The Effects of Border Enforcement on Migration to the United States: New Evidence from Oaxaca, Mexico

Transcription

1 The Effects of Border Enforcement on Migration to the United States: New Evidence from Oaxaca, Mexico Kristen Parks, Gabriel Lozada, Miguel Mendoza, and Lourdes Garcia Santos 2009 Congress of the Latin American Studies Association Rio de Janeiro, Brazil, June 11-14, 2009

2 2 Strategies for Success: Border Crossing in an Era of Heightened Security KRISTEN PARKS, GABRIEL LOZADA, MIGUEL MENDOZA, AND LOURDES GARCÍA SANTOS They told me the first time, If we apprehend you a second time, we re going to put you in jail for two weeks. If we apprehend you a third time, it s going to be a month; the fourth time, three months. You could be kept in jail for up to a year. No matter what they say to you, you re still going to try again. I said, Well, I just have to cross. They asked me if I was sure. Maybe you should just go home, they said. But I have to cross, I told them. No matter what, the majority of us Mexicans, we re going to keep trying. Briseida, a 24-year-old migrant who crossed in The U.S. Border Patrol apprehended Briseida six times in the month before her most recent clandestine entry into the United States. Her experience and those of many other Tlacotepenses offer a corrective to the conventional wisdom that potential migrants underestimate the difficulty of unauthorized crossings and will give up and go home after one or two failed entry attempts. Potential migrants do not make their decision to cross the border unsystematically. They weigh the benefits of migrating to the United States against the potential costs of a clandestine entry and present conditions in Mexico. Like Briseida, many migrants will persist in border-crossing attempts until they are successful. Although previous research suggests that U.S. policy has failed to reduce unauthorized entry into the country, government officials continue to emphasize the success and promise of ongoing border fortification efforts. Between 1993 and 2008, heightened security along the U.S.-Mexico border cost U.S. taxpayers more than $30 billion. MixtecaBook.indb 31 2/6/09 9:31:50 AM

3 32 Parks, Lozada, Mendoza, and García Santos During this fifteen-year period, both the Border Patrol and the resident population of undocumented migrants in the United States tripled in size (GAO 2007; Pew Hispanic Center 2007). The Mexico-born population in the United States has been increasing steadily over time, by as much as 500,000 individuals per year. The Pew Hispanic Center estimates that 80 to 85 percent of these new arrivals are undocumented. There is some evidence, however, that the rate of growth in this population has slowed somewhat since the 1990s. A reduction in the number of Border Patrol apprehensions on the U.S.-Mexico border and a drop in remittances sent to Mexico would appear to confirm this slowing in unauthorized migration. Apprehensions declined by about 25 percent in fiscal years 2001 and Though they rebounded in 2004 and 2005, they began to drop again in 2006 and thereafter (see figure 2.1). We argue later in this chapter that fluctuating apprehension rates reflect the geographic effects of concentrated border operations but actually reveal little about the ultimate success of potential undocumented crossers. Figure 2.1 Southwest Border Apprehensions, FY1980 FY2007 Number of apprehensions (millions) Concentrated border enforcement Year The flow of remittances from migrants working in the United States to their relatives in Mexico has also begun to slow. Remittances grew by only 1 percent in 2007, compared with sharp increases in the previous six MixtecaBook.indb 32 2/6/09 9:31:51 AM

4 Strategies for Success in Border Crossing 33 years (see figure 2.2). These data are unusual because migrant transfers tend to be remarkably stable even during economic recessions in the host country. Declining remittances could mean that migrants are receiving smaller paychecks, that fewer new migrants are entering the labor force, or that migrants are settling and bringing their families to the United States, obviating the need to send money home. Figure 2.2 Remittances to Mexico, U.S. dollars (billions) Year Sources: Inter-American Development Bank; the 2008 figure is a Banamex-Citigroup estimate. SUMMARY OF FINDINGS Various explanations have been advanced for the downward trends in apprehensions, remittances, and growth of the population of undocumented Mexicans in the United States. These range from tougher border enforcement to cyclical weaknesses in the economy. This chapter uses survey data on migration behavior and the perceptions of migrants and nonmigrants from San Miguel Tlacotepec, Oaxaca, to test these alternative explanations. We reconfirm previous findings that border enforcement efforts do not discourage most undocumented migrants from attempting to enter the United States. Instead, we find that economic and social factors provide the best explanations for trends in undocumented MixtecaBook.indb 33 2/6/09 9:31:51 AM

5 34 Parks, Lozada, Mendoza, and García Santos migration. Our data suggest that migrants are reducing the frequency of their return trips to Mexico because of the rising costs and risks of clandestine entry. In effect, tougher border enforcement and escalating costs of clandestine entry are discouraging circular migration and keeping migrants bottled up in the United States (see chapter 3 for further discussion of this topic). The current generation of border crossers has an extremely high success rate in clandestine entry, partly because of an increased reliance on coyotes. These professional people-smugglers have a remarkable ability to adapt to the ever-changing obstacles at the border. Thanks to this learning effect, coyotes seem to be increasing the likelihood of a successful crossing and, concomitantly, reducing U.S. apprehension rates. Finally, we suggest that the recent contraction of the U.S. economy has reduced employment prospects for some potential migrants, weakening the pull of the jobs magnet. Given that the primary reasons Mexicans give for migrating to the United States are better salaries and more jobs, 1 the economic slowdown is likely of equal or perhaps even greater importance than the deterrent impact of border enforcement in explaining the decrease in apprehensions. If unauthorized migration is driven by macrolevel economic forces, will border enforcement be most successful when it raises the costs and risks for the potential migrant, or will migrants and their smugglers always remain one step ahead of U.S. officials? A brief discussion of border policy in recent decades sheds light on what we consider to be a cyclical pattern of enforcement innovations followed by adaptations in border crossers strategies. POLICY HISTORY Immigration enforcement in the United States has long focused on preventing unauthorized entry at the southern border, but more aggressive policies were pursued beginning with the Clinton administration (see table 2.1). In 1993, Operation Hold the Line, the first in a series of concentrated border enforcement operations, was launched in the El Paso, Texas, area. The objective of this and similar operations was to curb illegal 1. Nearly 69 percent of undocumented respondents in our sample of Tlacotepec migrants came to the United States to work. See chapter 1 for a more detailed discussion of motives for migration. MixtecaBook.indb 34 2/6/09 9:31:51 AM

6 Strategies for Success in Border Crossing 35 crossings by concentrating a large number of Border Patrol agents, vehicles, and surveillance technology on a small segment of the border. This tactic led to a 72 percent drop in apprehensions in the El Paso sector (Bean et al. 1994). In 1994, the Clinton administration launched Operation Gatekeeper, another concentrated enforcement effort, in San Diego County. Subsequent border initiatives emphasizing prevention through deterrence include Operation Safeguard in Arizona (begun in 1995) and Operation Rio Grande in the McAllen sector in Texas (launched in 1997). Table 2.1 Key Developments in U.S. Border Enforcement Policy, Operation Hold the Line (El Paso, Texas) 1994 Operation Gatekeeper (San Diego, California) 1995 Operation Safeguard (Nogales, Arizona) 1997 Operation Rio Grande (Southeast Texas) 2005 Operation Streamline (Del Rio, Texas, and Yuma, Arizona) 2006 Secure Border Initiative Secure Fence Act Operation Jump Start Recent border enforcement initiatives have shifted emphasis from physical obstacles to high-tech monitoring systems and punitive measures. Under the Secure Fence Act, approved by the U.S. Congress in 2006, 700 miles of new primary fencing and vehicle barriers are to be installed, at an estimated cost of $3 million per mile (Archibold et al. 2008, 2). In some urban areas a new secondary fence of steel mesh will be added; and in the San Diego sector, razor-sharp concertina wire is being installed atop a five-mile section of the border fence (Marosi 2008). If all of the new fencing mandated by the Secure Fence Act were constructed, about onethird of the nearly 2,000-mile U.S.-Mexico border would have physical fortifications, at a cost of at least $2.5 billion The U.S. Corps of Engineers estimates that the construction of a double-layer fence would cost anywhere from $1.2 to $1.3 million per mile. This does not take into account the cost of acquiring the land, much of which is privately owned, nor the estimated $16.4 to $70 million per mile for repairing the fence when it is damaged or broached (Kim et al. 2008, 25). MixtecaBook.indb 35 2/6/09 9:31:51 AM

7 36 Parks, Lozada, Mendoza, and García Santos Under the Secure Border Initiative, mandated by Congress in 2006, a virtual fence consisting of 1,800 high-tech towers and other technologically advanced surveillance systems are to be built along the border. Homeland Security Secretary Michael Chertoff promised that this project, known as SBInet, would address all aspects of the border security problem across the board deterrence, detection, response, apprehension, detention, and removal (Ewing 2008, 7). Yet the initiative has been hampered by hardware and software glitches which have delayed implementation by several years. An SBInet pilot project, built along a 28-mile stretch of the border in central Arizona, proved unsatisfactory for replication elsewhere. Nevertheless, the George W. Bush administration asked Congress for an additional $775 million in funding for the SBInet project in FY Secretary Chertoff defended the project as follows: Yes, you can get over it; yes, you can get under it. But it is a useful tool that makes it more difficult for people to cross. It is one of a number of tools we have, and you ve got to use all of the tools (Archibold et al. 2008, 1). In addition to physical fortifications, the manpower available to police the Southwest border grew threefold over a decade. There were 14,923 Border Patrol agents in FY 2007, up from 5,878 agents in The largest single-year increase came in 1998, when 1,102 new agents were added (see figure 2.3). The Bush administration pledged to create a force of more than 18,000 Border Patrol agents before it left office in January In addition, the administration deployed about 6,000 National Guard troops to the Southwest border for Operation Jump Start in These troops helped build border fencing and access roads, and also backstopped the Border Patrol in various support roles. Guardsmen were not authorized to apprehend unauthorized migrants, and their deployment ended in the summer of In the face of mounting evidence that physical barriers and increased manpower are failing to deter many would-be unauthorized entries, U.S. border enforcement policy is shifting to a penalty-based or zero-tolerance approach. The Border Patrol s Operation Streamline is an example of this shift. It began in the Del Rio, Texas, sector as a pilot project in December 2005 and was later implemented at least partially in the Yuma, Laredo, and Tucson sectors. Theoretically, Streamline subjects all unauthorized migrants apprehended in the sectors where it operates to MixtecaBook.indb 36 2/6/09 9:31:51 AM

8 Strategies for Success in Border Crossing 37 mandatory incarceration and prosecution for unauthorized entry, a misdemeanor under federal law. Detainees can be jailed for up to 180 days, and they are deported after serving their sentences. Figure 2.3 Growth of U.S. Border Patrol, Number of agents (thousands) Fiscal year Sources: GAO 2007; 2008 and 2009 numbers are estimates. There has been a significant decline in apprehensions in some sectors where Streamline has been implemented. Homeland Security officials report that since December 2006, apprehensions have dropped by nearly 70 percent in the Yuma sector (U.S. Department of Homeland Security 2008). During the first nine months of Streamline, repeat crossings by apprehended migrants in the Del Rio sector fell from 14 to 4 percent. In the 45 days after Streamline was implemented in the Laredo sector, apprehensions dropped 33 percent (Preston 2008, 2). Border Patrol officials also claimed that in a 15-mile target area within the Tucson sector where Streamline was in operation, a 79 percent recidivism rate plummeted to 46 percent (Holstege 2008, 2). Of course, falling apprehensions and recidivism rates along one segment of the border do not necessarily mean that migrants are not bypassing those areas to cross at other points. In fact, rising apprehensions in other sectors, such as San Diego, suggest that this is precisely what may be happening (see figure 2.4). MixtecaBook.indb 37 2/6/09 9:31:52 AM

9 38 Parks, Lozada, Mendoza, and García Santos Figure 2.4 Southwest Border Apprehensions, by Border Patrol Sector, Number of apprehensions (thousands) San Diego, CA Yuma, AZ El Paso, TX Del Rio, TX Rio Grande Valley, TX El Centro, CA Tucson, AZ Marfa, TX Laredo, TX Fiscal year One of the principal limitations of an increasing-the-penalties approach is a lack of capacity in the federal criminal justice and detention system, and fewer than 10 percent of migrants apprehended in some Streamline sectors have actually been incarcerated (Archibold et al. 2008, 2). In May 2008, only 32,000 beds were available for immigration detainees. Federal courts are overwhelmed with immigration and drug cases. Though the Justice Department will add sixty-four federal prosecutors along the border to support the new caseload, there are additional challenges to expansion, particularly the lack of detention space and the cost of incarceration (currently between $90 and $119 per day per migrant). Nevertheless, there is a strong constituency in the U.S. Congress for expanding Streamline to cover more of the border. For example, in 2008 Senator Jeff Sessions (R-Alabama) introduced legislation to extend Operation Streamline to all sectors of the border by the end of EVALUATING THE SUCCESS OF ENFORCEMENT The U.S. government s explanation for the recent decline in border apprehensions is rooted in the border fortification efforts undertaken in the MixtecaBook.indb 38 2/6/09 9:31:52 AM

10 Strategies for Success in Border Crossing 39 past fifteen years. Yet scholars have pointed out that apprehension rates are not evidence of deterrence, partly because recorded apprehensions do not distinguish between repeated arrests of the same migrant and thus likely overestimate the number of illegal entrants (Donato, Durand, and Massey 1992). We argue that, in addition, they are an unsatisfactory indicator of deterrence because they only include migrants who have been apprehended (fewer than half of crossers are caught, according to our interviewees in Tlacotepec). Our data on clandestine entry substantiate the claim that apprehension statistics are a poor metric of the actual flow of undocumented migrants into the United States. Orthogonal to the official explanation for the downward trend in apprehensions is the notion that border enforcement has reached a tipping point. Beyond this tipping point, some analysts expect to see a decline in attempted illegal entries as a direct result of the accumulation of Border Patrol agents, physical barriers, new technologies, and increased legal penalties for clandestine entry. We asked 22-year-old Héctor, an experienced migrant who crossed the border most recently in 2007, on which occasion he felt that crossing was most difficult. This last time, he reported. It is much harder to cross today.... There s more vigilance; there used to be less. Yeah, we probably noticed vigilance back then, but it was not as obvious as this last time. If border enforcement continues to expand at the current rate, at some point clandestine entry will become so improbable, costly, and risky that potential migrants will not bother to attempt crossing. But there is no convincing evidence that this tipping point has been reached. Indeed, as discussed below, changing bordercrossing strategies such as increased use of people-smugglers and crossings through legal ports of entry may be reducing the risks as well as the probability of detection, despite the continuing border enforcement buildup. ALTERNATIVE EXPLANATIONS FOR BORDER APPREHENSION TRENDS Another explanation for fluctuations in border apprehension numbers proposes a lag or learning effect. Cornelius argues that apprehensions rose in the late 1990s because migrants and people-smugglers were still learning how to evade the new obstacles, but by the end of 2001, the learning process was complete and the probability of apprehension once again began to decline (2005, 782). Historical evidence of learning curves MixtecaBook.indb 39 2/6/09 9:31:52 AM

11 40 Parks, Lozada, Mendoza, and García Santos can be found by examining the shift in border-crossing locations that occurred in the late 1990s as migrants began to cross through more remote desert and mountainous areas, circumventing the concentrated enforcement operations in urban areas like San Diego and El Paso. The general pattern of crossing behavior after concentrated border enforcement operations were implemented has been for migrants to cross in increasingly remote areas, like the central Arizona desert, in order to reduce the risk of detection. However, since FY 2004, apprehensions in the newly fortified Tucson sector have been declining (see figure 2.4). On the other hand, apprehensions in the heavily fortified San Diego sector have been rising since 2005, after a decade of sharp declines. Border Patrol agents traditionally have referred to such geographic redistributions in illegal entries as the squeezing-the-balloon effect. In most years since 1993, these sector shifts have not translated into an appreciable, borderwide decline in apprehensions. Progress in gaining what the Border Patrol terms operational control in certain sectors has been offset by loss of control in the areas to which migrants and people-smugglers have redirected their efforts. Still another plausible explanation for the drop in border apprehensions is less circularity in Mexico-to-U.S. migration. If fewer return trips are being made, fewer unauthorized migrants are vulnerable to apprehension on return trips to the United States. Migrants may also be delaying return trips to Mexico in hopes of amnesty legislation that has been debated in the U.S. Congress since Being apprehended at the border would jeopardize their eligibility in the event that such legislation is approved. The methods that unauthorized migrants use to cross the border can also affect apprehension statistics. More of them are using professional people-smugglers to cross the border, increasing their probability of entering without detection. Coyotes can guarantee successful crossings for their clients because they are constantly devising new routes and methods of entry, and many use the same technology as the Border Patrol (infrared night-vision scopes, for example). 3 Finally, the weakening of labor demand in certain sectors of the U.S. economy that have become heavy users of foreign-born labor could be 3. For a more detailed discussion of the people-smuggling industry, see chapter 3 in Cornelius, Fitzgerald, and Borger MixtecaBook.indb 40 2/6/09 9:31:52 AM

12 Strategies for Success in Border Crossing 41 discouraging some potential migrants from going north. 4 The Pew Hispanic Center reported a steep rise in Hispanic unemployment in the first quarter of 2008: 6.5 percent, compared to 4.7 percent for non-hispanics. Nearly 250,000 Hispanics in the United States have lost their jobs over the past five years (Pew Hispanic Center 2008), many of them in the construction sector. The construction industry employs approximately 16 percent of the total Mexican workforce in the United States (Banamex-Citigroup 2008, 26). While agricultural employment rates are considered somewhat inelastic, the construction sector is highly vulnerable to changing economic conditions. Unemployment in construction nationwide doubled from July 2006 to March 2008 from 6.1 to 12 percent. Figure 2.5 shows the sharp decline in the number of construction jobs in California since the housing bubble burst in Between January 2006 and March 2008, employment in that sector went from a 9.5 percent rate of growth to a 9.6 percent contraction. Moreover, employment in many service-sector and manufacturing jobs is dependent upon home construction. Figure 2.5 Changing Employment in California s Construction Sector, Change in construction employment in California (percent) Year While a recession in the United States clearly reduces the strength of the jobs magnet that pulls migrants out of Mexico, a downturn in the U.S economy could also produce contractions in the Mexican economy, thus strengthening the push factors that drive migration (see Banamex-Citigroup 2008). MixtecaBook.indb 41 2/6/09 9:31:53 AM

13 42 Parks, Lozada, Mendoza, and García Santos The agricultural sector has also experienced recent difficulties. Nationwide, unemployment in agriculture rose from 9.7 percent in March 2007 to 13.3 percent in March Agricultural employment in North San Diego County, where most Tlacotepense migrants seek work, has contracted somewhat less but mirrors the national trend nonetheless, with unemployment rising from 4.4 percent in March 2007 to 5.1 percent in March 2008 (Bureau of Labor Statistics, Information about negative trends in the U.S. job market is transmitted quickly through transnational social networks. Thus we can expect that major job losses in construction and other sectors would have been known to potential migrants in Alfonso, a 42-year-old Tlacotepense migrant who works construction in Vista, California, sized up the situation: It s critical. This is going to stop many illegals.... The U.S. economy is in a downward spiral. Many are losing their homes Mexicans who bought houses but bought them at too high a price. More will lose their houses if construction work keeps going down. EFFECTS OF BORDER ENFORCEMENT ON THE PROPENSITY TO MIGRATE According to Reyes et al. (2002), stronger border enforcement does not have a statistically significant effect on the probability of migration. Economic conditions in the United States and Mexico and social networks were found to be more potent predictors of migration decisions. Some analysts even argue that increased enforcement can encourage migration, as potential migrants choose to cross sooner in fear of future enforcement (Massey and Espinosa 1997). One could argue that the success of U.S. border enforcement relies on the perceptions of millions of would-be migrants across Mexico. Undocumented entry will only decrease when the majority of potential migrants conclude that the costs and risks of illegal entry are greater than the potential benefits awaiting them on the other side of the border. To make this calculation they must be aware of border fortifications and consider them formidable threats. In our interviews among Tlacotepenses, we delved into their knowledge and perceptions of the obstacles to clandestine entry. Their responses help us understand which factors are most likely to deter unauthorized migrants. MixtecaBook.indb 42 2/6/09 9:31:53 AM

14 Strategies for Success in Border Crossing 43 Using respondents stated intent to migrate to the United States during the next twelve months (January December 2008) as the dependent variable, we analyzed the influence of several types of knowledge and perceptions related to increased border enforcement on the propensity to migrate. These include the perception that undocumented entry is very dangerous, that it is very difficult to cross the border successfully, and knowing someone who died attempting to cross the border (see figure 2.6). Figure 2.6 Knowledge and Perceptions of Border Hazards as Predictors of Intent to Migrate Intend Do not intend Percent Crossing is very dangerous Know someone who died trying to cross Crossing is very difficult Perceived hazard Our data show that knowledge and perceptions of border-crossing difficulties and hazards are not useful predictors of the propensity to migrate. These border enforcement related variables do not differentiate between those who intend to go to the United States in the next twelve months and those who plan to stay home. The only statistically significant difference between groups is that those planning to migrate are slightly more likely to know someone who died trying to cross the border. 5 This may be because those who do not intend to go to the United States do not interact regularly with experienced migrants, whose knowledge of such fatalities is likely to be higher. 5. Significant at a 90% confidence level (p = 0.094). MixtecaBook.indb 43 2/6/09 9:31:53 AM

15 44 Parks, Lozada, Mendoza, and García Santos A multivariate regression analysis of these data reveals that perceptions of border-crossing difficulty and danger have no statistically significant effect on the intent to migrate in 2008 when controlling for age, sex, marital status, education, previous migration experience, and social network ties (see table 2.2). Our results are consistent with those of several previous studies of other migrant-sending communities in Mexico conducted between 2005 and 2007 (Cornelius and Lewis 2007; Cornelius and Salehyan 2007; Cornelius, Fitzgerald, and Lewin Fischer 2007; Cornelius, Fitzgerald, and Borger 2009). Previous migration experience and family members in the United States emerge as the most significant indicators of intent to migrate. Of the variables having a negative impact on the propensity to migrate, only a lack of legal documents is statistically significant, a finding that simply reflects the fact that individuals possessing legal permissions are more likely to make a trip to the United States because they have the ability to do so without risk or payments to people-smugglers. Our data firmly establish the absence of any connection between attitudes about border enforcement and migration behavior. Respondents who were not planning to migrate to the United States in 2008 were asked to explain this decision. While many nonmigrants cited lack of economic need as their primary reason for staying home, nearly a quarter of them mentioned tougher border enforcement. 6 This result may be influenced by the large representation of seniors and the very poor in our sample. Older people are more likely to feel a physical disadvantage that would make them more vulnerable to arrest when attempting to cross the border. People with very low incomes may not feel able to pay for the professional assistance that would be needed to evade detection by the Border Patrol. Indeed, a logit analysis reveals that older respondents and those with lower weekly earnings were significantly more likely to specify border enforcement as their reason for not planning to migrate in A 2005 study in Zacatecas and Jalisco also found that the most-cited reason for not going to the United States was border enforcement, though the effect was statistically insignificant in the authors multivariate analysis (Cornelius and Lewis 2007, 62). 7. The coefficients for the effects of age and income on probability of mentioning border enforcement as a reason to stay home are.087 and.012, respectively. Both are statistically significant at a 95% confidence level. MixtecaBook.indb 44 2/6/09 9:31:54 AM

16 Strategies for Success in Border Crossing 45 Table 2.2 Logit Models: Effect of Border Perceptions on Propensity to Migrate Variable Model 1 Difficulty of crossing.105 (.470) Model 2 Danger in crossing.028 (.566) Know someone who died trying to cross Male.319 (.364) Age.006 (.073) Married.650 (.410) Education.035 (.040) Previous migrant 1.15*** (.404) Undocumented 1.56*** (.510) Social network a.102*** (.039) Constant.456 (1.32).385 (.362).007 (.074).643 (.400).048 (.042) 1.09*** (.404) 1.60*** (.510).100** (.039).045 (1.36) Model (.370).361 (.364).017 (.073).670 (.397).055 (.041) 1.12*** (.400) 1.70*** (.510).103*** (.040).246 (1.32) Model (.482).189 (.592).556 (.370).321 (.368).010 (.076).702 (.411).040 (.038) 1.17*** (.405) 1.66*** (.516).103*** (.037).330 (1.40) N Chi-squared Robust standard errors in parentheses; * 90% confidence level, ** 95% confidence level, *** 99% confidence level. a The social network variable measures the number of immediate family members currently residing in the United States. SOURCES OF CONCERN ABOUT CLANDESTINE BORDER CROSSINGS What do prospective migrants worry about most when they contemplate going to the United States without documents? We asked our interviewees to choose three images from a set of seven pictures representing common sources of concern for unauthorized migrants (see figure 2.7). These could be considered discouragement factors affecting the propensity to migrate. Forty-three percent of our respondents ranked the harsh, life-threatening climate of the southwestern deserts as the principal MixtecaBook.indb 45 2/6/09 9:31:54 AM

17 46 Parks, Lozada, Mendoza, and García Santos source of anxiety among potential migrants (see figure 2.8). The Border Patrol was the second most frequently reported concern, followed closely by Mexican bandits. We found no significant differences among men and women or migrants and nonmigrants on responses to this question. Figure 2.7 Factors of Concern to Potential Migrants What worries you most when thinking of migrating? The National Guard Extreme Weather Border Patrol The Fence Mexican Authorities Minutemen Gangs MixtecaBook.indb 46 2/6/09 9:31:55 AM

18 Strategies for Success in Border Crossing 47 Figure 2.8 Relative Importance of Discouragement Factors Percent of responses Harsh climate (desert) Border Patrol Bandits Fence National Guard Minutemen (vigilantes) Mexican police Discouragement factor Concerns about the physical danger of clandestine border crossings are well founded. Since 1995, more than 4,500 bodies of migrants who perished in illegal entry attempts have been recovered a clear undercount of fatalities because of the unknown number of victims still missing. Despite intensified search-and-rescue efforts by the Border Patrol, the probability of dying versus that of being apprehended on the border nearly doubled over a seven-year period, from 1 death per 5,812 apprehensions in 1998 to 1 death per 3,109 apprehensions in 2004 (Cornelius 2005, 783). Yvón, a 31-year-old Tlacotepec resident, reflected on her husband s most recent border-crossing experience: There are many risks in crossing through the desert today. The last time my husband crossed the border, about four years ago, he said that he crossed through the desert and there was a man who could not keep up walking. The coyote abandoned him there. That s a big risk. That is why I have not left. While the risk of death may be a strong deterrent for some potential or experienced migrants, we find that most individuals who elect clandestine entry are well informed about the difficulties and dangers they will face at the border, and having such information does not fundamentally alter the calculus that leads to a decision to migrate. MixtecaBook.indb 47 2/6/09 9:31:55 AM

19 48 Parks, Lozada, Mendoza, and García Santos BORDER-CROSSING OUTCOMES Despite a rising probability of being caught by the Border Patrol at least once on a given trip to the border, apprehended migrants persist until they succeed. Evidence of the success of repeated entry attempts can be found in data from the Border Patrol s IDENT system (Espenshade 1990; Cornelius 2005, 780; Donato, Durand, and Massey 1992). Our survey respondents were asked how many times they were apprehended on their most recent trip to the border and whether they eventually made it across on that same trip. Although 48 percent of unauthorized migrants had been apprehended on their most recent trip to the United States, 97 percent eventually succeeded in entering, without having to return to Tlacotepec (see table 2.3). This finding is consistent with results from studies of migrant-sending communities conducted by our research program in (see figure 2.9). Table 2.3 Border-Crossing Experiences of Tlacotepenses, by Year of Most Recent U.S. Trip Entered Undetected on First Attempt Succeeded on Second Attempt Succeeded on Third or Later Attempt Failed to Gain Entry Before % 20% 19% 2% % 21% 32% 2% % 17% 19% 1% % 20% 19% 3% While apprehension rates have risen over this period, the eventual success rates among migrants interviewed in these studies have remained remarkably high and consistent, varying from 92 to 98 percent despite a continuous strengthening of border enforcement. Our data from Tlacotepec (figure 2.10) show that the eventual success rate is virtually unchanged between migrants whose most recent crossing occurred before 1995 (before heavy border fortification) and those crossing in the most recent period. These data corroborate previous reports that the subsequent border enforcement buildup has made no difference in terms of keeping undocumented migrants out of the United States. Extremely high success rates do not occur by chance; rather, they are achieved through an evolving array of border-crossing strategies. We now focus our analysis on these evolving MixtecaBook.indb 48 2/6/09 9:31:55 AM

20 Strategies for Success in Border Crossing 49 methods of clandestine entry, hypothesizing that border-crossing strategies change over time as border enforcement levels and tactics evolve. Figure 2.9 Apprehension and Eventual Success Rates among Undocumented Mexican Migrants from Selected Rural Mexican Migrant-Sending Communities Percent Apprehended on most recent trip to border Able to enter eventually, on same trip Jalisco 05 Yucatán 06 Jalisco 07 Oaxaca Migrant-sending community Sources: Surveys by the Mexican Migration Field Research and Training Program, in Tlacuitapa, Jalisco, and Las Ánimas, Zacatecas (2005); Tunkás, Yucatán (2006); Tlacuitapa, Jalisco (2007); and San Miguel Tlacotepec, Oaxaca ( ). Figure 2.10 Eventual Success in Crossing the Border among Tlacotepec Migrants Percent of undocumented migrants Apprehended on most recent trip to border Able to enter eventually, on same trip Before Year of most recent entry MixtecaBook.indb 49 2/6/09 9:31:56 AM

21 50 Parks, Lozada, Mendoza, and García Santos RELIANCE ON PEOPLE-SMUGGLERS Many people suffered a lot. I saw people dying of thirst. When I found them I gave them water and went without drinking myself. I felt a lot of responsibility because a person s life depended on me. That person may have left his children here, his father, his mother.... I don t know what he left when he paid us and gave us the responsibility of crossing him to the United States. Raúl, a former coyote living in California. Raúl s words poignantly illustrate the human dimension of the peoplesmuggling industry. Raúl began assisting family members and acquaintances cross the border and was gradually sucked into the lucrative coyotaje business over the course of a decade. Although he later served jail time for people-smuggling, the sentiments he shared with us demonstrate a yearning for social legitimacy: What I like about the United States is the law, because it is applied. We deserve punishment because we are breaking the law... and laws need to be respected. But sometimes we must break the law out of necessity, because we need to feed our families, give them shoes, educate our children. Professional people-smugglers (known as coyotes or polleros) have played a role in unauthorized immigration from Mexico for decades. However, our data from Tlacotepec reveal a significant increase in the utilization of coyotes since 2000 in response to the intensification of U.S. border enforcement efforts in recent years. Physical fortifications, increased numbers of Border Patrol agents on the line, and technological advances in the Border Patrol s surveillance capabilities increase the probability of apprehension and the danger of clandestine entry. Crossing through remote and hazardous desert and mountainous areas requires an experienced and knowledgeable guide. Migrants invest in the services of professional smugglers both to reduce their physical risk and to increase the likelihood of evading the Border Patrol. Logistical decisions about when and where to cross the border are being delegated increasingly to expert coyotes (see Cornelius, Fitzgerald, and Lewin Fischer 2007, chap. 5; Cornelius, Fitzgerald, and Borger 2009, chap. 2). Figure 2.11 shows a sharp increase in coyote usage among MixtecaBook.indb 50 2/6/09 9:31:56 AM

22 Strategies for Success in Border Crossing 51 Tlacotepense migrants between and , which coincides with the full implementation of concentrated border enforcement operations in California and Arizona. Using a coyote essentially guarantees success for most undocumented migrants. Among the Tlacotepenses we interviewed, all who had employed the services of a coyote eventually entered the United States successfully on their most recent trip to the border. This can be attributed in part to the fact that coyotes are generally not paid until their clients are delivered safely to their final destination, at which point the migrant s U.S.-based relatives pay the coyote for his services. Figure 2.11 Coyote Usage among Undocumented Migrants Percent of undocumented migrants Before Year As the demand for coyotes has risen, so have the fees they can command. Coyote fees have tripled or quadrupled borderwide in the post period (Cornelius 2001, 668). The run-up in fees is a direct consequence of heightened border enforcement. There is evidence that Border Patrol efforts (measured by line watch hours) in a given sector are correlated with increased coyote fees (Massey, Durand, and Malone 2002, 58). Among our Tlacotepense respondents, payments to coyotes have risen by 5 percent per year on average, controlling for inflation (see figure 2.12). The average fee paid to a coyote in 1995 was $978; by it had risen to $2,124. MixtecaBook.indb 51 2/6/09 9:31:57 AM

23 52 Parks, Lozada, Mendoza, and García Santos Figure 2.12 Coyote Fees Paid by Tlacotepense Migrants 2,500 Median coyote fee (in 2007 dollars) 2,000 1,500 1, Increased border enforcement Year Our data reveal that walking across the border with a coyote is substantially cheaper (56 percent less) than crossing by other methods, such as paying for false or borrowed documents to pass through a legal port of entry, for example. However, it is considerably more risky in terms of physical danger. As shown in the logit analysis reported in table 2.4, neither the migrant s age nor gender affects how much a coyote will charge. Also, finding a coyote through relatives or friends does not significantly affect the fee paid, nor does the number of social network contacts in the United States. The most significant predictors of the coyote fees that our respondents paid are the year of crossing and whether they walked across the border. Migrants find their coyotes in a variety of ways, but the most common is through a trusted relative or friend who is consulted prior to beginning the trip to the border. On their most recent trip to the border, nearly two-thirds arranged for the services of a coyote before migrating, receiving a recommendation from relatives living either in the United States or in Mexico; 31 percent hired a coyote only after arriving at the border (see table 2.5). The prevailing perception among migrants is that finding a trusted coyote through family connections is an important safety precaution. Most would-be migrants believe that coyotes hired by U.S.-based relatives provide better service than coyotes based at the border, who MixtecaBook.indb 52 2/6/09 9:31:57 AM

24 Strategies for Success in Border Crossing 53 typically charge less but work in higher volume and may not be as reliable. Briseida, who used U.S.-based relatives to hire a coyote, explained: The family that is here in Vista will help you find a coyote. When you are in Tijuana, many people come up to you and ask if you want to cross with them, but you cannot trust everyone.... Many are there to rob you. I have seen many awful cases. We do not hire coyotes from Tijuana. Table 2.4 Logit Analysis of Determinants of Coyote Fee Paid (most recent trip to United States) Variable Age Male Coyote hired through family/friends Number of U.S. social network contacts Crossing as mode of entry (dummy) Crossing year Constant Model (.006).180 (.160).092 (.161).009 (.015).575*** (.123).051*** (.014) 4.92*** (.794) N 146 R-squared.284 Robust standard errors in parentheses; * 90% confidence level, ** 95% confidence level, *** 99% confidence level. Table 2.5 How Did You Find Your Coyote? (on most recent trip to United States) Method Employed to Locate/Hire Coyote Friends/family in United States Percentage 45.5% Friends/family in Mexico 19.8% At border (no referral) 31.2% Met coyote in hometown 3.5% Total 100% MixtecaBook.indb 53 2/6/09 9:31:57 AM

25 54 Parks, Lozada, Mendoza, and García Santos Social networks also play a significant role in financing journeys to el otro lado. U.S.-based relatives have been the leading source of financing for Tlacotepense migrants, except for the most recent period (see figure 2.13). Since 2005 a plurality of migrants have financed their trips to the United States mainly out of personal savings. These may be experienced migrants who financed their trips with money earned during previous stays in the United States. Mexico-based relatives and unrelated creditors in Mexico have become less important over time. A possible explanation is that local moneylenders now charge exorbitant interest rates (as much as 10 percent per month); and because of sharply rising coyote fees, family members in Mexico no longer have the resources to pay them. Figure 2.13 Sources of Financial Assistance for Migration 60 Percent Pre Family member in U.S. Non-family member Family Creditors member in Mexico in Mexico Source of financial assistance Personal savings Other BORDER-CROSSING LOCATIONS AND METHODS We found that the overwhelming majority of Tlacotepense migrants crossed the border most recently in the San Diego/Tijuana area, although the percentage that chose to cross through all points on the California- Mexico border declined from 97 percent before 1995 to 79 percent after The most obvious reason that Tlacotepenses choose to cross the border via Tijuana is because most of them have San Diego County as their final destination. Other scholars consider final destination to be a useful predictor of crossing location (Orrenius 2004). Crossing through the San MixtecaBook.indb 54 2/6/09 9:31:58 AM

26 Strategies for Success in Border Crossing 55 Diego/Tijuana area is also the logical choice for unauthorized migrants who opt to cross through a legal port of entry. The San Ysidro port is the world s busiest land-border crossing, with more than 105,000 people passing through it every day. The sheer volume of traffic makes close inspection of vehicles and passengers impossible. The vast majority of undocumented Tlacotepenses (82 percent) walked across the border on their most recent trip (see table 2.6). This mode of entry encompasses the treacherous, multiple-day journeys on foot through the remote desert areas of Southern California and Arizona that have garnered much media attention due to the large numbers of migrant fatalities that result. However, most of our Tlacotepense interviewees reported crossing through the Tijuana/San Diego sector at points not so far east of the legal ports of entry in San Ysidro and Otay Mesa. While the journey over the cerro (mountainous areas) in East San Diego County is considerably shorter than crossing through more remote desert regions, it nonetheless involves grave physical dangers, as noted by Alfonso, whose wife and children crossed there: It s very ugly, very ugly.... It s cold and windy, there are sandstorms.... It s very tiring, with dust blowing into your eyes. You get very tired. Concerns about personal safety when crossing on foot were common among our respondents, as illustrated by Consuelo, a 64-year-old migrant living in California: The biggest problem is the hills. Not only are people dying there, but the coyotes are stealing from people and raping the women. Regardless of the perceived dangers, for the immigrant who cannot afford the exorbitant fees charged to cross through a legal port of entry in a vehicle or with false documents, crossing on foot continues to be a popular and by our statistical indicators highly successful mode of unauthorized entry. Table 2.6 Border-Crossing Methods, Pre-1995 to Crossing Method Before 1995 (N = 99) Year of Most Recent Crossing (N = 60) (N = 74) (N = 36) Walking 85.9% 81.7% 77.0% 80.6% Vehicle 11.1% 18.3% 21.6% 19.4% Other 3% 0% 1.4% 0% Total 100.0% 100.0% 100.0% 100.0% MixtecaBook.indb 55 2/6/09 9:31:58 AM

27 56 Parks, Lozada, Mendoza, and García Santos Nearly one-fifth of the undocumented Tlacotepenses we interviewed had passed through a legal port of entry (garita) on their most recent trip to the border. This method of crossing has increased significantly in popularity since 1995, as shown in figure The most common method of crossing through a legal port of entry is in a car, truck, or van, either concealed in a compartment of the vehicle or as a passenger holding false or borrowed documents. This approach yields a lower probability of being apprehended, given the thousands of people crossing daily through legal ports of entry. Additionally, corrupt customs agents have been known to work in tandem with coyotes, eliminating the risk of apprehension. Nevertheless, clandestine entries in vehicles may involve physical discomfort. Briseida recalls her third attempt to cross the border hidden in a vehicle: I was hidden in the car, where the glove box is. They had made a hole and taken out that part. And they fit me in there.... They folded me up like a taco. Figure 2.14 Border Crossings through a Legal Port of Entry, Pre-1995 to Percent Pre Year Among undocumented Tlacotepenses, we find that women are more likely than men to cross through a legal port of entry: 27 percent of women crossed through a garita on their most recent trip to the United States, while only 19 percent of men did so. No women swam across a river or MixtecaBook.indb 56 2/6/09 9:31:58 AM

28 Strategies for Success in Border Crossing 57 irrigation canal, drove their own cars through a legal port of entry, or crawled through transborder tunnels (a few men did). Briseida chose to go hidden in a vehicle, and despite having been apprehended several times, she still recommends this as the best option for a single woman: The thing is, it s less dangerous. They apprehend you, but that s it. You know they are going to arrest you and send you back, but you re alive and then you try again. On the other hand, crossing through the desert, you don t know if you re going to make it. What if they leave you halfway? It seems a lot more risky to me. CONCLUSION Has increased border enforcement discouraged potential migrants from deciding to migrate or prevented them from crossing the border illegally? Our research shows that the Border Patrol apprehends most Tlacotepense migrants at least once on an average trip to the border. Yet despite the rising probability of apprehension, we have seen that nearly all trips to the border eventually result in successful (undetected) entry into the United States. We have explained this high success rate as a function of changing border-crossing strategies that depend on coyotes with access to advanced technologies and border expertise. Since 1993, U.S. border enforcement strategy has entailed more fences, more Border Patrol agents, and more advanced surveillance technology, such as remote video surveillance systems. The annual budget of the U.S. Border Patrol in FY 2006 was $1.6 billion, an increase of 332 percent since 1993 (Ewing 2008, 7). Moreover, the Customs and Border Protection agency s budget increased from $5 billion in 2002 to $9.3 billion in 2008 (Barry 2008). Nevertheless, this approach to immigration control has yielded only marginal successes and has generated numerous unintended consequences, including higher rates of permanent settlement by undocumented migrants and their dependents in the United States, a burgeoning people-smuggling industry, and a sharp increase in the number of migrant fatalities at the border. This chapter has presented evidence of a learning effect to explain cyclical changes in the efficacy of border enforcement. The United States remains susceptible to creative forms of clandestine entry from Mexico. Migrants have consistently devised new methods of entry as security MixtecaBook.indb 57 2/6/09 9:31:58 AM