Beyond The Border Buildup: Towards A New Approach To Mexico-US Migration
A proper understanding of the causes of international migration suggests that punitive immigration and border policies tend to backfire, and this is precisely what has happened in the case of the United States and Mexico. Rather than raising the odds that undocumented immigrants will be apprehended, U.S. border-enforcement policies have reduced the apprehension rate to historical lows and in the process helped transform Mexican immigration from a regional to a national phenomenon. The solution to the problems associated with undocumented migration is not open borders, but frontiers that are reasonably regulated on a binational basis.
Among the findings of this report:
Before September 11, 2001, President Bush and President Vicente Fox of Mexico appeared to be moving toward an agreement to manage Mexican labor migration by expanding the quota for legal immigrants, creating a reasonable temporary worker program, facilitating the return of migrants and the investment of their dollars in Mexico, and regularizing the status of undocumented Mexicans in the United States. Unfortunately, the hijackers derailed this negotiation and President Fox was left standing at the border looking northward with his hand extended as President Bush turned his back to launch the War on Terror.
In the end, the United States must learn that national security involves more than toppling ruthless dictators in distant lands. It also requires attending to the political stability and economic security of a country of 100 million people with whom we share a 2,000-mile border. The administration’s inattention to migration in the context of North American integration has undermined the stature and standing of Mexico’s first democratically elected President in 70 years; and every day that passes without a labor agreement makes it more difficult for Mexico to realize its full potential for economic growth.
Rather than accepting immigration as a logical consequence of America’s dominant position at the core of a global market economy, U.S. political leaders have enacted repressive unilateral policies that seek to create the impression that immigration is not occurring, that U.S. borders are “under control,” and that U.S. citizens are protected from the presumed ill effects of immigrants. In fact, such policies achieve the opposite: immigration continues, but in a way that undermines the status and welfare of U.S. residents and immigrants alike.
Indeed, a proper understanding of the causes of international migration suggests that punitive immigration and border policies tend to backfire, and this is precisely what has happened in the case of the United States and Mexico. Rather than raising the odds that undocumented immigrants will be apprehended, U.S. border-enforcement policies have reduced the apprehension rate to historical lows and in the process helped transform Mexican immigration from a regional to a national phenomenon. Rather than discouraging Mexicans from coming, these policies have induced millions of legal immigrants to become U.S. citizens, thereby allowing them to petition for the entry of relatives under U.S. immigration law. The net result has been an unprecedented increase in the number of Mexicans living north of the border.
Buildup at the Border
In 1986 the United States embarked on a determined effort to restrict Mexican immigration and tighten border enforcement. The arrival of this new era was heralded by passage of the Immigration Reform and Control Act (IRCA) in October of that year, which decisively expanded funding for the U.S. Border Patrol. The renewed effort at border enforcement intensified around 1994, just as the North American Free Trade Agreement (NAFTA) took effect and promoted greater integration of the Mexican and U.S. economies. During the 1980s, border control was framed by U.S. politicians as an issue of “national security” and illegal migration was portrayed as an “alien invasion.” Between 1986 and 1996, Congress and the President undertook a remarkable series of restrictive actions to reassure citizens that they were working hard to “regain control” of the Mexico-U.S. border.
Despite expectations that IRCA would somehow slow unsanctioned Mexican migration, both legal and illegal immigration from Mexico continued to rise, and Congress returned to the drawing board in 1990 to pass another revision of U.S. immigration law. This legislation focused strongly on border control and authorized even more funds for the hiring of additional Border Patrol officers. Early in the Clinton administration (1993-94), the agency developed a new border-enforcement strategy that took full advantage of this increased funding. Known as “prevention through deterrence,” the strategy aimed to prevent Mexicans from crossing the border in key sectors in order to avoid having to deport them later. The strategy originated in September of 1993, when the Border Patrol Chief in El Paso, Texas, launched Operation Blockade – an all-out effort to prevent illegal border-crossing within the El Paso sector. Within a few months, immigrants had been induced to go around the imposing wall of enforcement, and traffic through El Paso itself was reduced to a trickle.
Officials in Washington took note of the favorable outcome in El Paso and incorporated this approach into the Border Patrol’s national strategic plan for 1994. In October of that year, a second mobilization was authorized for the busiest sector of the border: San Diego, California. Operation Gatekeeper installed high-intensity floodlights to illuminate the border day and night and built an eight-foot high steel fence along 14 miles of border from the Pacific Ocean to the foothills of the Coastal Range. Border Patrol officers were stationed every few hundred yards behind this formidable steel wall, and a new array of sophisticated hardware was deployed in the no-man’s-land it faced.
This buildup of enforcement resources was further accelerated by Congress when it passed the Illegal Immigration Reform and Immigrant Responsibility Act of 1996 (IIRIRA). Once again, the legislation focused heavily on deterrence, authorizing funds for the construction of two additional layers of fencing in San Diego and enacting tougher penalties for smugglers, undocumented migrants, and visa over-stayers. It also included funding for the purchase of new military technology and provided funds for hiring 1,000 Border Patrol agents a year through 2001 to bring the total strength of the Border Patrol up to 10,000 officers.
The effect of these successive policy actions on border enforcement is illustrated in Figure 1, which shows increases relative to 1986 in the budget of the Border Patrol and the Immigration and Naturalization Service (INS), now Citizenship and Immigration Services (CIS) of the Department of Homeland Security. In 1986, the INS budget stood at just $474 million and that of the Border Patrol was $151 million. IRCA began the acceleration of funding for border enforcement, but it was the innovation of border blockades in 1993 that really opened the spigot of money. By 2002 the Border Patrol’s budget had reached $1.6 billion and that of the INS stood a $6.2 billion, ten and thirteen times their 1986 values, respectively. With this additional revenue, more Border Patrol officers were hired, as shown in Figure 2. Between 1986 and 2002 the number of Border Patrol officers tripled and the number of hours they spent patrolling the border (“linewatch” hours) grew by a factor of around eight.
Border Enforcement Backfires
The fundamental weakness of blockading particular sectors of the Mexican border is that there are always other, less-defended sectors within which to cross. The mobilization of enforcement resources in El Paso and San Diego simply diverted the flow of migrants into Arizona, causing U.S. authorities to launch new blockades there, which in turn channeled the flow into New Mexico and the Rio Grande Valley of Texas, bringing about a mobilization of enforcement resources in those sectors. However, the border is 2,000 miles long and systematically blockading this entire length in the manner of San Diego or El Paso is prohibitively expensive.
Ultimately, the net effect of the border blockades has been to push undocumented Mexican migrants into crossing at more remote and less accessible locations in mountains, deserts, and untamed sections of the Rio Grande River. The tragic result for undocumented migrants has been a tripling of their death rate during entry. But if migrants are more likely to die while crossing remote sectors of the border, they are also less likely to be caught, and a little-known consequence of U.S. border-enforcement policy has been that it has decreased the odds that undocumented Mexican migrants are apprehended while attempting to enter the United States.
Figure 3 draws upon data from the Mexican Migration Project (MMP) at Princeton University and the University of Guadalajara to show trends in the location of border-crossing and the probability of apprehension among undocumented Mexicans from 1980 to 2002, the latest year for which reliable estimates are available. From 1980 through 1987, the proportion of migrants crossing in either Tijuana-San Diego or Juarez-El Paso increased. By 1988 around 70 percent of all border crossings occurred within these two “traditional” sectors. The militarization of the border begun by IRCA in 1986 was naturally targeted to these high-volume points of entry, a tendency that was amplified beginning in 1993-1994 with the launching of blockade operations. As a result, the proportion of migrants crossing at “non-traditional” sectors along the border has steadily risen from 29 percent in 1988 to 64 percent in 2002. Obviously, undocumented migrants are simply going around hardened sectors of the border.
Through the 1970s and early 1980s, the probability of apprehension along the border was relatively steady and averaged about 33 percent: roughly one third of attempted entries led to an apprehension. Thereafter, however, the probability of apprehension fell into the 20-30 percent range and following the implementation of operations Blockade and Gatekeeper in 1993 and 1994, the likelihood of arrest plummeted. By 2002 the probability of apprehension had reached an all-time low of just 5 percent! Rather than increasing the odds of apprehension, U.S. border policies have reduced them to record lows.
Given this fact, it is not surprising that U.S. border-enforcement policies have had little detectable effect in deterring undocumented migrants from leaving for the United States in the first place. Figure 4 uses MMP data to compute the probability that Mexican men and women took a first trip to the United States from 1980 onward. There is little evidence in either series that the border buildup has dissuaded undocumented Mexicans from heading northward. There is considerable variation in the trend for males, whose probabilities of first undocumented migration fluctuate between .015 and .025, with variations being closely tied to economic conditions on both sides of the border. Although the likelihood of female migration is much lower, the trend is virtually flat. Available data thus indicate that the inflow of undocumented Mexican immigrants continues apace, but that once at the border the odds of being apprehended are much lower. As a result, more undocumented migrants are gaining entry to the United States than ever before.
Crackdown on Legal Immigrants Backfires
Over the same time period as the border-enforcement buildup, legal immigration from Mexico has also grown, despite measures enacted by Congress to make it more difficult to qualify for documents and to reduce the rights and privileges of legal immigrants once they are here. Figure 5 shows trends in legal immigration from Mexico using official statistics. In response to an anti-immigrant backlash in the early 1990s, Congress in 1996 barred lawful permanent residents (LPRs) from receiving federally-supported public benefits and financial assistance. This action dramatically increased the incentives for naturalization among LPRs. Moreover, at about the same time, Mexico enacted a new policy of allowing dual nationality, permitting those who naturalize in the United States to retain their Mexican citizenship. Both events occurred just as the 2.3 million Mexicans who legalized under IRCA between 1988 and 1990 became eligible to apply for U.S. citizenship (following five years as LPRs).
The end result was predictable: a huge and unprecedented surge in the number of Mexicans naturalizing to U.S. citizenship. As Figure 5 shows, between 1986 and 1996, the number of Mexicans being naturalized increased by a factor of nine. This surge in naturalizations fell in 1997 as the cohort of Mexicans legalized under IRCA passed and the INS bureaucracy staggered under the administrative load. Administrative reforms reduced INS backlogs in 1998, however, and the number of Mexicans acquiring citizenship once again mushroomed to nearly seven times its 1986 level. Although the latest data show declines in the years 2001 and 2002, the rate of naturalization among Mexicans is still running at three times its 1986 level.
If by stripping legal immigrants of social rights Congress sought to discourage legal immigration from Mexico, the effort backfired. In fact, it encouraged millions of Mexicans who otherwise would have happily remained in LPR status to apply for U.S. citizenship, the acquisition of which in turn increased the rate of legal immigration. Although LPRs have the right to petition for the legal entry of their spouses and minor children, these visas are limited in number and immigrants must wait in line, typically for many years in countries such as Mexico, where the quota of numerically limited visas is perennially filled. Once an LPR becomes a U.S. citizen, however, spouses and minor children, as well as parents, are entitled to enter outside the quotas, not subject to any numerical limitation whatsoever. In addition, U.S. citizens acquire the right to sponsor the entry of brothers and sisters as well as older, married children and their spouses, though these categories are subject to numerical limitations.
In short, each person who becomes a U.S. citizen acquires new and powerful rights under U.S. immigration law to sponsor the entry of relatives, many of whom are not subject to numerical limitations. Each naturalization thus creates the potential for more immigration in the future. By taking rights away from LPRs, Congress pushed Mexicans decisively toward naturalization and therefore guaranteed the acceleration of legal immigration from Mexico. The dashed line in Figure 5 shows a “bump” in legal entries corresponding to the surge in naturalizations during 1996 and a steady increase thereafter. By 2002, legal immigration was running at more than three times its 1986 level. Reflecting the increased sponsorship of entries by U.S. citizens rather than LPRs, the number of “non-quota” immigrants not subject to numerical limitations also began to grow rapidly in 1994 and accelerated markedly after 1996. By 2002 it was approaching four times its 1986 level.
The Costs of Misguided Policies
Although the size of the Border Patrol budget increased by a factor of 10 between 1986 and 2002, and the number of Border Patrol Agents tripled, more Mexican immigrants – both documented and undocumented – are arriving than ever before. The combination of huge budget increases with rising immigration rates suggests a marked deterioration in the efficiency of U.S. border enforcement. American taxpayers are spending far, far more to achieve much less in the way of deterrence and relatively fewer arrests along the border. This hypothesis is confirmed by the data in Figure 6, which divides the Border Patrol’s annual budget by the number of apprehensions achieved along the Mexico-U.S. border, expressing the average cost of arresting one undocumented Mexican migrant.
Before 1986 the cost of one apprehension was roughly constant at around $100 per arrest. Beginning with the passage of IRCA in 1986, however, the cost of enforcement began to rise, tripling to around $300 per arrest in 1989 before stabilizing for a time. Beginning with the launching of operations Blockade and Gatekeeper in 1993 and 1994, however, the cost of making one arrest immediately jumped to more than $400 and then gradually increased to $600 in 1999. The events of September 11, 2001, brought another huge infusion of resources to the Border Patrol that was in no way connected to the threat of either terrorism or undocumented migration emanating from south of the border. The cost of an apprehension then skyrocketed. Whereas the cost of making one arrest along the border stood at just $300 in 1992, ten years later it reached $1,700, an increase of 467 percent in just a decade.
If this increase in the cost of enforcement, expensive as it was, had slowed the flow of undocumented immigrants, then one might consider it money well spent. But as we have seen, in 2002 the probability of apprehension was lower than at any point in the modern history of Mexico-U.S. migration and the number of Mexicans entering the United States was greater than ever. Whatever one thinks about the goal of reducing migration from Mexico, U.S. policies have clearly failed, and at great cost to U.S. taxpayers. The money allocated to border enforcement since 1986 has been a complete and total waste of billions of dollars.
Data presented so far have shown that, despite massive increases in border enforcement and congressional actions undertaken to discourage legal immigration, the number of legal and illegal entries from Mexico has continued to grow, implying the waste of billions of dollars (not to mention hundreds of lives) in the futile effort to prevent the movement of labor within a rapidly integrating North American economy. As grim as this assessment may be, it gets worse. Not only have U.S. policies failed to reduce the inflow of people from Mexico, they have perversely reduced the outflow of people back to Mexico and thus produced an unprecedented increase in the population of the United States. America’s unilateral effort to prevent a decades-old pattern of migration from continuing has paradoxically transformed a circular flow of Mexican workers into a settled population of families and dependents in the United States.
Rather than electing not to enter the United States without inspection, undocumented immigrants quite rationally invested more money to minimize the risks and maximize the odds of a successful border crossing. As U.S. authorities deployed a more formidable array of personnel and materiel at key points along the border, smugglers on the Mexican side simply upgraded the package of services they offered. Smugglers used to simply accompany small parties of undocumented migrants on foot across well-trod pathways from Tijuana to San Diego and deliver them to some urban setting. Now they had to transport people to remote sectors of the border, guide them across, and transfer them on the other side to other personnel who would arrange transport to destinations throughout the United States.
The net effect of U.S. policies, in other words, was to increase the quality and price of border-smuggling services. After the various blockades were launched, undocumented migrants faced rising out-of-pocket costs to ensure a successful border crossing. The extent of this increase is indicated by Figure 7, which shows the average amount that undocumented migrants paid someone to smuggle them into the United States by year. From 1980 to 1992, the cost of hiring a coyote or pollero (as smugglers are colloquially labeled) was relatively flat, averaging around $400 per crossing. With the launching of the new strategy of prevention through deterrence in 1993, however, the cost of purchasing a smuggler’s services rose to around $1,200 in 1999 before leveling off.
In other words, compared to 1990 and prior years, by 2000 it cost undocumented migrants three times as much to gain entry to the United States. If the first order of business on any trip to the United States is to recover this cost, then holding constant the rate of remuneration and hours worked per week, the trip would have to be three times as long. Although militarizing the border may not have reduced the inflow, it did substantially increase the length of trips and thereby reduced the outflow back to Mexico.
Another way of viewing the increase in trip lengths is in terms of a decline in the probability of return: fewer migrants return within one year of their original entry. This fact is illustrated in Figure 8, which uses MMP data to compute the probability of returning to Mexico within 12 months of entry. As can be seen, before IRCA the annual likelihood of return migration fluctuated between 40 percent and 50 percent with no clear trend. After 1986, however, we observe a steady, sustained decline in the likelihood of return migration, which bottoms out at 24 percent in 1996 and begins to oscillate. Roughly speaking, the average probability of return migration goes from around 45 percent before IRCA to around 25 percent today. If 1,000 migrants were to enter the United States each year at the former rate, 950 or 95 percent would be back in Mexico within five years and the average length of trip would be 1.7 years. At the latter rate, of 1,000 migrants who entered the United States within a given year, only 763 or 76 percent would have returned to Mexico within five years and the average trip duration would have grown to 3.5 years.
If the number of undocumented Mexicans entering the United States each year after 1986 remained constant or was increasing, as the evidence suggests, and the probability of return migration was simultaneously falling, then only one outcome is possible: a sharp increase in the size of the undocumented population living in the United States at any point in time. In demographic terms, if the number of entries to a population persists or grows while the number of exits falls, it can only grow.
The growth in the size of the Mexican population of the United States as recorded by the U.S. Census Bureau is shown in Figure 9. From 1980 through the mid 1990s, the Mexican population of the United States grew at a steady if rapid rate, roughly tripling in the 15 years from 1980 to 1995. After 1990 the rate of Mexican population growth shifts sharply upward, with the population growing from 7 million in 1997 to around 10 million in 2002 – an increase of 43 percent in just five years. After results from the 2000 Census were published, it was evident that Hispanics had overtaken blacks to become the nation’s largest minority far earlier than most demographers had predicted.
A New Approach to Mexican Immigration
The solution to the problems associated with undocumented migration is not open borders, but frontiers that are reasonably regulated on a binational basis. Under current U.S. immigration law, all countries are allotted the same quota of 20,000 legal immigrants per year, no matter what their size or relationship to the United States. Thus, our largest and closest neighbor and most important trading partner has the same limited access to U.S. visas as Botswana, Nepal, and Paraguay. A more realistic policy would recognize Mexico’s unique status by increasing the annual immigrant quota, establishing a flexible temporary labor program, and regularizing the status of those already here. By bringing the flow of immigrants above board, we would mitigate the downward pressures on wages and working conditions in the United States while raising tax revenues that could be used to offset the costs of immigration and to assist Mexico in overcoming the market failures that motivate so many people to move north of the border.
Specifically, I propose that the United States create a new category of temporary visa that permits the bearer to enter, live, and work in the country without restriction for two years, with an option for renewal once in the lifetime of the migrant, but only after he or she has returned home. The visas would be issued to persons and not tied to specific jobs. Such a program would guarantee the rights of temporary migrants, protect the interests of American workers, and satisfy the demands of employers by moving toward a relatively free and open North American labor market.
I would make these new visas generously available to residents of Canada and Mexico. If 300,000 two-year visas were issued annually, there would be 600,000 temporary migrants working in the United States at any time, a small share of the U.S. workforce but a large fraction of undocumented migrants. Moreover, I would charge a $400 fee to migrants for each visa issued, to be paid up front in cash or in low-interest installments from the migrant’s U.S. earnings. This money could be used for the benefit of the migrants themselves, in ways described below. The data presented above indicate that migrants are perfectly willing to pay this amount to gain entry to the United States, but up until now the money has gone into the pockets of border smugglers rather than toward more beneficial purposes. A $400 fee paid by 300,000 temporary migrants per year would yield annual revenues of $120 million.
As an additional source of revenue, the government could earmark federal taxes (Social Security, Medicare, and income taxes) withheld from the paychecks of temporary migrants for immigration-related initiatives. If 600,000 temporary migrants were to earn annual incomes of just $15,000 and have taxes withheld at a rate of only 25 percent, the annual revenues would be $225 million per year. Additional revenues could be raised by drastically reducing the personnel and resources devoted to border enforcement. There is no evidence whatsoever that the costly militarization of the U.S. border has raised the odds of apprehension or prevented the entry of undocumented immigrants, but the human costs in terms of injury and death have been great. The Border Patrol would be equally effective, more efficient, and violate fewer human rights with a smaller number of officers assigned to the border.
In addition to reducing the size of the Border Patrol, I would redeploy its enforcement effort from the border to the interior of the United States, targeting regions of the country and economic sectors known to employ immigrant workers. Rather than focusing on the identification and apprehension of undocumented immigrants, however, worksite inspections would vigorously enforce U.S. tax, labor, environmental, and occupational health and safety laws, eliminating many of the incentives for employing undocumented immigrants and ensuring that employers are deducting taxes to finance immigration initiatives. It is essential that employers be held responsible for violations, even if workers are hired through a labor subcontractor.
With revenues from visa fees and tax withholding from temporary migrants, the United States would join with Mexican authorities to establish a binational insurance program that allows all migrants to purchase low-cost insurance for a variety of purposes, giving them a means of risk management other than migration. The revenues would also be used to establish a binational agency for economic development that would make matching grants to Mexican communities for the construction or improvement of local infrastructure. Such an agency would offer migrants a way of multiplying the positive benefits of their remittances back to Mexico (currently on the order of $20 billion per year) by pooling them for local development initiatives and doubling them dollar for dollar. Migrant-generated revenues would also be used to create a migrant savings bank that pays dollar depositors above-market interest rates as a means of attracting earnings back to Mexico, and would make low-interest loans to individuals and families for both production and consumption, using revenues from visa fees and tax withholding to finance the subsidies.
It is also imperative to increase the number of permanent resident visas available to Mexicans to 100,000 per year. The current quota of 20,000 visas for a nation to which we are so closely bound by history, geography, and treaty is absurdly low, yielding excessively long waiting times for many legally qualified immigrants and virtually guaranteeing undocumented migration. At the same time, however, I would eliminate the preference category that confers rights of entry on brothers and sisters of U.S. citizens. This is an unnecessary provision that bears more responsibility than any other feature of U.S. immigration law for reinforcing the process of chain migration that propels so much immigration from Mexico.
The foregoing actions go well beyond what President Bush proposed in his January 7, 2004, speech on immigration reform, but are not so distant from reforms broached by senators and representatives in the wake of his announcement. If enacted, these policy reforms will not eliminate undocumented immigration from Mexico, of course, nor solve all of the problems associated with it. They will, however, reverse the deleterious consequences of our current policies by eliminating the black market in immigrant labor, minimizing the long-term settlement of Mexican immigrants, encouraging the repatriation of capital and people to Mexico, promoting economic growth within migrant-sending communities, and overcoming the prevailing weaknesses in Mexican capital, credit, and insurance markets.
In the short run, the disruptions that follow from the consolidation of the North American market will continue to produce migrants to the United States. But long-term economic growth and development within Mexico will gradually eliminate most of the incentives for international migration. We should not seek to stamp out the inevitable migratory flows, but move North America toward a more balanced economy in which fewer Mexicans will experience the need to migrate northward.
NOTE: Portions of this report, along with a discussion of U.S. trade policies, were published in Backfire at the Border: Why Enforcement without Legalization Cannot Stop Illegal Immigration. Washington, DC: Cato Institute Center for Trade Policy Studies, June 13, 2005.
1 Peter Andreas, Border Games: Policing the U.S.-Mexico Divide. Ithaca, NY: Cornell University Press, 2000.
3 Joseph Nevins, Operation Gatekeeper: The Rise of the ‘Illegal Alien’ and the Remaking of the U.S.-Mexico Boundary. New York: Routledge, 2002.
4 Timothy J. Dunn, The Militarization of the U.S.-Mexico Border 1978-1992: Low Intensity Conflict Comes Home. Austin: University of Texas Press, 1996.
5 Joseph Nevins, Operation Gatekeeper: The Rise of the ‘Illegal Alien’ and the Remaking of the U.S.-Mexico Boundary. New York: Routledge, 2002.
6 Jason Ackleson, “Fencing in Failure: Effective Border Control is Not Achieved by Building More Fences,” Immigration Policy IN FOCUS 4(2). Washington, DC: Immigration Policy Center, American Immigration Law Foundation, April 2005.
7 Douglas S. Massey, Jorge Durand & Nolan J. Malone, Beyond Smoke and Mirrors: Mexican Immigration in an Age of Economic Integration. New York: Russell Sage Foundation, 2002.
Douglas S. Massey for the Immigration Policy Center is professor of sociology and public affairs at Princeton University and coauthor of Beyond Smoke and Mirrors: Mexican Immigration in an Era of Economic Integration (New York: Russell Sage Foundation, 2002).
The opinions expressed in this article do not necessarily reflect the opinion of ILW.COM.