level, Mexico benefits enormously from migrant remittances, estimated at six billion dollars per year. However, because undocumented immigrants receive below-market wages, and because there are high transaction costs associated with the transmission of remittances, these flows are lower than they "should" be. Second, there are political and diplomatic costs associated with the failure to include Mexico as a partner in immigration policy-making. At least since the debate over the 1986 IRCA, Mexican politicians have generally agreed that "migration is an international problem, and its solutions ought to be reached by the countries involved and not by the unilateral decisions of a government." ₂₀ In recent years, Mexican politicians of the left and the right have staked out their opposition to U.S. restrictions; and there is a consensus among Mexican politicians that "our number one priority . . . is to try to guarantee that while in the United States all Mexicans, regardless of their migratory status, have their rights respected." ₂₁ The hostility among Mexicans to what is commonly perceived as the U.S. "criminalization" of immigration is equally strong at the mass level as well (Rosenblum 1998).

Three Alternative Proposals

If the current system is broken, how should it be fixed? In this section I describe three alternative proposals. One of these is currently being debated in the U.S. Senate as bills S.1814 and S.1815 to expand and streamline the existing H-2A guest worker program. The other two are at opposite ends of the possible spectrum of enforcement responses: a strict unilateral enforcement regime, and a cooperative market-based regime. I summarize each proposal, describe its likely results if implemented, and analyze how each would affect the seven groups discussed above.

Expanded H-2A Guest Worker Program

Proposal for Guest Worker Reform (Summary of S.1814 and S.1815) In 1996, when Congress debated the IIRIRA, several members raised concerns (as they had in 1986) that legislation to limit undocumented immigration would cause an unskilled labor shortage. Three years later, S.1814 and S.1815 propose a two-pronged program to expand and streamline the H-2A guest worker program and to provide a limited amnesty for undocumented immigrants working in agriculture. According to the bills' sponsor, Senator Gordon Smith (R-OR), the goals of the program are to "reform the agricultural labor market, establish and maintain immigration control, provide a legal workforce for our farmers, and restore the dignity to the lives of thousands of farmworkers who have helped make the U.S. economy the powerhouse that it is today." ₂₂

S.1814 proposes to reform the H-2A program in three ways. First, the program would allow a greater number of H-2A visas to be issued and allow H-2A workers to be employed in a wider range of agricultural jobs. Second, the program would streamline the H-2A application process by limiting the efforts required of employers to recruit U.S. workers before H-2A workers, and by shortening the time frame for the DOL to respond to employer requests. Third, "guest worker reform" would lower wages and eliminate housing requirements.

S.1815 would complement H-2A reform by providing amnesty for certain undocumented immigrants. To qualify for temporary legal status, immigrants would be required to prove that they worked in agriculture at least 150 days in the previous year. Individuals who also work in agriculture for at least 180 days in five of the following seven years will qualify for permanent legal status. Of the total number who qualify after seven years, twenty percent will be given legal permanent resident (LPR) status in each of the following five years. Once immigrants receive LPR status, their families become eligible for family-based visa waiting lists. Five years after immigrants receive LPR status, they may apply for U.S. citizenship.

Anticipated Effects of Guest Worker Reform

The proposed guest worker reforms are unlikely to meet their stated goals. The most important goal of the program, based on the rhetoric of its supporters, is to assure an adequate supply of farm labor. But, as discussed above, there is no evidence that an agricultural labor shortage currently exists. Likewise, according to the GAO (1997, 24): "a sudden widespread farm labor shortage requiring the importation of large numbers of foreign workers is unlikely to occur in the near future" (see also U.S. Commission on Immigration Reform 1994 and 1997).

The guest worker reform proposal is also unlikely to convince current undocumented immigrants to legalize their status because the amnesty requirements are prohibitively difficult. The informal nature of seasonal agricultural employment guarantees that relatively few migrants who worked 150 days in the previous year will be able to document their labor. Moreover, given that the average seasonal worker only works 176 days each year, ₂₃ it will be difficult for individuals to work 180 days in five of the following seven years, much less to obtain documentation of that work. If the likelihood of receiving citizenship by staying in the H-2A program seems doubtful, many workers will abandon the program in favor of better wages and working conditions in other industries.

Moreover, the revised H-2A program will have even less effect on overall undocumented immigration, especially since only twelve percent of undocumented immigrants work in agriculture, according to INS estimates. By expanding guest worker recruitment, but making a long-term commitment to the program unattractive to immigrants, the most likely result is that the revised H-2A program will have the same result as virtually every other guest worker program in history: a high degree of "leakage" as guest workers leave the program to become undocumented immigrants.

The fourth stated goal of the guest worker proposal is to protect foreign and domestic workers, and "restore dignity" to their lives. Once again, it seems doubtful that the proposed program would achieve these goals. In its current form, the H-2A visa program is criticized by workers for its weak enforcement mechanisms. ₂₄ The proposed revisions do nothing to strengthen enforcement, and they lower official wages and reduce benefits.

Implications of S.1814 and S.1815 for Various Economic Groups

U.S. Workers

The effect on U.S. workers would mainly be limited to workers who compete with H-2A visa-holders for agricultural jobs. These workers would face increased competition as regulatory changes would make it easier to hire guest workers. ₂₅ The revised H-2A program would also maintain a bias against hiring U.S. workers by excusing H-2A employers from paying social security or unemployment taxes. For these reasons, S.1814 is opposed by every labor group in the country.

U.S. Employers

As strongly as U.S. workers oppose "guest worker reform," U.S. agricultural producers favor the legislation. As Western Growers Association President David Moore explained, "it is extremely important that growers of highly perishable fruit and vegetable crops be able to harvest their crops in a timely manner" (Stockwin 1999). The streamlined application procedure found in the proposed legislation is considered critical to making the H-2A program more attractive to growers. Thus, it is no surprise that the fourteen cosponsors of S.1814, the bill dealing specifically with the agricultural guest worker program, received an average 1999 rating of 81.4 (out of 100) from the American Farm Bureau Federation (author's calculation; data available at www.votesmart.org).

U.S. Consumers

U.S. consumers are likely to benefit slightly from the implementation of the proposed guest worker reforms, assuming savings from lower agricultural wages will be passed on to consumers. Agricultural wages are already lower than for any other sector of the economy, however, and it is unlikely that savings to U.S. consumers will be dramatic.

Other U.S. Citizens

Guest-worker reform is not costly to taxpayers. Assuming undocumented immigration is inherently costly on some level, the program would benefit U.S. citizens if a significant number of immigrants participate in the program rather than maintaining illegal status. But if, as seems more likely, the program fails to attract participants and instead results in a new round of guest-worker "leakage" and net increases in undocumented migration, then it would be somewhat costly to U.S. citizens.

Undocumented Immigrants

Immigrants who do participate in the amnesty program and gain citizenship from S.1815 would benefit from its passage: even when controlling for education and other social capital, legal immigrants receive higher wages than do undocumented workers. The program is more likely to have little effect since immigrants will choose not to participate. It may even have a negative effect for undocumented immigrants who attempt to participate, but fail to meet the requirements at some point in the following seven years, and are subsequently deported. Participation in the amnesty program could also force H-2A workers to choose between finding a living wage in non-agricultural work (since oversight of the H-2A program fails to protect wages) and protecting their amnesty status (by staying in the H-2A program).

Legal Immigrants

For legal immigrants already participating in the H-2A program, the proposed reforms would be costly. The biggest difference immigrants would notice would be falling wages. ₂₆ In addition, the revised H-2A program would allow forced production speed-ups without any wage increase. Finally, the proposed revisions would no longer require a minimum wage for each employee, but rather would require that the average employee wage rate be as high as the mandated minimum, creating the possibility that a large number of workers will receive even lower wages.

Other Mexican Citizens

Mexico has not taken an official position on either S.1814 or S.1815, and will probably resist doing so. While immigration has always been a sensitive topic between the two countries (see Rico 1992, Rosenblum 1998 and forthcoming, Thorup 1989), pre-election pressures in both countries make it especially unlikely that Mexico will take an official position on the bills in the current session of Congress. It seems likely that Mexico would oppose the bills for all the reasons discussed above. They offer little benefit for Mexico or Mexicans. From Mexico's perspective, a "good" guest worker bill would probably include Mexican participation in the conceptualization and implementation of a bilateral agreement.

A Strict Enforcement Regime

Current U.S. policy rhetorically emphasizes the deterrence of undocumented immigration, but fails to deter unauthorized entries. A strict enforcement regime would diversify current enforcement efforts with the goals of eliminating the job magnet that attracts undocumented immigrants to the United States, increasing the likelihood that immigrants will be apprehended and deported even after they have successfully crossed the border.

Proposal for a Strict Enforcement Regime

The INS devotes 304 staff-years to noncriminal investigations, or six staff-years per state; only 224 staff-years are devoted to work-site enforcement. Thus, work-site enforcement accounts for less than four percent of INS enforcement activities (GAO 1997, 31). In 1999, only 8,600 undocumented immigrants were arrested for deportation in the U.S. interior. ₂₇ Given an undocumented immigrant population of 5million, these trivial interior enforcement efforts imply that the vast majority of undocumented immigrants who clear the border area or enter legally and then fail to depart when their visas expire are free to live and work at will in the United States.

Thus, if the goal of immigration policy is to maximize deterrence of undocumented immigration, it is necessary to focus on workplace enforcement to eliminate the job magnet, which is the main draw for most undocumented migrants. An effective work-site enforcement program would include three components. First, INS staff-years devoted to work-site enforcement should be increased overall and as a percentage of the INS's total budget. Second, to eliminate the job magnet, it is necessary to make it costly for employers to hire undocumented migrants. The current employer sanctions impose a fine on employers of no more than $2,500 (and as little as $250) for each undocumented migrant employed for up to three separate offenses. If an employer can expect to save up to $6,000 per year, at a minimum, by employing undocumented immigrants, ₂₈ it is worth employing them even if an employer expects to be caught (which the employer does not, currently). Fining employers a figure closer to $20,000 per migrant for a first offense and a higher figure and/or jail time for subsequent offenses would be a stronger deterrent.

Third, in order for a system of employer sanctions to be effective, it will be necessary to implement a national counterfeit-proof ID system along with a database of all citizens and legal residents eligible for employment in the United States. Such databases exist in most European countries, and have been used with varying degrees of success to allow employers to check the legal status of prospective workers. A national clearinghouse would keep track of all changes in employment status and could easily identify cases in which the same ID number appears too often. ₂₉

Anticipated Effects of a Strict Enforcement Regime

Two sets of empirical cases provide insight into how a strict enforcement regime would function. First, the two most serious efforts to limit undocumented immigration into the United States occurred during the 1930s and during "Operation Wetback" in 1954. In both cases, enforcement focused on rounding up groups of immigrants, checking their documents, and deporting all those who were unable to prove citizenship. It is estimated that between 300,000 and 500,000 Mexican immigrants departed in the 1930s, and 1 million Mexicans departed during the single year 1954.

Second, the experience of European countries provides some insight into the feasibility of stronger workplace enforcement. Although most European countries (and Japan) have tough sanctions on paper, actual enforcement is weak due to low enforcement budgets, lack of political will, and liberal democratic institutions which make it difficult to punish employers (see, for example, SOPEMI 1999, 243?45; Cornelius et al. 1994; Hollifield 1992). The one possible exception to this generalization is the German case, in which workplace enforcement against undocumented immigration is integrated within a broader effort to limit "irregular" (that is, off-the-books) employment of natives and legal residents as well as undocumented immigrants (Hailbronner, Martin, and Motomura 1998, 204?207; Stobbe 2000). In fact, employers pay higher fines for evading social insurance taxes than they do for hiring undocumented immigrants (P. Martin 1994, 220). As a result, relatively few undocumented immigrants find work in Germany.

There are, however, several reasons to doubt that these "successful" models could easily be applied to the United States today. First, while both U.S. crackdowns were successful in removing and preventing undocumented immigrants, they benefited from historically unique settings. In the 1930s, the Great Depression-as well as the depression in Mexico-made it difficult and unappealing for migrants to return given the high cost of migrating (especially compared to the ease of doing so now). In the case of Operation Wetback, the crackdown on immigration was greatly facilitated by the presence of the bracero program, a large-scale guest-worker program that brought in 430,000 legal guest workers in 1954. To a large extent, Operation Wetback was successful because it substituted legal for undocumented immigration (Rosenblum forthcoming, Calavita 1992). Most importantly, in both of these cases, a significant number of legal residents were also deported when they were unable to produce their proof of residence. Success of the programs relied on wide-scale (and unconstitutional) harassment, which made it uncomfortable for immigrants to stay in the United States, a feature unlikely to be tolerated in the post-1960s era.

Second, in the case of the German comparison, work-site enforcement is facilitated by the fundamentally different approach to regulation, which includes highly centralized employment and housing systems for all residents. As Hailbronner, Martin, and Motomura (1998, 205) point out, citizens and noncitizens alike must register with the government, providing extensive information on family status, former residence, date of move, and so on each time they change their residence, a requirement that is inconceivable in the United States. Thus, cultural norms in Germany appear more permissive of workplace enforcement, and workers and employers might both conspire against such a program were it to be attempted in the United States.

How successful would a strict enforcement regime be in controlling undocumented migration? In theory, there is a level of border enforcement and a likelihood of not finding work that will deter undocumented immigration; and there is a level of fine and a probability of being caught that will deter employers from hiring undocumented workers. Achieving these enforcement levels, however, would require a massive dedication of resources and a reorientation of American attitudes about the role of the state.

Implications of a Strict Enforcement Regime for Various Economic Groups

U.S. Workers

U.S. workers would benefit from strict enforcement of work-site restrictions on immigration. While organized labor formally opposes employer sanctions, the decision by unions to adopt this position is based on the assumption that no meaningful employer sanctions will be enforced. Unions prefer no sanctions to asymmetric sanctions that benefit employers, but they would prefer strong sanctions that actually deter hiring to either of the other possibilities.

U.S. Employers

A greater emphasis on work-site enforcement shifts the costs of migration enforcement from immigrants to U.S. employers. On one hand, universal strict enforcement would help employers overcome their collective action problems, and all employers could afford to stop hiring undocumented immigrants without worrying about being undercut. But overall, U.S. employers would oppose these measures for three reasons. First, strict work-site enforcement imposes regulatory costs on employers, who then have less flexibility in their hiring decisions. Second, as the labor pool shrinks, employers would have less access to certain desirable workers and wages would rise. ₃₀ Third, as a result, immigrant-dependent import-competing or export-producing firms would be less competitive in the global market.

U.S. Consumers

It is hard to estimate with any degree of accuracy the effect of undocumented immigrants on consumer goods, but there is little question that it is significant: undocumented immigrants subsidize the production of U.S. agricultural products, manufactured goods, and services by working under difficult conditions for low wages. The price increase of these goods and services would reflect not only an increase in wages, but also the likelihood that employers pay appropriate taxes, extend additional benefits, and improve working conditions.

Other U.S. Citizens

The benefit to the U.S. population at large of eliminating undocumented immigration is also hard to estimate. On an economic level, most analysts agree that the overall effect of undocumented immigration on the tax base is marginal and perhaps slightly positive. But presumably, there are other non-economic benefits that would follow from curtailing undocumented immigration, which might include greater respect for the law and an improved civic culture.
At the same time, the social and economic costs of enforcing a strict control program like the one described here would be extreme. At the general societal level, increasing interior enforcement by a factor of, say, one hundred (i.e., 3,000 additional staff-years devoted to interior enforcement) would cost a minimum of 200 million dollars. To increase interior enforcement by a factor of one thousand, which might be necessary, would cost close to 2 billion dollars. Second, it is inevitable that the general attack on immigrants implied by the strict enforcement regime would spill over to negatively impact all residents and citizens of recent immigrant descent. This problem will be especially severe in areas with high concentrations of immigrants.

Undocumented Immigrants

The cost of this program to undocumented immigrants would be high. In fact, the thinking behind the program is specifically to set enforcement at a high enough level that would-be undocumented immigrants are deterred. Given the strength of push factors in sending states, these negatives will have to be substantial, and immigrants will be forced to choose between undesirable conditions in their countries of origin or even less desirable conditions in the United States.

Legal Immigrants

Legal immigrants would face even more discrimination than would U.S. citizens of immigrant descent. A strict enforcement program would also have an impossible-to-measure negative impact on immigrant communities who would be made to feel unwelcome and under siege. It is likely that "desirable" migrants with high social capital will also be deterred.

Other Mexican Citizens

Finally, a strict unilateral approach to immigration control would be costly to Mexico on social and economic levels, and would be damaging to U.S.-Mexican relations. Within Mexico, the sudden return of 2 million or more undocumented Mexican migrants would have a massive impact on Mexico's already fragile economy. Rising unemployment and the loss of migrant remittances would place a strain on the Mexican social welfare system, require new state spending, and encourage a balance of payments crisis. These negative impacts would fall disproportionately on Mexico's poorest citizens.

The damaging impact on the Mexican economy would have immediate economic effects in the United States as well, given that Mexico is the United States' number-two trading partner. Finally, given Mexican responses to previous enforcement efforts (see Jungmeyer 1988, Rosenblum 1998 and forthcoming), it is a near certainty that such a crackdown in the United States would provoke a general deterioration in the bilateral relationship.

A Market-based Solution to Undocumented Immigration

At the other policy extreme would be an immigration policy which sets as its goals the improvement of standards for workers in the United States and in Mexico, the reduction of enforcement costs, a guarantee of sufficient low-skilled labor in the United States, and the elimination of undocumented immigration from Mexico. All of these goals would be pursued by moving from an enforcement-based regime which seeks to deter undocumented immigration, to a market-based policy which combines the free flow of labor with improved enforcement of wage and labor standards in a North American Common Market (NACM).

Proposal for a Cooperative Market-based Regime

At the center of a cooperative approach to immigration control would be a bilateral agreement between the United States and Mexico: the United States would allow the free flow of labor, raise wages and standards for low-skilled workers, and establish or oversee a migrant remittance banking system; and Mexico would enforce improved labor standards in Mexico, invest in Mexican migrant-sending communities, and increase its police efforts at the U.S.-Mexican border to prevent the flow of drugs and (non-Mexican) undocumented immigrants. Within the NACM area of Canada, the United States, and Mexico, the flow of goods, capital, and people would be deregulated. At the same time, the NACM would adopt a universal set of labor standards, and enforcement would be monitored at the supra-national level by a NACM body including representatives of all three member states. ₃₁ In the United States, the majority of resources employed along the U.S.-Mexican border would be transferred to integrated internal oversight of immigration status, wages, and standards.

An NACM approach to immigration control implies two important changes. First, Mexican members of the NACM oversight board would be empowered to bring complaints against U.S. employers found to be discriminating against or taking advantage of Mexican migrant workers. Supranational enforcement would be designed to ensure that Mexican workers in the United States enjoy the same wage and benefit guarantees as U.S. workers. Second, for Mexicans (and U.S. workers) in Mexico, the NACM would establish a set of wages and standards comparable to those in the United States, adjusted for cost of living differences. Mexico would be required to penalize employers failing to meet these standards.

Revenues to pay for Mexican oversight and to subsidize Mexican employers who currently benefit from low standards would come from three sources. First, Mexico could reallocate current social spending because Mexican communities would receive a greater flow of migrant remittances due to rising immigrant wage rates and improved mechanisms for remittance transferal (lower transaction costs). ₃₂ Second, Mexican workers would be taxed at the same rate as their U.S. counterparts, but the majority of the Mexican workers' tax dollar would be returned to Mexico to support job creation and improved standards. Mexican migrants would only contribute taxes to the U.S. social security system, unemployment insurance, and other social welfare programs for which they would be eligible. Third, the remaining funds needed to pay for Mexican reforms would come from direct transfers from the U.S. government.

Anticipated Effects of a Market-based Proposal

What would be the likely effects of such a radical change in U.S.-Mexican immigration policy? For the United States, the key question is whether it is possible to improve working conditions in low-skilled industries, and to combine the enforcement of wages and standards with immigration eligibility (with work-site immigration enforcement against non-Mexicans). As noted above, the U.S. culture of loose regulations presents an obstacle to effective employer sanctions. But three changes implied by a NACM suggest that work-site enforcement could be improved. First, by drastically reducing the INS border presence, it would be possible to move thousands of agents into work-site enforcement with minimal new expense. ₃₃ Second, by linking DOL wage and hour oversight with immigration inspections, work-site enforcement would achieve greater economies of scale, benefiting from DOL expertise and from bringing U.S. unions into a pro-enforcement coalition. Third, by including Mexican agents through a NACM oversight committee, regulations regarding Mexican employees would also become more enforceable. The 43 Mexican consulates already spread throughout the United States could provide an infrastructure such as that which successfully represented Mexican interests during the early years of the U.S.-Mexican bracero program (see Rosenblum forthcoming, Craig 1971).

What would be the effect of a move to a NACM on undocumented immigration? By definition, the problem of undocumented immigration from Mexico would be eliminated, but several other questions must be addressed. First, what would be the effect of establishing a NACM on undocumented immigration from other source countries? It is reasonable to expect that the legalization of all Mexican workers, combined with stepped-up work-site enforcement and Mexican cooperation on the prevention of non-Mexican undocumented immigration across the U.S.-Mexican border, would reduce additional undocumented immigration. These conditions would affect both the demand for and the supply of undocumented labor. On the demand side, employers who hire undocumented immigrants are motivated the desire to pay low wages and the belief that immigrants have a better work ethic. Work-site enforcement would minimize the former incentive, and the availability of legal Mexican workers would meet the second concern. Mexican workers would have a comparative advantage over undocumented immigrants, so the demand for undocumented labor should decrease. On the supply side, combining U.S. and Mexican enforcement efforts would improve border and interior enforcement, and would allow the INS to target a group of undocumented immigrants (nonborder crossers) who are currently unregulated.

Second, what would be the effect of a NACM on labor flows from Mexico? Would the movement to a NACM greatly increase the flows of Mexicans to the United States, resulting in a costly "brain drain" and/or a loss of Mexico's best workers on the one hand, and an oversupply of labor in the United States on the other? There are at least four reasons to think this would not be the case. First, to the extent that undocumented migrants are relatively undeterred by current border enforcement efforts, dropping border enforcement would have little effect on the overall numbers of immigrants. ₃₄ Second, if it is possible to improve wages and standards in Mexico, and so to limit the incentive for Mexicans to emigrate, the demand for immigration would decrease. Indeed, the U.S. Commission for the Study of International Migration and Cooperative Economic Development (1990, xiv) concluded that economic development and job creation in sending states is the only way to diminish undocumented immigration pressure. Third, improved wage and standards enforcement would lower the demand for Mexican labor because many employers would prefer to hire better-trained U.S. workers for the same salary. Finally, many noneconomic deterrents to emigration, including the social, psychological, and cultural costs of migrating, would remain in place. ₃₅

A third question about the likely implications of a cooperative market-oriented approach to immigration is whether it is possible to improve working conditions and increase employment opportunities in Mexico, and what the effects of these efforts would be. Much has been written about the developmental approach to fighting undocumented immigration (see, for example, U.S. Commission 1990, Stalker 1994, Díaz-Briquets and Weintraub 1991), and the consensus is that this approach requires careful targeting of investment to sending communities and labor-intensive industries. Analysts also emphasize that financial aid should be accompanied by debt forgiveness, access to receiving-state markets, and other policies that maximize benefits to migrant-sending states. Finally, job creation as a tool of migration control requires a long time horizon; increased investment will not result in immediate drops in emigration.
Finally, will the movement of U.S. agents away from the border promote an increase in the flow of narcotics into the United States? U.S. agents would continue to play a role in counter-smuggling efforts; but because large-scale drug shipments tend to enter via vehicles, the expensive border strategy of deterrence and apprehensions in desert areas could be greatly reduced. Second, it is quite likely that the carrot of a NACM would be more effective in motivating intense Mexican cooperation than has been the stick of de-certification; and Mexican enforcement goals could be set as a precondition for the NACM. ₃₆

Implications of a Market-based Regime for Various Economic Groups

U.S. Workers

U.S. workers might be skeptical of an NACM, but stand to benefit enormously from it. If the free flow of labor within a NACM is not accompanied by improved U.S. enforcement of wages and standards, then this proposal would simply expand the labor pool and depress wages. Likewise, if the free flow of Mexican labor is not accompanied by job growth in Mexico, then immigration will increase with the same negative effect. But both of these implementation issues are open questions; and there is a level of work-site enforcement and Mexican job growth which makes a NACM good for U.S. workers. The qualified support that the AFL-CIO offers the WTO (pending improved standards abroad), and the strong pro-immigrant position being adopted by U.S. unions, suggest that organized labor would support cooperative immigration control.

U.S. Employers

For U.S. employers, the cost of moving to a NACM depends upon employers' dependence on low-skilled labor, on the importance of exports to overall sales, and on the degree to which employers are exporters or import-competing. Increased enforcement of wages and standards would primarily effect industries for which unskilled labor is an important component of costs. For these immigrant-dependent businesses, production costs would increase even if the labor pool expands slightly. For producers of nontradables-which include many immigrant-dependent industries such as hotel and food services, construction, landscaping, and so on-uniform enforcement would prevent collective action problems so that rising labor costs could be passed along to consumers. Thus, it is only labor-dependent exporters and import-competing industries that are hurt by rising labor costs because their international competitiveness falls. Even these businesses would not be wiped out, however, because many have already adjusted to changes in the world economy by modernizing and developing niche markets.

U.S. Consumers

For nontradable industries, it is likely that these increasing production costs would be passed largely or entirely along to consumers. Likewise, U.S. imports from Mexico (currently valued at $860 billion) would experience price increases as Mexican wages and standards are raised as a condition of the NACM establishment. Thus, to a large extent, it is consumers who would be asked to subsidize the effort to limit undocumented immigration and upgrade Mexican standards.

Other U.S. Citizens

For U.S. citizens in general, the benefits of moving to a NACM would outweigh the costs.

First, there is the great likelihood that a NACM would reduce the overall level of undocumented immigration for the reasons noted above. The higher proportion of legal migrants would raise wages, promote respect for the law, and so on. Second, cooperative immigration control would generate millions of dollars in savings compared to the current approach by obtaining Mexican cooperation at the U.S.-Mexican border and by eliminating the need for expensive migration detention facilities. ₃₇ Even if an equivalent level of enforcement is necessary in the interior, the economies of scale involved in combining INS and DOL oversight, and the greater ease of policing fixed urban work-sites rather than an expansive rural border, will generate savings. Eliminating the current game of hide-and-seek at the U.S.-Mexican border would provide noneconomic benefits for Latinos, civil libertarians, and for all property owners near the border who would no longer be subjected to INS agents or undocumented immigrants violating their property rights.

In addition to these benefits, there are several potential costs to the U.S. public. First, if the move to a NACM resulted in a large inflow of Mexican workers, the increasing strain on social services and other public goods would compromise services available to U.S. residents. If new migrants were to use a disproportionate share of public goods, the additional inflows would also result in additional tax burdens. However, given that most migrants would be young, healthy, and looking for work, it is likely that new legal migrants-like most current migrants-would be net contributors to the tax base, rather than net users, and that they would support the aging U.S. Social Security system. A second potential cost to U.S. citizens would be funding to subsidize Mexican producers and oversight agents. Although Mexico would contribute to these efforts in exchange for the benefits of participating in a NACM, it is likely that U.S. resources would be required to avoid placing excessive strain on the Mexican system. A final cost, which would be an important aspect of any debate over a NACM, would be the cultural-social cost to nativists who object to the possibility of increased integration with Mexico.

Undocumented Immigrants

For Mexicans currently in the United States illegally, the benefits of legalization are obvious. Undocumented immigrants' wages and working conditions would improve. These immigrants would no longer live in fear of apprehension and removal, but rather would be in a position to make long-term investments in their communities. Remittances by migrants to their families and source communities would increase as wages go up and as transaction costs of transmission decrease. It would also be less costly for immigrants to return to Mexico regularly, making it more likely that immigrants with new skills eventually will settle back in Mexico. The only conceivable cost to undocumented immigrants is that a larger portion of their wages will be taken in the form of taxes, but the tax losses will be outweighed by wage and service gains. Non-Mexican undocumented immigrants, on the other hand, are clear losers under this plan, as the likelihood of apprehension and removal would sharply increase with enforcement moving from the U.S.-Mexican border to the interior.

Legal Immigrants

Legal immigrants would enjoy similar benefits to undocumented immigrants. Wages and working conditions would improve throughout low-skilled sectors of the economy. Legal Mexican immigrants would further benefit from the reduction in anti-immigrant hostility and from the further integration of the U.S. and Mexican economies. While non-Mexican legal immigrants would share the benefits of increasing wages, they might suffer from the transposition of current anti-Mexican hostility to other ethnic groups.

Other Mexican Citizens

Moving to a NACM would not be without cost to Mexico. The greatest obstacle from within Mexico would be the mandate that wages and standards be raised to a level proportional to the United States. Mexico has spoken out against previous efforts to "export U.S. values," claiming (not without merit) that less developed countries deserve to exploit their comparative advantage in low-cost labor. Second, Mexico would be asked to expend greater national resources on border policing and on investing in migrant-sending communities. And third, the possibility exists that successful efforts by Mexico to raise wages and standards would result in decisions by would-be investors to avoid Mexico in favor of less well-regulated countries.

But these costs would be outweighed by the benefits of a NACM. Each of the costs identified above would be small. First, participation in a NACM would not require that Mexico raise standards to the U.S. level, but rather that they reach a proportional level, controlling for cost of living. Second, border policing efforts would not be politically costly since they would not regulate Mexican emigration; and investment in sending communities would be politically popular, especially if accompanied by U.S. matching funds. And third, while it is possible that raising Mexican standards might discourage some investors, the support from the United States implicit in a NACM would attract others, and the net effect of a NACM on investment would be positive.

Mexico would also reap important benefits from the NACM. On an economic level, Mexico would benefit from increased remittances as discussed above and from the elimination of transfers from migrants to smugglers, an illicit industry generating 100 million dollars each year. The greater ease of return and circular migration, noted above, would also increase the flow of resources into Mexico. Factor price equalization would accelerate, and the entire Mexican economy would benefit from closer economic integration with the larger United States.

Finally, the Mexican government would reap political benefits from reaching a NACM agreement. First, the abandonment of the U.S. border enforcement strategy of pushing undocumented immigrants into inhospitable desert and mountain terrain would reverse the disastrous trend of increasing border deaths since the early 1990s. Mexico's inability, thus far, to protect emigrants from these dangers has been costly both in terms of loss of life and on a domestic political level. All three major Mexican political parties and their constituents consider the protection of migrants' rights an important foreign policy goal (Rosenblum 1988 and forthcoming). The creation of a NACM would offer concrete steps in that direction.

Conclusions: Recommendation and Feasibility

In evaluating the relative benefits of competing policy options, I consider the likely effectiveness and cost efficiency of the different proposals. Distributive effects raise questions about the political feasibility of the plan that otherwise appears most promising: the bilateral market-oriented approach.

Effectiveness and Efficiency

Unilateral Enforcement Regimes

The first three plans discussed rely on some combination of unilateral border control, employer sanctions, and (in all but the strict enforcement regime) some type of guest worker program. This basic approach fails for at least three reasons. First, unilateral enforcement-especially when concentrated at the border-fails to recognize the forces of globalization and, in particular, the effects of increasing U.S.-Mexican integration (see Sassen 1998, among others). The worldwide trends of increasing flows of goods and services, falling costs of travel, and growing transnational communities make it more difficult for countries to prevent immigration through border enforcement. The United States and Mexico have implemented a free trade area, and Mexico has become the United States' second largeest source of imports and exports; more goods flow across the U.S.-Mexican border than any other land border in the world. It is simply not possible to prevent people from crossing that border as well. Border enforcement fails to actually deter migration, while damaging bilateral relations and slowing economic integration.

The second problem with all three of these unilateral enforcement approaches is that they create opponents of control, but fail to create a political constituency in favor of immigration control. At the workplace, the one thing employers, workers, and labor unions all agree on is a desire to thwart INS enforcement efforts, though for different reasons. The only individuals likely to help enforce immigration controls at the work-site are disgruntled employees and union-busting employers. Nor is there a powerful constituency for border and non-work-site interior control. Civil libertarians oppose the invasive steps required to make such controls effective; and Latinos and property owners near the border have lost patience with Border Patrol efforts to prevent crossing. The only groups inherently supportive of border-oriented enforcement are INS bureaucrats and anti-immigration activists.

The third problem with the unilateral approaches is that they discourage cooperation among executive branch agencies. The INS itself is given a schizophrenic mission of providing services to immigrants while simultaneously apprehending and deporting them. The DOL and other federal agencies devoted to protecting workers often find themselves hampered by immigrants' fear of the federal government.

In sum, none of the unilateral enforcement regimes is likely to be effective, and comparing them is a matter of estimating which is the least of three evils. The worst option would be a strict enforcement regime along the lines I have described. Producing results through strict unilateral enforcement would be far too costly in terms of federal spending, the rights of migrants and U.S. residents, and U.S.-Mexican relations. Among the other options, it would clearly be a mistake to maintain the current system while adding the proposed "guest worker reform" provisions. The guest worker bills fail to address the existing shortage of agricultural labor, yet they are likely to have the unfortunate effects of lowering wages for U.S. and non-U.S. workers while encouraging additional undocumented immigration. This leaves the current system, expensive and ineffective as it is, as the best option among unilateral policies.

Bilateral Market-based Enforcement

In contrast, a cooperative, market-based solution embraces the reality of economic integration and makes the most of the U.S.-Mexican relationship. Deepening integration has not only lowered the cost of immigration, it has also created opportunities for bilateral cooperation. The existence of NAFTA and of current bilateral immigration institutions, including the Border Liaison Mechanisms and Mechanisms of Consultation (see Rosenblum 1999 and forthcoming), lower the costs to both countries of building new cooperative immigration institutions. Mexico's expanded importance as a U.S. trading partner raises the benefit to the United States of sharing the costs of raising Mexican employment standards. Economic integration means that both countries benefit from eliminating the deadweight losses of immigrant transfers to smugglers in favor of increased consumption and production. ₃₈

A second general advantage to the bilateral, market-oriented approach is that the focus on interior enforcement creates a powerful coalition of pro-enforcement interest groups. In particular, labor unions, immigrant and non-immigrant workers, and Mexican Consuls would all support DOL enforcement of wages and standards. Third, the focus on interior enforcement would encourage cooperation among federal agencies, allowing the DOL to cooperate with immigration enforcement agents instead of opposing them. The INS would be freed up, consistent with reform efforts currently pending in Congress, to focus more attention on visa and naturalization services. Responsibility for border enforcement could be concentrated in the hands of traditional enforcement agents, such as the Customs Service. Such a move that would foster bilateral cooperation since agents would share common enemies on both sides of the border (for example, smugglers), unlike under the current system (in which the United States is mainly focused on immigrants).

Fourth, in contrast to unilateral enforcement, a system that focused on deregulating the bilateral labor market would allow market forces to determine the supply of labor. In a segmented economy with a finite number of unskilled jobs, legal Mexican immigrants would have a comparative advantage for these positions, reducing the job magnet for other undocumented immigrants, freeing up resources to protect labor standards, and increasing the flow of remittances to Mexico. Thus, only the market approach is likely to achieve the twin goals of legalizing the Mexican portion of the undocumented flow and limiting undocumented immigration from other source countries.

The Politics of Immigration Reform

If a cooperative approach to immigration policy-making would result in equal or better outcomes for most affected actors, why is it the one solution least discussed by U.S. immigration policy-makers? The answer to this question is threefold: the distributive effects of immigration policy benefit powerful groups under the current system, the current system is stable for non-distributive reasons, and there are several as yet undiscussed obstacles to moving toward a NACM. On the other hand there are reasons to expect that the political feasibility of a NACM may be increasing.

Obstacles to Immigration Policy Reform

The stability of the current unilateral enforcement regime is explained, in part, by Olsonian interest group politics. First, the beneficiaries of the current regime are a combination of a powerful privileged group (agribusiness) and diffuse consumers who are not mobilized on the issue, but who are present in every single district in the country. In contrast, opposition to the current regime has consisted of a disenfranchised latent group: undocumented immigrants. Only recently has an economic interest group with any organizational strength (labor unions) dropped its support for the existing system, and labor remains somewhat divided over how forcefully to reject employer sanctions. Finally, the INS itself has a vested interest in maintaining the status quo, and with 20,000 employees and a $4.1 billion budget the INS is a non-trivial bureaucratic actor.

There are several additional obstacles to policy reform. At least two general features of immigration policy-making suggest that a cooperative solution is unlikely. First, immigration politics is subject to policy inertia, so it is difficult to implement substantial policy reform until reform is overdue. As a result, by the time major immigration reform occurs, the issue is highly salient in domestic politics. High salience raises the political stakes for Congress and the president, making it difficult to bring international issues into the debate (Rosenblum forthcoming). A second obstacle to bilateralism is the role of racism in immigration policy-making. None of the cost-benefit calculations I have described in this paper attempt to take the racism of interest groups and/or policy-makers into account. There is no question, however, that racism has played a consistent and fundamental role in shaping U.S. immigration policy (see, for example, Loescher and Scanlan 1986, Divine 1956) and would be a hindrance to forming a NACM.

In addition to these general obstacles, at least three specific problems make enacting a bilateral immigration policy difficult. First, international cooperation is always risky (Axelrod 1984, Keohane 1986, Snidal 1986), and a cooperative effort which includes a U.S. commitment to demilitarize its southern border would carry special risks. In short, if Mexico were to fail to complete its promised responsibilities, the United States could be vulnerable to a massive influx of new unwanted immigrants. This is a contracting problem that can be solved through monitoring, but doing so raises the cost of a bilateral agreement. A second case-specific obstacle is the long time horizon inherent in a cooperative solution. It would take years, at best, before increased investment in Mexico would lead to a reduction in immigration push factors (Díaz-Briquets and Weintraub 1991, U.S. Commission 1990, xvi), and it could take equally long before the message that the United States is no longer welcoming to non-Mexican undocumented immigrants filters back to other sending communities. Third, the creation of a NACM raises questions about limiting the free flow of labor to the NAFTA area: why should Central America and the Caribbean be excluded? Why should China be excluded? There is no obvious answer to this last point, and it is likely that if the Clinton goal of expanding free trade throughout the Americas is ever to come to fruition, then a common market restricted to current NAFTA countries could become untenable.

Possibility of Immigration Reform

Despite these obstacles, there are several reasons for optimism about the possibility of a cooperative approach to immigration control. The first, and most obvious, reason is that the current system is a failure, and a growing number of policy-makers recognize that throwing more money at the Border Patrol, building more fences along the border, and denying more services to migrants have little influence on the strong forces affecting immigration. Increased economic integration between the United States and Mexico assures that these forces will remain in place, and makes a border-oriented immigration policy increasingly anachronistic. Second, the political landscape in the United States is evolving in a way that is favorable to a bilateral approach to policy-making. Five, ten, or twenty years from now, when a NACM is in place, analysts will look back to the February 2000 vote by the AFL-CIO to oppose employer sanctions as a turning point in its development. Equally important is the growing participation and political power of Mexican-Americans as an ethnic group (González Gutiérrez 1993, Perez-Godoy 1999). In addition, immigrants are increasingly active in politics in the United States and Mexico, and they explicitly focus transnational political power on immigration policies. ₃₉

Finally, the government of Mexico and other migrant-sending states are increasingly willing to participate directly in the U.S. political process, an effort that has been taboo until recently. In Mexico's case, high-level actors in the Foreign Ministry describe a learning process after Mexico's "mistake" in not lobbying against the IRCA, a process which resulted in Mexico's successful, full-scale lobbying campaign in support of the NAFTA agreement. ₄₀ In pursuit of these goals, Mexico has also begun explicitly encouraging Mexican-Americans and Mexican immigrants to become more active in the U.S. political process through the so-called Program for Mexican Communities Abroad. Finally, the Mexican government has several attractive concessions it can offer the United States in exchange for moving to a NACM, including the lifting of restrictions on U.S. banks wishing to operate in Mexico (a move that Mexico successfully resisted during NAFTA negotiations) and a major expansion of drug control efforts. Thus, a bilateral approach to immigration policy-making may be closer than it seems.

References

Axelrod, Robert. The Evolution of Cooperation (New York: Basic Books, 1984).

Binational Commission. Binational Study on Migration Between Mexico and the United States (Mexico City: Secretaría de Relaciones Exteriores, 1997).

Calavita, Kitty. Inside the State: The Bracero Program, Immigration and the INS (New York: Routledge, 1992).

_____. "U.S. Immigration and Policy Responses: The Limits of Legislation." In Wayne A. Cornelius, et al., eds., Controlling Immigration (Stanford, Calif.: Stanford University Press, 1994).

Chang, Howard. "Migration as International Trade: The Economic Gains from the Liberalized Movement of Labor." Journal of International Law and Foreign Affairs 3 (1998).

"The United States' Demand for Mexican Labor." In Wayne A. Cornelius and Jorge A. Bustamante, eds. Mexican Migration to the United States (La Jolla, Calif.: UCSD Center for Mexican Studies, 1989): 25?47.

Cornelius, Wayne A. "The Structural Embeddedness of Demand for Mexican Immigrant Labor: New Evidence from California." in Marcelo Suárez-Orozco, ed., Immigration and the Socio-Cultural Remaking of the North American Space (Cambridge, Mass.: Harvard University Press for the David Rockefeller Center for Latin American Studies, 1998).

Cornelius, Wayne A., Philip L. Martin, and James F. Hollifield, eds. Controlling Immigration: A Global Perspective (Stanford, Calif.: Stanford University Press, 1992).

Craig, Richard B. The Bracero Program: Interest Groups and Foreign Policy (Austin: University of Texas Press, 1971).

Delgado, Héctor L. New Immigrants, Old Unions: Organizing Undocumented Workers in Los Angeles (Philadelphia, Penn.: Temple University Press, 1993).

Díaz-Briquets, Sergio, and Sidney Weintraub, eds. Regional and Sectoral Development in Mexico as Alternatives to Migration (Boulder, Colo.: Westview Press, 1991).

Dunn, Timothy. The Militarization of the U.S.?Mexico Border, 1978?1992 (Austin: CMAS Books, University of Texas Press, 1996).

Farmworker Justice Fund. "The Agribusiness Proposal for a New 'Guest worker' Program Should be Rejected." Available at http://www/clraf.org/gwanalys.htm.

Fix, Michael, ed. The Paper Curtain: Employer Sanctions Implementation, Impact, and Reform. (Washington, D.C.: Urban Institute Press, 1991).

González Gutiérrez, Carlos. "The Mexican Diaspora in California: Limits and Possibilities for the Mexican Government." In Abraham Lowenthal and Katrina Burgess, eds. The California-Mexico Connection (Stanford, Calif.: Stanford University Press, 1993).

Hailbronner, Kai, David Martin, and Hiroshi Motomura. Immigration Controls: The Search for Workable Policies in Germany and the United States (Providence, R.I.: Breghahn Books, 1998).

Hamermesh, David, and Frank Bean, eds. Help or Hindrance? The Economic Implications of Immigration for African Americans (New York: Russell Sage Foundation, 1998).

Haus, Leah. "Openings in the Wall: Transnational Migrants, Labor Unions, and U.S. Immigration Policy." International Organization 49 (1995): 285?313.

Hollifield, James. Immigrants, Markets, and States: The Political Economy of Postwar Europe. (Cambridge, Mass.: Harvard University Press, 1992).

Jacobs, Eva. Handbook of U.S. Labor Statistics. 3d ed. (Lanham, Md.: Bernan Press, 1999).

Jungmeyer, Robert. "The Bracero Program, 1942?1951: Mexican Contract Labor in the United States." Ph.D. diss., University of Missouri at Columbia, 1988.

Keohane, Robert O. After Hegemony: Cooperation and Discord in the World Political Economy (Princeton, N.J.: Princeton University Press, 1986).

Loescher, Gil, and John A Scanlan. Calculated Kindness: Refugees and America's Half-Open Door, 1945 to the Present. (New York: The Free Press, 1986).

Lorey, David. United States?Mexico Border Statistics Since 1900. (Los Angeles: UCLA Latin American Center Publications, 1990).

Marcelli, Enrico, and David Heer. "The Unauthorized Mexican Immigrant Population and Welfare in Los Angeles County: A Comparative Statistical Analysis." Sociological Perspectives 41 (1998): 279?302.

Martin, David A. "Disentangling the Strand of U.S. Immigration Policy Reform." In Wayne A. Cornelius et al., eds., Controlling Immigration (Stanford, Calif.: Stanford University Press, 1994). Martin, Philip L. "Good Intentions Gone Awry: IRCA and U.S. Agriculture." Annals, AAPSS 534 (1994): 44?57.

Massey, Douglas, Joaquín Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino, and J. Edward Taylor. "Theories of International Migration: A Review and Appraisal." Population and Development Review 19 (1993): 431?66.

_____. "An Evaluation of International Migration Theory: The North American Case." Population and Development Review 20 (1994): 699?751.

McCain, Johnny M. "Contract Labor as a Factor in United States?Mexican Relations, 1942?1947." Ph.D. diss., University of Texas at Austin, 1970.

Ness, I. "Organizing Immigrant Communities: UNITE's Workers Center Strategy." In Bronfenbenner et al., eds., Organizing to Win (Ithaca, N.Y.: Cornell University Press, 1998).

Perez-Godoy, S. Mara. "Social Movements and International Migration: The Mexican Diaspora Seeks Inclusion in Mexico's Political Affairs, 1968?1998." Ph.D. diss., University of Chicago, 1999.

Portes, Alejandro, ed. The Economic Sociology of Immigration (New York: Russell Sage Foundation, 1995).

Rico, F. Carlos. "The Making of U.S. Policy Toward Mexico: Should We Expect Coherence?" In Rosario Green and Peter H. Smith, eds., Foreign Policy in U.S.-Mexican Relations (La Jolla, Calif.: UCSD Center for U.S.-Mexican Studies, 1989).

_____. "Migration and U.S.-Mexican Relations, 1966?1986." In Christopher Mitchell, ed. Western Hemisphere Immigration and United States Foreign Policy (College 1992).

Rosenblum, Marc R. "My Neighbor, Myself: Mexican Influence on U.S. Migration Policy." Journal of International Law and Foreign Affairs 3 (1998).

_____. "Immigration as Foreign Policy: The International Political Economy of U.S. Migration Policy-making." Paper presented at the Annual Meeting of the American Political Science Association, Atlanta, Georgia, Septermber 1999.

_____. "Abroad and at Home: The Foreign and Domestic Sources of U.S. Migration Policy." Ph.D. diss., University of California, San Diego. Forthcoming.

Sassen, Saskia. Globalization and Its Discontents (New York: The New Press, 1998).

Schuck, Peter. "The Re-Evaluation of American Citizenship." In Christian Joppke, ed. Challenge to the Nation-State: Immigration in Western Europe and the United States (New York: Oxford University Press, 1998).

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

Snidal, Duncan. "The Game Theory of International Relations." In Kenneth A. Oye, ed., Cooperation Under Anarchy (Princeton, N.J.: Princeton University Press, 1986).

SOPEMI. Trends in International Migration (Paris: OECD, various years).

Stalker, Peter. The Work of Strangers: A Survey of International Labour Migration (Geneva: International Labour Organization, 1994).

Stobbe, Holk. 2000. "Undocumented Migration in the United States and Germany: An Analysis of the German Case with Cross-reference to the U.S. Situation." La Jolla: UCSD Center for Comparative Immigration Studies.

Stockwin, Will. "H2 Reform Bill in Senate Addresses Farm Labor Supply." AgAlert, 3 November 1999.

Thorup, Cathryn. "U.S. Policy Toward Mexico: Prospects for Administrative Reform." In Rosario Green and Peter H. Smith, eds., Foreign Policy in U.S.-Mexican Relations (La Jolla, Calif.: UCSD Center for U.S.-Mexican Studies, 1989).

U.S. Commerce Department. Statistical Abstract of the United States (Washington, D.C.: U.S. Government Printing Office, various years).

U.S. Commission on Immigration Reform. U.S. Immigration Policy: Restoring Credibility (Washington, D.C.: U.S. Government Printing Office, 1994).

_____. Becoming an American: Immigration and Immigrant Policy (Washington, D.C.: U.S. Government Printing Office, 1997).

U.S. Commission for the Study of International Migration and Cooperative Economic Development. Unauthorized Migration: An Economic Development Response (Washington, D.C.: U.S. Government Printing Office, 1990).

U.S. General Accounting Office. "H-2A Agricultural Guest Worker Program: Changes Could Improve Services to Employers and Better Protect Workers (Washington, D.C.: U.S. Government Printing Office, 1997).

_____. "Illegal Immigration: Status of Southwest Border Strategy" (Washington, D.C.: U.S. Government Printing Office, 1999).

U.S. Immigration and Naturalization Service. Statistical Yearbook of the Immigration and Naturalization Service (Washington, D.C.: U.S. Government Printing Office, various years).

U.S. Select Commission on Immigration and Refugee Policy. Final Report (Washington, D.C.: U.S. Government Printing Office, 1981).

Waldinger, Roger, and Mehdi Bozorgmehr. Ethnic Los Angeles (New York: Russell Sage Foundation, 1996).

Watts, Julie. "An Unconventional Brotherhood: Why Spanish, Italian and French Unions Support More Open Immigration Policies." Unpublished manuscript, 2000.

Notes

Note 1: The limited recent efforts at bilateral policy-making have been important, however. In particular, the U.S.-Mexican Binational Commission has been very successful at building local institutions at the border which have mitigated the harsh effects of Congressional "restrictionism" and insulated the U.S.-Mexican relationship against the possible fallout from North American anti-immigrant hysteria (Rosenblum 1999 and forthcoming). Back.

Note 2: Dunn (1996, 40) argues that Carter favored delaying an increase in enforcement efforts pending the publication of the final report by the Select Commission on Immigration and Refugee Policy. I believe more important factors were Carter's effort to deepen overall cooperation between the United States and Mexico and his unsuccessful effort to establish cooperative border enforcement policies (Rosenblum forthcoming). Back.

Note 3: Legislation making it illegal to harbor, shelter, or transport undocumented immigrants was first passed in 1952 to encourage participation in the U.S.-Mexican bracero program, but the so-called Texas proviso explicitly stated that employing undocumented immigrants did not fall within the proscribed acts (see Calavita 1992, Rosenblum forthcoming). Back.

Note 4: The IRCA also included the "carrot" of amnesty for three different groups of undocumented immigrants based on their proven residence in the United States since 1982 or their proven employment in U.S. agriculture for 90 days in each of three previous years. I do not include this as an aspect of the current policy regime since it was a "one-time" policy tool. Back.

Note 5: INS Associate Commissioner Robert Bach, as quoted in Louis Uchitelle, "INS Looks the Other Way on Illegal Immigrant Labor," The New York Times, 9 March 2000. Back.

Note 6: Sam Dillon, "Agua Prieta Journal: Boom Turns Border to Speed Bump." The New York Times, 18 January 2000. Back.

Note 7: These jobs are low-wage, and are referred to as 3D jobs because they are dirty, dangerous, and difficult. In the U.S. case, immigrant-dominated sectors of the economy include agriculture, hotel and restaurant, construction, low-skilled manufacturing, food processing, and maintenance and cleaning. Even to the extent that immigrants are becoming widely distributed throughout the economy, they still occupy migrant niches because many are employed by migrant entrepreneurs who prefer to hire co-ethnics (Waldinger and Bozorgmehr 1996, Cornelius 1998). Back.

Note 8: U.S. immigration policy has also addressed the availability of high-skilled labor: the 1990 Immigration Act created 140,000 immigrant visas distributed to (mostly skilled) individuals meeting labor demands, and there is currently a debate about expanding the H-1B high-skilled temporary visa program from 115,000 to 200,000 visas. I do not address issues surrounding high-skilled immigration in this paper. Back.

Note 9: Uchitelle, "INS Looks the Other Way on Illegal Immigrant Labor." Back.

Note 10: See, for example, Ruben Rosario, "Illegal Hotel Workers Both Win and Lose," St. Paul (MN) Pioneer, 9 January 2000. Back.

Note 11: As Watts (2000) shows, similar transformations in unions' preferences have occurred in Spain, Italy, and, to a lesser extent, France. Back.

Note 12: See Timothy Burn, "Unions Halt Decline in Membership," Washington Times, 20 January 2000; Ruben Rosario, "Illegal Hotel Workers Both Win and Lose"; and Maureen Fan, "Unions' Latest Organizing Push: Immigrants," San Jose Mercury News, 21 January 2000. Back.

Note 13: See full text at http://www.aflcio.organize/publ/estatements/feb2000/immigr.htm. Back.

Note 14: Marisa Taylor, "INS Struggles with Criminals, Crowding," San Diego Union-Tribune, 21 February 2000. Back.

Note 15: See Philip Pan, "INS Chief Puts Office on New Path," The Washington Post, 10 November 1999; and Marjorie Valbrun, "Border Patrol Job (Yawn) Isn't What It Used to Be," The Wall Street Journal., 4 January 2000. Back.

Note 16: Jim Yardley, "Some Texans Say Border Patrol Singles Out Too Many Blameless Hispanics," The New York Times, 26 January 2000 Back.

Note 17: See Ken Ellingwood, "U.S. Residents, Border Staff Clash." Los Angeles Times, 21 January 2000; and Ignacio Ibarra, "Rancher Questions Border Patrol Tactics After Cow is Shot." The Arizona Daily Star, 10 February 2000. Back. 18

Note 18: Yardley, "Some Texans Say Border Patrol Singles Out Too Many Blameless Hispanics." Back.

Note 19: Mexican consuls were especially active in bracero contract oversight between 1942-1947 and 1951-1954, and bracero contracts were well enforced during these periods. In 1954, the United States insisted on renegotiating the bilateral agreement to strip consuls of their oversight ability (Rosenblum forthcoming). Back.

Note 20: Resolution passed by the Mexican Chamber of Deputies, 1986. Back.

Note 21: Author interview with Frederico Salas, Mexican Foreign Ministry, 26 June 1998; see also Perez-Godoy 1999). Back.

Note 22: Congressional Record 1999, S13543. Back.

Note 23: Carlos Guerra, "More 'Legal' Farm Workers a Cruel Hoax," San Antonio Express News, 24 January 2000. Back.

Note 24: The INS and DOL consider agricultural labor exceptionally difficult to oversee because production occurs in open spaces, involves numerous workers, and is located at great distances from INS and DOL field offices (GAO 1997, 34), exacerbating the weakness of H-2A enforcement. Back.

Note 25: Specifically, the proposal would eliminate the 50 percent rule which requires that employers hire qualified locals who appear before the season is half over; eliminate the "positive recruitment" requirement that requires employers to seek out U.S. workers; limit required recruitment to those workers enrolled in a state registry within the same state as the employer; and limit the registry search, in some cases, to fourteen days before allowing growers to recruit guest workers over the next six months. Back.

Note 26: Currently, employers participating in the H-2A program must offer the highest of three wages: the federal or state minimum, the local "prevailing wage" defined by the DOL, or the H-2A adverse effect wage rate (AEWR), which is defined as the regional average wage for field and livestock workers from Department of Agriculture surveys. Under Reagan, the methodology for calculating the AEWR was revised so that the AEWR fell by about twenty percent in most cases, although it is still usually higher than the prevailing wage. Under the proposed changes in S.1814, the AEWR would be redefined as the local prevailing wage plus five percent, up to a maximum of the old AEWR, so S.1814 would lower the wage paid to many H-2A workers and not raise the wage for any of them. Back.

Note 27: Uchitelle, "INS Looks the Other Way on Illegal Immigrant Labor." Back.

Note 28: This figure assumes employers save three dollars per hour, 40 hours per week, by paying lower wages and avoiding unemployment, workers' compensation, and other payments. Actual savings are probably greater. Back.

Note 29: As a side benefit, the existence of a national ID card that all citizens would be required to carry would also facilitate stop-and-investigate operations b29INS and Border Patrol officials away from the work site. Back.

Note 30: According to many employers of undocumented immigrants, wages are not the primary motivation, and so rising wages would not be the biggest cost of strict enforcement: 54 percent of 114 San Diego County employers interviewed believe that immigrants, in general, are harder working than native-born workers (Cornelius 1998). Back.

Note 31: While an NACM would presumably include the United States, Mexico, and Canada, this paper does not deal with the effects of labor flows to and from Canada. Back.

Note 32: Western Union and other money transfer services have fraudulently failed to disclose rates of commission on international transfers to Mexico, with some migrants paying rates of twenty percent or higher. Even if no other aspects of a NACM are established, the United States and Mexico would both benefit from developing an improved system for transferring migrant savings back to Mexico. Back.

Note 33: In 1999, the INS enforcement budget was $1.93 billion dollars, while the DOL's Wage and Hour Division's 1997 budget was $118 million. Shifting half or two-thirds of the INS resources to work-site enforcement would have a dramatic effect on the DOL's ability to enforce labor standards. Back.

Note 34: The fact that many migrants attempt entry repeatedly until successful suggests that this is the case. Back.

Note 35: One empirical comparison is the case of the creation of the European Union: to a surprising degree, immigration rates between countries in Europe remained virtually unchanged (SOPEMI 1993-1999). However, the comparison is imperfect for two reasons: the wage gap is smaller, along the order of four-to-one between high and low wages countries in Europe compared to eight- or nine-to-one between the United States and Mexico; second, the cultural obstacles to migration are probably higher in Europe, with European countries demonstrating lower rates of internal migration than the United States or Mexico. Back.

Note 36: As with immigration, drug enforcement focuses overwhelmingly on the border (and on sending states) because these enforcement mechanisms are least costly to U.S. consumers. The creation of a NACM would be opposed by some of the same forces opposed to drug legalization. An integrated approach that encouraged Mexican cooperation and discouraged U.S. consumption would have the best chance of success. Back.

Note 37: In 1999, the INS estimated the need for over 1 million detention-days, over two-thirds of which were for Mexicans. Back.

Note 38: From an economic perspective, if migrants are simply a mobile factor of production, then imposing quotas—or immigration controls—is the least efficient way to regulate production, resulting in deadweight losses and economic distortions. At a minimum, it is more efficient to impose a tariff (to charge immigrants for the right to enter the United States), rather than impose a quota (Chang 1998). Charging Mexican immigrants a figure roughly equivalent to the fee now charged by coyotes would have little or no effect on immigration flows, but would ensure that the costs of migration are reinvested in the United States and Mexico (assuming the payment was divided between the two states) rather than simply enriching smugglers. Optimal tariff theory would allow policy-makers to set a fee-for-entry level that would result in the ideal level of unskilled inflows. Back.

Note 39: See Terri Langford, "Hispanic Leaders Seek Solutions to Mexican Expatriates Problems," The Dallas Morning News, 6 February 2000; and Alfredo Corchado, "Mexican Immigrants' Group to Meet in Dallas," The Dallas Morning News, 4 February 2000. Back.

Note 40: Author interview with Carlos Rico, Director General for North America, Mexican Ministry of Foreign Affairs, Mexico City, 8 July 1998. Back.