Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant? [1 ed.] 9781593326876, 9781593324698

Citation preview

The New Americans Recent Immigration and American Society

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Edited by Steven J. Gold and Rubén G. Rumbaut

A Series from LFB Scholarly

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved. Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies From Sojourner to Emigrant?

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Florian K. Kaufmann

LFB Scholarly Publishing LLC El Paso 2011

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011 by LFB Scholarly Publishing LLC All rights reserved. Library of Congress Cataloging-in-Publication Data

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Kaufmann, Florian K. (Florian Kristof), 1979Mexican labor migrants and U.S. immigration policies : from sojourner to emigrant? / Florian K. Kaufmann. p. cm. -- (The new Americans: recent immigration and American society) Includes bibliographical references and index. ISBN 978-1-59332-469-8 (hbk. : alk. paper) 1. United States--Emigration and immigration--Economic aspects. 2. Mexico--Emigration and immigration--Economic aspects. 3. United States--Emigration and immigration--Government policy. 4. Foreign workers--United States. 5. Mexicans--United States. I. Title. JV6471.K38 2011 331.5'440973--dc23 2011017858

ISBN 978-1-59332-469-8 Printed on acid-free 250-year-life paper. Manufactured in the United States of America.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

To Katrin

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved. Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Table of Contents

List of Tables ........................................................................................ix List of Figures.......................................................................................xi Acknowledgments ..............................................................................xiii Chapter 1: U.S. Immigration Policy: Attracting Permanent, Undocumented Mass Migration? .................................................1 Chapter 2: Social Network Assistance in Mexico – U.S. Migration......7 Chapter 3: Emigrant or Sojourner? Migration Intensity and Its Determinants ..............................................................................37 Chapter 4: A Comprehensive Model of International Migration Behavior .....................................................................................65 Chapter 5: The Impact of U.S. Border Enforcement............................93 Chapter 6: Mutually Beneficial Immigration Policies: A Utopia?..... 135 Appendix A: Data Appendix ............................................................. 151 Appendix B: Mathematical Appendix ............................................... 155 Appendix C: Additional Tables and Figures...................................... 165 Appendix D: Additional Specifications and Robustness Tests .......... 189 Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

References.......................................................................................... 193 Index .................................................................................................. 213

vii

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved. Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

List of Tables

Table 2.1: Migration Process and Social Network Assistance (Original Data)…………………………………………………... 25 Table 2.2: Relative Importance of Types of Social Network Assistance (Original Data)............................................................ 26  Table 2.3: Frequency and Assurance of Network Assistance by Network Type ............................................................................... 28  Table 2.4: Frequency of Social Network Assistance by Assistance Type (Secondary Databases) ........................................................ 30  Table 2.5: Explaining the Receipt of Various Network Assistance Types (MMP) ............................................................................... 32  Table 3.1: Migration Intensity Indicators for U.S. - Mexican Case (Cross-Section) ............................................................................. 50  Table 3.2: Linear Correlation Coefficients of Mexican Migration Intensity Indicators ....................................................................... 52  Table 3.3: Mexican Index of Migration Intensity (PCA Results) ........ 54  Table 3.4: Cross-Sectional Determinants of Migration Intensity......... 58  Table 4.1: Optimal Solution of Internal Migration Strategy ................ 76  Table 4.2: Simulating Various Migration Strategies............................ 82  Table 5.1: The Determinants of Mexicans’ Propensity to Migrate Illegally....................................................................................... 110  Table 5.2: Mexicans’ Propensity to Migrate Illegally: Additional Specifications and Compositional Effects .................................. 114  Table 5.3: The Determinants of Mexican Migration Intensity........... 118  Table 5.4: Mexican Migration Intensity: Additional Specifications and Compositional Effects.......................................................... 122  ix

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

x

List of Tables

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Table 5.5: Border Enforcement’s Direct and Indirect Impact on Undocumented Migration ........................................................... 128  Table 6.1: Potential U.S. Immigration Policies and Their Expected Effects......................................................................................... 142  Table C.1: Summary of Literature on Social Network Assistance .... 166  Table C.2: Social Network Assistance in Other Mexican Migration Studies ........................................................................................ 171  Table C.3: Social Network Assistance by Type of Helper (MMP).... 171  Table C.4: Profile of Mexican Migrants and Social Networks .......... 172  Table C.5: The Cross-Sectional Determinants of Migration Intensity: Descriptive Statistics .................................................. 173  Table C.6: Correlation Coefficients of IMI Indicators (Alternative Indicators)................................................................................... 175  Table C.7: Robustness of the IMI ...................................................... 176  Table C.8: Simulation Baseline Parameter Values and Ranges ......... 177  Table C.9: Econometric Studies on the Effect of U.S. Border Enforcement on the Flow of Mexican Migration........................ 178  Table C.10: Determinants of Mexican Migration Behavior: Descriptive Statistics .................................................................. 180  Table C.11: First Stage Regression Results of Extensive Migration Behavior...................................................................................... 182  Table C.12: Intensive Migration Behavior: Relative Magnitudes of Coefficients................................................................................. 183  Table C.13: Determinants of Mexican Migration Behavior: Additional Specification Tests.................................................... 184 

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

List of Figures

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Figure 3.1: Sample Distribution of Mexican Index of Migration Intensity ........................................................................................ 55 Figure 3.2: Mexican Migration Intensity over Time............................ 62  Figure 4.1: Simulation of Policy Changes ........................................... 85  Figure 5.1: Mexicans' Average Propensity to Migrate Illegally ........ 102  Figure 5.2: Index of Migration Intensity over Time .......................... 104  Figure 5.3: Line-Watch Hours of U.S. Border Patrol Agents ............ 105  Figure C.1: Robustness of the IMI to Changes in Extensive Migration Behavior..................................................................... 186 Figure C.2: Simulation of Optimal Migration Strategy ..................... 186 Figure C.3: Simulation of Condition 1 and Condition 2 .................... 187 Figure C.4: Simulation of Additional Conditions .............................. 188

xi

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved. Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Acknowledgments

This book originates from my doctoral research at the University of Massachusetts Amherst. Many people have assisted me during the writing of this book. I am particular grateful to James Boyce, whose inspiring advice and guidance has been crucial. I am indebted to: Melissa Gonzalez-Brenes for her support with my field research; Peter Skott for his help with the migration model and critical comments; and Bernard Morzuch for improving previous drafts of this manuscript. I thank Samuel Bowles for rousing my interest in economic modeling. I thank Aline Sayer, Leo Chavez, Jennifer Cohen, Mathieu Dufour, Lynda Pickbourn, Mike Carr and participants of the News School UMass Conference for their comments on my manuscript. I am thankful to Phil Mellizo and Hwok Aun and Jacqui Lee for their fellowship. The presented field research benefited from many. I am especially grateful to Martin, Leonardo, Hilda, Pablo, Gorge and his family, JoseLuis, Silverio, Javier, Toribio, Alejandra, Soledad, Noe, Antonio, Gabriela, Bieto, Maelli, Arturo, and Margarita. I was humbled by the generosity of the residents in my research communities. Their life stories inspired me to write this book. I want to thank the Political Economy Research Institute and the Barkin family for their financial support. I am grateful to the staff of the Mexican Migration Project and the Encuesta sobre Migración en la Frontera Norte de México as well as Gordon Hanson and Pia Orrenius for making their databases available to me. I dedicate this book to my beloved wife Katrin. Without her patient support this book would not have been written.

xiii

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved. Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

CHAPTER 1

U.S. Immigration Policy: Attracting Permanent, Undocumented Mass Migration?

“Not only were undocumented migrants dispersing more widely in the wake of the new enforcement regime, but they were staying longer north of the border. A perverse consequence of draconian border enforcement is that it does not deter would-be migrants from trying to enter the country so much as it discourages those who are already here from returning home. The end result of a border buildup is typically longer trip durations, lower probabilities of return migration, and a shift toward permanent settlement.” [Massey et al. 2002, pp. 128-129]

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

INTERNATIONAL MIGRATION AND ITS IMPORTANCE TO SENDING AND RECEIVING COUNTRIES Since 1980, the worldwide number of international migrants has more than doubled, reaching 214 million in 2008 (United Nations 2008). Economic theory would support such a development, since the efficiency gains from exploiting the sizable international wage differences are estimated to be enormous (World Bank 2006b). Low- or middle-income countries often possess an abundant labor stock but insufficient capital to serve it. Hence, a labor export-led growth model, that is, exporting cheap domestic labor into other countries, may be a 1

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

2

Mexican Labor Migrants and U.S. Immigration Policies

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

desirable strategy. Migrant workers’ increased incomes and skills, if repatriated, could then help to develop the domestic capital stock. Moreover, the generated foreign exchange could be used to service foreign debt in the source country, reducing its country risk rating, which should attract additional foreign investment flows. For highincome countries, the inflow of foreign workers could offset the projected future labor shortages due to aging societies (United Nations 2002). However, unlike in the case of foreign direct investment or international trade flows, no powerful lobby has emerged to advance the liberalization of international labor markets.1 One reason thereof, seems to be that sending countries have long doubted that the outmigration of their citizens could benefit their own country. Another reason is that a public discussion of free international labor markets is effectively off-limits in high-income countries. Apparently, industrialized nations’ main interests are to adjust the inflow of foreign labor to the perceived need of their domestic labor market - often shortages in cheap labor or certain skill sets - but most importantly, to keep the inflow of “undesirable” migration in check. Integration of immigrants is a major concern for high-income countries, especially in Europe, as ongoing debates underline. To achieve this desirable and controllable immigration mix, external border enforcement has emerged as one of the most popular migration management techniques, despite its often devastating effects on migrants (McKinley 2010, GAO 2006) and debatable effectiveness (Hanson & Spilimbergo 2001, Reyes et al. 2002). In understanding the popularity of border enforcement the collaboration of common sense argumentation and economic mainstream theories should not be underestimated: If potential migrants seek higher earnings and have only two choices, staying at home to earn a marginal wage or emigrating to earn an industrial-level wage, then border enforcement must deter migration, if only it incurs sufficiently high costs to outweigh migrants' wage gain.

1

Since 1995, the General Agreement on Trade in Services (GATS) Mode 4 of the World Trade Organization allows the temporary movement of natural persons to supply services abroad. Yet, Winters et al. (2002) note that only “limited commitments” have been made, which largely relate to the movement of inter-company personnel.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Attracting Permanent, Undocumented Mass Migration?

3

Yet, restricting migration behavior to a stay-versus-migrate decision, as major economic theories of cross-border labor movement do (Harris and Todaro 1970, Lucas and Stark 1985, Borjas 2009), fails to account for the complexity of international migration. International migration affects multiple facets in the life of migrants and societies. Due to the involved transaction costs, such as border crossing expenses, physical or psychological deprivations and other, the decision to migrate should not be separated from related aspects of international migration behavior, such as family reunification, remittance sending, or private investment. Unlike investors or traders, migrants cannot separate the location of their economic activity markedly from the location of their private life. Furthermore, migrants are not restrained to either stay or emigrate, but have versatile migration strategies (e.g. Massey et al. 1987, Schiller et al 1992, Portes and Rumbaut 2006, Donato et al. 2010). Accounting for these facts has an importance beyond mere theoretical interest, as they could affect policy making in both sending and receiving countries. If migrants can choose among a suite of migration strategies, then the main consequence of tightening borders may not be to deter migration, as the simplistic stay-versus-migrate framework suggest, but to encourage more temporary migrants to become permanent, i.e. stop visiting home to avoid re-entry barriers. Furthermore, as return becomes more difficult, temporary migrant workers may start to move their family abroad to settle. Governments in sending countries might wonder whether different “migrant types” contribute differently to domestic development. In this respect, a discussion has emerged among economists on whether “skilled” emigration may cause a detrimental “brain drain” from sending countries (Carrington and Detragiache 1999) or on the contrary stimulates a “brain gain” (Stark et al. 1997, Beine et al. 2003). What has received little attention so far is whether migrants’ migration strategy, “permanency,” or degree of socio-economic detachment from their country of origin should be of concern to sending countries. For example, if “remittance decay” exists (Stark 1978), then emigration would subvert the above-sketched labor export-led growth model because few remittances and human capital may be repatriated to benefit the domestic capital stock. In this case short-term migration programs might be more desirable than permanent legalization for sending countries.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

4

Mexican Labor Migrants and U.S. Immigration Policies

MEXICO – U.S. MIGRATION: FROM TEMPORARY TO PERMANENT MASS MIGRATION?

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

The long-standing tradition of unskilled Mexican labor migration to the U.S. offers many interesting insights into the nexus of migration and development.2 For one thing, a vast, naturally easy-to-cross border separates two countries whose income levels are drastically different. Unsurprisingly, this has resulted in mass migration, traditionally mostly seasonal or temporary (Massey et al. 1987).3 For another thing, both governments seem to have long accepted this mass migration flow, at least tacitly (Massey and Singer 1998; Rosenblum 2004), but have recently started to manage it more actively. On the one hand, the Mexican government has started to reach out to its large diaspora in the U.S., intending to incorporate it into domestic development initiatives (Goldring 2004b). On the other hand, the public outcry in the U.S. against the rising “flood of illegal migration”, mainly Mexican, has grown very loud, prompting the U.S. government to “tighten borders” (Massey et al. 2002). However, the escalation of U.S. border enforcement has not reduced the volume of undocumented Mexican migration. Instead there are indications that the flow of undocumented Mexican migration has steadily grown up until the early 2000s (Passel & Suro 2005), has diversified and has become more permanent in the 1990s (Cornelius and Lewis 2007; Reyes et al. 2002). This new situation – unabated undocumented mass migration, but now more permanent – suits neither the U.S. nor Mexican government, nor does it necessarily reflect the first choice of migrants themselves, if given different options.

2

An overview of Mexico – U.S. migration may be taken from the U.S. Commission on Immigration Reform (1998), Massey et al. (1987, 2002), or Hanson (2006). 3 In 2006, 28.3 million residents in the U.S. were of Mexican descent (Pew 2006), about 6.7 million were undocumented Mexican immigrants (DHS 2010).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Attracting Permanent, Undocumented Mass Migration?

5

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

UNDERSTANDING PERMANENT, UNDOCUMENTED MASS MIGRATION The book is structured around five major questions: First, if destination countries try their best to keep “unwanted” migrants out, how do hundreds of thousands of Mexicans manage to successfully enter the U.S. without documentation each year? More specifically, what accounts for social networks importance in facilitating undocumented migration, both theoretically and empirically? Second, can and should migrants' degree of socio-economic reorientation be measured? Third, what effects do different immigration policies have on individuals' decision whether or not to migrate and their degree of socio-economic reorientation, as reflected in return, remittance sending, family reunification and other choices? Fourth, does U.S.-Mexican migration data support these hypotheses? In particular, what accounts for the evolution of permanent, undocumented mass migration? Lastly, which suggestions follow for U.S. and Mexican policy makers? Is there hope for mutually beneficial policies? What may other policy makers, such as European, learn from the U.S.-Mexican case? As a first entry point, the following chapter investigates the significance of social networks in facilitating undocumented Mexican migration to the U.S. I structure the migration process of an undocumented first-time Mexican migrant into six major steps and conduct an in-depth analysis of his challenges. I argue that the main challenges he faces require trust to overcome, accounting for the importance of social networks. Drawing on secondary data sets as well as own field research, I quantify the extent of social network assistance, disaggregated by type of assistance, helper and reliability. Chapter 3 develops the concept of migration intensity, defined as the degree to which a migrant shifts his attachment, association and engagement from his place of origin to the migration destination. I argue that “intensive migration behavior” should be seen as complementary to “extensive migration behavior,” the stay-versusmigrate decision. I propose a general, country-independent measurement procedure to quantify per se unobservable migration intensity by applying principal components analysis to a set of observable choices that reflect migrants’ degree of socio-economic reorientation. Using data for male Mexican migrants in the years 1950 to 2005, I find strong complementarities among remittances, migration patterns, and investment decisions, allowing me to derive an Index of

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

6

Mexican Labor Migrants and U.S. Immigration Policies

Migration Intensity (IMI). The IMI shows that male Mexican migrants use a continuum of labor migration strategies. Its cross-sectional variation accords with a priori expectations. In Chapter 4 I model intensive and extensive migration behavior jointly. Augmenting a conventional Harris-Sjaastad-Todaro model, I develop a simultaneous model for the initial migration, return, repeat migration, remittance and other decisions of migrant workers. This model can incorporate various migration strategies, including “circular migration,” “target earning,” and “emigration.” I suggest that excluding migrants’ choice of their degree of socio-economic reorientation gives an incomplete or even misguiding policy guide of their behavior. For example, I find that stricter border enforcement has two consequences: an intended deterrence effect, and an unintended intensification effect whereby stricter border controls lead migrants to make fewer return trips, prolong total U.S. time, reduce remittances, and move family members to the destination country. The impact of the latter on origincountry incomes may induce others to migrate as well. I use numerical simulation, calibrated with U.S.-Mexican migration data, to illustrate the effects of various immigration policies. I then examine the determinants of Mexicans’ propensity to migrate illegally (extensive migration behavior) and their degree of socio-economic reorientation from home (intensive migration behavior; summarized by the IMI). This is done by estimating two reduced-form equations, using instrumental variables estimation and individual fixed effects on a panel of micro data. My findings support the hypotheses that stricter U.S. border enforcement leads to higher migration intensity, which in turn leads to an increase in the volume of undocumented Mexican migration, presumably due to the reduced transfer-incomes of “stayers.” As the deterrence effect of U.S. border enforcement is negligible, the net effect of increased U.S. border enforcement is to spur more undocumented Mexican migration. Politicians seeking to limit illegal migration would be better advised to attempt to narrow the U.S. – Mexican wage gap, since the resulting direct as well as indirect effect – the fostering of migrants’ home attachment – would both deter Mexicans from migrating. The final chapter evaluates different policy options as to their benefits and costs. I argue that, in their own national interests, the U.S. and Mexican governments should favor regularized circular migration programs in the short-run and initiate a development partnership in the long- run.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

CHAPTER 2

Social Network Assistance in Mexico – U.S. Migration

The Department of Homeland Security (2010) estimated that 11.6 million undocumented migrants resided in the U.S. in January 2008, of whom 62 percent were of Mexican descent. The net inflow of undocumented aliens in-between 2000 and 2004 alone was estimated to average 600,000 per year; an amount about the population of Wyoming, Alaska, North Dakota or Vermont (see also Passel and Cohn 2009). Mexican migration has not only grown in numbers over the last two decades, but has also expanded by including migrants from untraditional sending states or dependents (Cerrutti and Massey 2001); both migrant types tend to be most vulnerable. The conundrum is that at the same time the U.S. has quadrupled its border enforcement resources.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

SOCIAL NETWORK ASSISTANCE IN MIGRATION Sociologists and anthropologists have long pointed to a “family and friends effect” (Levy and Wadycki 1973) or “chain migration” (Macdonald and Macdonald 1974) in explaining mass migration (see also Tilly and Brown 1967 or Jones 1982). Field researchers have reported that the lives of Mexican migrants are characterized by social network assistance (e.g., Reichert and Massey 1979; Mines 1981; or Lopez 1986).4 This lead Massey et al. (1987) to suggest a process of 4

I consider social networks to be a means by which members of social networks access and exchange resources, such as information, services or

7

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

8

Mexican Labor Migrants and U.S. Immigration Policies

“cumulative causation” in which social networks form a “social infrastructure” that enables Mexican mass migration. Indeed, searching the migration literature within the social sciences for indications of network assistance gives an abundance of results. Table C.1 in Appendix C lists about 60 studies that I have found to relate to social network assistance in migration. Most are case studies of Mexico – U.S. migration, but I include also literature reviews, econometric studies, and other types. The table summarizes reports of social network assistance in migration, which I classify into six categories: information and risk, initial financing, domestic dislocation, border crossing, initial U.S. sustenance, and U.S. employment. In an attempt to categorize the studies, I noted if it focused on a certain type of network assistance (F), gave further examples or explanations for it (E), provided statistics on it (S), or merely listed the assistance (L). Furthermore, I include information on the investigation’s main research question. Types of Social Network Assistance

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Scholars have long pointed out that migrants provide valuable information about opportunities in foreign countries to fellow countrymen (MacDonald and MacDonald 1964). Durand (1994) notes that in rural Mexico adolescents learn the “rules of the game” and the “costs, sacrifices, and objectives of migration” when joining the public gatherings of return migrants.5 Information from social networks is generally assumed to reduce the risks associated with migrating6 but may also influence a potential migrant’s expectations. Some scholars even claim that a “migration culture” (Massey et al. 1987) or “migrant syndrome” (Reichert 1981) has emerged in many rural Mexican communities.

goods. Social networks can thus be considered as the source of “social capital” (Bourdieu 1986; Coleman 1988; Dasgupta 2002). 5 Boyd (1989) and Goldring (1998) note other ways in which social networks may convey migration information, such as phone calls, letters or public consumption of remittances. 6 Taylor (1986) argues that social networks’ information reduces migrants’ risks by aligning expected and true variances. See also Menjivar (2002) and Zahniser (1999).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Social Network Assistance in Mexico – U.S. Migration

9

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Social networks have been found to help undocumented migrants enter the U.S. They can inform prospective migrants about available border crossing techniques, offer advice on how to avoid capture by the U.S. Border Patrol and how to behave if captured, recommend a reliable and affordable smuggler, or even pre-arrange a smuggler contract (e.g., Massey and Singer 1998; Durand 1994). A more advanced service, which still seems not well known in the literature, is the cash on delivery payment (COD) of smugglers by a member of the migrant’s social network to guarantee his safe border crossing (Durand 1994; Spener 2009). Moreover, the travel companionship of family and friends has been found to reduce travel costs, provide moral support, and protect migrants against victimization (Graves and Graves 1974; Lopez 1988; Bastida 2001). A handful of studies have examined how social networks support migrants upon arriving at the destination. Initially, Mexican arrivals often appear to be treated as guests without the host’s expectation of compensation (Lomnitz 1977; Espinosa 1997). This seems to change if the co-residence lasts or the guest finds a job (Mines 1981; Menjivar 1995). Chavez (1992) argues that long-term co-residence of undocumented migrants in the U.S. is a strategic choice to pool resources against economic hardship and the threat of deportation (see also Glick 1999). Van Hook and Glick (2007) find that recent Mexican migrants choose different household strategies than Mexican “stayers” or more settled immigrants, characterized by a high likelihood of shared living arrangements and high household turnover (“port-in-astorm host households”). Social networks’ help to migrants in finding employment is the most cited assistance. Economists stress social network’s information provision function, both general labor market advice and information on job vacancies (Banerjee 1984; Caces 1986).7 Kin or friends also seem to refer or recommend new arrivals to U.S. employers or even pre-arrange employment for them (Banerjee 1990; Waldinger 1997). Migrants that have advanced to supervisor positions have been found to rely on network-based recruitment (Waldinger).8 Social networks may 7

Munshi (2003) reports that migration networks improve Mexican migrants’ odds and types of U.S. employment (see also Espinosa 1997; Aguilera and Massey 2003). 8 Adler (2002) argues that active recruitment of a migrant-supervisor improves his position towards employer and co-workers. Yet, Waldinger (1997) shows

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

10

Mexican Labor Migrants and U.S. Immigration Policies

also provide rookies with training or job orientation (Lomnitz 1977; Mines 1981; Caces 1986). Some studies report that social networks help to finance migration trips (Morawska 1990; Papail 1998). Such family financing is a central element of the New Economics of Labor Migration (Stark and Bloom 1985). Goldring (1992) reports that social networks help migrants to sustain themselves while waiting to cross the border. Other studies have analyzed social networks’ impact on migrants’ settlement or acculturation (Portes 1990), as well as networks’ moral support (Kemper 1977), emergency assistance (Lomnitz 1977), insurance function (Cuecuecha 2008), or help in daily chores (Lomnitz 1977).

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Empirical Findings on Social Network Assistance In spite of an abundance of qualitative investigations on social networks and migration, very few studies quantify social network assistance in migration. The few quantitative studies that I have found rely on small samples and cover only some types of network assistance. Espinosa (1997), who interviewed 518 Mexican migrants, is the most representative and comprehensive study. Moreover, as Table C.2 in Appendix C shows, studies differ considerably in their estimates of the prevalence of social network assistance to Mexican migrants. Between 40 and 95 percent of Mexican migrants appear to receive initial U.S. lodging assistance and between 40 and 75 percent seem to be helped in finding a U.S. job. Espinosa ranks several types of social network assistance that Mexican migrants receive by frequency. She reports that help in getting a U.S. job comes first, then U.S. lodging and food support, and last aid in crossing the border. Jointly consulting the scattered sources reported in Table C.2 would support this ranking, except for placing initial U.S. lodging and food assistance first. However, deducing the relative importance of a given type of assistance from solely considering observed frequencies may be dangerous: a frequently observed type of assistance may simply be an easy one to provide. Examining migrants’ efforts to ascertain their receipt of some assistance and looking at the type of helper they try to recruit for it – e.g., trustworthy or not - would

that exuberant network recruitment may undermine managers’ power towards a unified labor force.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Social Network Assistance in Mexico – U.S. Migration

11

be a better indicator of its relative importance. Furthermore, the ranking should be comprehensive.9 The consensus in the larger literature on social networks and social capital seems to be that relatives are more important helpers than nonrelatives since ties to kin are stronger, making their help more reliable (e.g., Choldin 1973; Lomnitz 1977). In the case of Mexican migration to the U.S. Massey et al. (1987), using qualitative observations, have suggested a ranking of helpers according to the “probability and reliability” of their help: nuclear family comes first, then extended family, friends, and paisanos (“fellow countrymen”). Yet, Van Hook and Glick (2007) and Espinosa (1997) challenge Massey et al.'s hierarchy of assistance. Van Hook and Glick find that extended families as well as non-kin are more important helpers for Mexican migrants than nuclear family relatives, when comparing to common practices in Mexico. Espinosa’s estimates show that help from friends is often the most frequent for Mexican migrants. She further detects indications of specialization: friends matter most for border crossing and job search, kin are most important for initial financing.10 When researchers included indicators of “migration networks” into regressions trying to explain Mexicans’ propensity to migrate, most found a significant spurring effect (e.g., Taylor 1987; Davis and Winter 2000). However, applying Manski’s (1995) critique to these econometric studies shows that these studies cannot be certain to have identified a social network effect rather than a shared contextual effect (see also Durlauf and Fafchamps 2004).11 Variables, which are

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

9

Drawing on bi-national census data, Van Hook and Glick (2007) find that recent immigrants are 24.4 percent more likely to live in an extended family or non-kin household than their peers in Mexico and 19.6 percent more likely than more settled immigrants. Interestingly, they also find that households of recent migrants show a high turnover, indicating a highly dynamic migration pattern. 10 In the case of migrants’ job assistance, qualitative studies disagree whether “strong ties” (Waldinger 1997) or “weak ties” (Garcia 2005; Sassen 1988) are more important. 11 Manski shows that if individual behavior is modeled as a linear function of the mean behavior of a group - such as the number of group members that have migrated (“migration networks”) - then it is not possible to distinguish whether the mean group behavior (i.e. migration networks), or personal or exogenous factors that the group shares - such as similar ethnicity or wealth, or sufficient rainfall - drives the individual behavior.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

12

Mexican Labor Migrants and U.S. Immigration Policies

exogenous to such confounding group-level effects and therefore may serve as instrumental variables for social networks, have not been found in the case of Mexican’s decision whether or not to migrate.12 The Case for a Micro Perspective on Social Network Assistance

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

An abundance of field researchers have reported incidences of social network assistance in migration. Econometric studies have generally found that social networks influence the decision whether or not to migrate (e.g., Palloni et al. 2001), but in the case of Mexican migration appear to have reached a dead-end in ruling out a Manskian “reflection problem.” One way to proceed may be to inquire why social networks appear to matter so much. Conceptually, an explanation of the “why” needs to study the migration process from the perspective of its main actor: a potential first-time migrant without immigration documentation. In analyzing the migration process, one should elicit the migrants' main challenges and uncertainties, which facilitating or assisting role social networks play, and what networks' competitive appeal is as opposed to other migration conduits. Conducting such a micro analysis should lead to empirically testable hypotheses about the expected overall prevalence of social network assistance in Mexican migration as well as an expected hierarchy of assistance types and helpers (those who provide assistance). Empirically, despite an abundance of qualitative studies, social network assistance in Mexican migration still lacks comprehensive and representative quantification to support or reject migration networks’ empirical relevance. Developing an empirical hierarchy of assistance types and helpers would contribute to the understanding of how migration networks work. Such a hierarchy should not only be based on frequency observations, which existing studies rely on, but also include other criteria, such as trustworthiness – a concept already used to rank different helpers theoretically (Massey et al. 1987), but still without empirical testing.

12

While Munshi’s (2003) instrumental variable - past rainfall - can be used to analyze migrants’ U.S. labor market outcomes, it cannot serve as an instrument for the stay-versus-migrate decision, because past rainfall is likely to directly affect employment in Mexico (but not in the U.S.).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Social Network Assistance in Mexico – U.S. Migration

13

My study also contributes to the emerging economic literature on “neighborhood effects” or “social interactions.” Overall this literature seems to assume that social networks’ main function is information provision, such as new business practices (Udry 2004), health practices (Behrmann et al. 2002; Miguel and Kremer 2003), job information (Munshi 2003), or other (for a summary see Durlauf and Fafchamps 2004). However in the case of undocumented migration, it is likely that social networks also provide more “concrete” assistance, such as in border crossing or initial sustenance abroad. Moreover, the research questions in this chapter are policy relevant. Driving up undocumented migrants’ costs and risks at different points along the migration process seems currently the focus of U.S. migration management practices. In order to better evaluate the potential of such “cost-shock policies,” it seems prudent to consider the possible influence of the apparently dominant facilitator of undocumented migration, social networks. Knowing the overall prevalence of social network assistance as well as its functional focus and distribution of helpers will indicate migration networks’ role, effectiveness and resilience in helping undocumented migrants to elude such costs-shock policies: Are social networks only a minor facilitator? Do they mainly provide unimportant types of assistance? Are migrants dependent on specific helpers or can they rely on various network members?

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

THE MEXICAN – U.S. MIGRATION PROCESS: A MICRO PERSPECTIVE This section develops a microeconomic analysis of the migration process of undocumented Mexican first-time migrants, taking into account first-time migrants’ challenges, information problems, and options, in order to better understand the apparently prominent role that social networks play in facilitating undocumented migration. Since my main interest is the impact of social networks on Mexicans’ first-time migration decision, I only consider the migration process from initial decision making up to the point at which a migrant obtains his first job in the U.S. (hereafter referred to as the “migration process”).13

13

Undocumented Mexican migrants face further challenges once they have obtained a job, such as settlement or integration, but these later migration steps

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

14

Mexican Labor Migrants and U.S. Immigration Policies

The interdisciplinary literature review in the previous section as well as my own interviews with Mexican migrants and non-migrants (see next section for details on data collection) lead me to the conclusion that an undocumented Mexican first-time migrant generally passes through six major migration steps: information gathering for decision making, initial financing, domestic dislocation, border crossing,14 initial U.S. sustenance, and U.S. job search. In pondering undocumented first-time migration I assume that a potential migrant considers each of these migration steps jointly, given his current information level. If he reaches the conclusion that the risk-adjusted benefits of undocumented migration outweigh the risk-adjusted costs, including opportunity costs, he will choose to migrate. From Decision Making to First U.S. Employment: Six Milestones

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Information on Migration: Before considering migration to the U.S., potential migrants need to at least have a general idea about migrating. Given that international migration has strongly penetrated Mexican daily life, most adult Mexicans are likely to possess such “latent generalized information” (Banerjee 1984). More detailed general knowledge about the workings of the U.S. labor market, methods to cross the border, or ways to survive in the U.S - “direct generalized information” (Banerjee) - also seems to be widespread in many Mexican regions. However, only the possession of “specific information” (Banerjee) - about a certain smuggler, accommodation option, or job vacancy - would seem to reduce the risk of undocumented migration to such a degree that broad masses would consider migrating (see also Tilly 1990). In general, only social

seem less likely to influence the initial migration decision, warranting omission. 14 Spener (2009), focusing on border crossing at the South-Texas border, breaks border crossing down into two micro-steps: crossing the first sentry point (e.g. getting over the Rio Grande or an international bridge) and getting past the border zone and interior immigration checkpoints. However, he reports that all forms of smuggler services he found, except one, cover both micro-steps. This supports the bundling of border crossing into one migration step.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Social Network Assistance in Mexico – U.S. Migration

15

networks are able to offer trustworthy specific information and are also likely the best source of direct generalized information.15 What I have found common in interviews with Mexican migrants and non-migrants is the concept of an “invitation.” Typically, an invitation meant that an experienced relative or friend, who resided in the U.S. or intended a repeat trip, invited a potential migrant to come to the U.S., offering specific information and assistance.16 In the absence of an invitation, undocumented first-time migrants frequently called U.S. contacts or asked return migrants for information and assistance. Either way, this facilitated migration considerably. Initial Financing: Undocumented Mexican first-time migrants need to finance domestic expenses and initial expenses in the U.S. In interviews I found indications that these are frequently financed separately (see also Espinosa 1997). As expenses often surpass undocumented first-time migrants’ savings, most migrants have to find a creditor. Yet, the common principal-agent problem in lending arises (e.g., Conning and Udry 2005) with an additional aggravation. Not only must the creditor fear that his debtor cannot pay back if he is insolvent, due to limited liability, but even in the case of solvency a migrantdebtor could easily refuse repayment by remaining outside of the national jurisdiction (“opportunistic default”).17 Informal moneylenders or standard group-based lending arrangements, which typically rely on their ability to access local information, for example, supervising the debtor’s harvest, or local enforcement systems, such as reputation (Bowles 2006), face the problem that the migrant-debtor leaves the local sphere in which such mechanisms operate.18 Social networks, on the other hand, have unique monitoring and enforcement advantages due to ongoing communication with the migrant-debtor and because other kin and friends are frequently co-residents or co-workers of the migrant-debtor. Moreover,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

15

Official labor recruitment programs seem the only true exception to this. Smugglers might offer information but, given their business interest, they should not be trusted. 16 The extent of offered information or promised assistance varied. For a minority of invitations no help was promised, in which case the “inviter” was often a rookie, too. 17 Collateralizing the debt could be one solution but will often be unavailable. 18 This may explain why local smugglers or U.S. employers sometimes seem to become “employer-lenders” or “smuggler-lenders” (see also Menjivar 2000).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

16

Mexican Labor Migrants and U.S. Immigration Policies

altruism and multiplex future exchanges reduce the conflict of interest between creditor and debtor. This should apply strongest to nuclear family networks. Domestic Dislocation: Before being able to enter into the U.S. an undocumented Mexican first-time migrant has to physically relocate himself to a border crossing point if he is not a border resident. Then he has to sustain himself in the border region until he manages to cross into the U.S. Domestic relocation and border sustenance have received little attention in the literature, likely because they do not pose major challenges.19 Border Crossing: Undocumented Mexican first-time migrants need to find an illegal way to enter the U.S. Due to the tightening of U.S. border controls most undocumented Mexican migrants now seem to contract a border smuggler instead of trying to cross on their own (Spener 2004). Contracting a smuggler constitutes another principalagent problem, as the quotation of an anonymous smuggler in Lopez (1998, p. 973) forcefully shows:

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

“Up there at the border there’s lots of coyotes, but there’s no safer way than going with one right from here, like’s one got more guarantee, more safety it’s somebody you know and you’re with a group and they take better care of you, cause you never know what’s it’s gonna be like up there. […] You could find yourself with a bunch of people who are just starting and that’s tough. But these guys [...] if you don’t get across but get caught, they take you back across again. […] If you got contacts that’ll answer for you, you give them the telephone number and they call and then they take you to your house and they pay him there.” A first-time migrant cannot know ex-ante if he picks a lemon (Akerlof 1970) - a rookie smuggler, a high risk smuggler, or a brutal abuser (see quote) - nor can he verify ex-post if the smuggler acted diligently, let alone court-enforce misbehavior. Without an extralegal enforcement system, a rational migrant will refuse to pre-pay and a 19

Some of my interviewees reported that they first migrated internally to a border city, working there for some time (sometimes several years), and then decided to migrate to the U.S. To enter the U.S, some of them obtained tourist visas, which are only available to border residents.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Social Network Assistance in Mexico – U.S. Migration

17

rational smuggler will not accept deferred compensation due to fear of non-payment.20 One solution could be that the migrant promises future rents to the smuggler, such as recommending his services to relatives or friends, contingent upon current satisfaction.21 However, this is unlikely to fully resolve the conflict of interest between migrants and smugglers. Smugglers, migrants, and social networks have thus jointly come up with a contract that avoids problems of information asymmetry by moving “residual claimancy” to the agent, i.e., the smuggler (Bowles 2006): “cash on delivery” (Spener 2009). The smuggler is paid in full only once the migrant safely reaches his final or intermediate U.S. destination.22 To avoid robbery in transit, the U.S. payment is commonly made by a trusted social network member, who “responds” for them.23 Due to the unlimited number of entry trials granted as part of the contract, the migrant is insured against both smuggler misbehavior and exogenous risks, such as changes in U.S. border enforcement levels. Moreover, the responding U.S. contact often also serves as creditor for the smuggler payment and picks the migrant up at the delivery point (Chavez 1992, Spener 2009). In the end, Spener (2009) points to a mixture of extralegal enforcement mechanisms that act together in the closing and compliance of smuggling contracts: (enforceable) trust (i.e., the COD system), reciprocity (i.e., reputation effects) and altruism (“responding” for arriving friends or kin). All depend on social networks.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

20

Spener (2001, 2004) and Menjivar (2000) provide examples that show the danger of pre-payment. 21 Migrants’ preference for local coyotes rather than commercial border businesses (Lopez, 1998; Spener 2009) is likely explained by the importance of this reputation effect and the ability of migrants to informally sanction the local coyote (or his family). 22 It should be noted that now the smuggler becomes contingent upon a promise by the migrant. To enforce payment, smugglers appear to keep migrants at a U.S. safe house until their contacts provide payment - sometimes as hostages under threat of violence (Chavez 1992, Spener 2004), but not as a rule (Spener 2009). 23 Other money transfers (e.g., wire transfers) appear not to be a common solution. Sometimes another third party, such as the employer, steps in and finances the smuggler.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

18

Mexican Labor Migrants and U.S. Immigration Policies

Initial U.S. Sustenance: Undocumented Mexican first-time migrants are very vulnerable upon arriving in the U.S. Inexperience, alienation, language barriers, and immobility make it difficult for them to obtain accommodations, buy food or clothes, commute to their workplace, or handle other daily chores (Chavez 1992). Reliance on ethnic entrepreneurs who specialize in assisting newcomers (Adler 2002), paternalistic employers (Goldring 1992), or ethnic enclaves (Portes 1995) could be a way out. However, migrants generally cannot afford high U.S. living costs until they have secured employment. Moreover, reliance on such strangers would make first-time undocumented migration very risky from the pre-departure perspective.24 The prospect of initial lodging and food assistance as well as initial guidance by a relative or friend – i.e., an “initial safe haven” - considerably reduces this uncertainty. Such an initial safe haven further helps undocumented migrants to remain invisible (Chavez 1992). First U.S. Job: A major problem of labor markets is that employers possess information about job vacancies, but only workers know which job they would be willing to take (Montgomery 1991). Private and governmental intermediation, which has been created to solve this dilemma, is unlikely to be an option for undocumented first-time migrants due to ignorance, language problems and fear of apprehension. Ethnic labor recruiters could step in (Adler 2002) but would make migration risky from the pre-departure perspective. Social networks may be able to pre-arrange a job (Banerjee 1990) or at least promise to help in the job search, reducing the migrant’s pre-departure uncertainty. Employers worry about adverse selection when hiring. Signals to attenuate this problem, such as educational or job references, are likely not valid in the case of newly arrived aliens. This makes social networks a helpful screening device for employers (Waldinger 1997).25

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

24

From the first-time migrant’s pre-departure perspective, trustworthy support by an employer or ethnic entrepreneur seems to require intermediation by someone trusted or trustworthy third-party supervision, as in the case of an official guest worker program. 25 Referred workers self-select since they possess better information to judge if they match the job vacancy. Referees pre-select workers unintentionally by referring persons that resemble themselves and intentionally to protect their own reputation (Montgomery 1991). Furthermore, newcomer’s concern for

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Social Network Assistance in Mexico – U.S. Migration

19

Hence, social networks’ labor market intermediation is appealing to both the employer and the migrant-employee. The Most Needed Assistances and Best Helpers: Hypotheses

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

In the previous section I have provided evidence that asymmetric information between migrants and lenders, smugglers, or employers, as well as illegality make trust a crucial element to most exchangebased solutions to the challenges of first-time migration.26 This makes formal court-enforced contracts infeasible (Bowles 2006). Common informal monitoring or enforcement mechanisms have only limited appeal due the migrant’s departure from the local sphere, weakening informal mechanisms, and offering him a strong one-sided exit option. Since kin and friends typically accompany the Mexican migrant to the U.S. as co-residents or co-workers, social networks are able not only to help best but can also monitor and sanction misbehavior best. Hence, it seems likely that social network assistance to Mexican undocumented first-time migrants is very frequent. Social network assistance should be most important for those migration steps in which the migrant needs exchange-based solutions most and trust is most crucial for successful exchange. Issues of trust are likely to be most pronounced in the cases of lending, border crossing, and job intermediation. General information on migration is relatively simple to obtain in most Mexican regions. Financial assistance for initial U.S. expenses is important, but such credits seem to be intertwined with other types of assistance - border crossing and initial U.S. sustenance. The observable part of lending, for domestic expenses, seems less crucial. Despite the economic literature’s focus on labor market intermediation, failed border crossing or failed initial U.S. sustenance are likely to inflict the worst misery on a migrant, making social network assistance more important. Hence, I hypothesize that for undocumented Mexican first-time migrants, social networks’ help with border crossing and initial U.S. sustenance is most important; their their referee and peer-monitoring tends to keep the work morale up (Waldinger 1997). 26 Pioneer migrants have always managed to solve migration challenges by themselves, but full self-reliance would seem to put migration out of the grasp of the broad masses. See also Tilly (1990).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

20

Mexican Labor Migrants and U.S. Immigration Policies

assistances with information on migration, job search, and financing domestic migration expenses are of intermediate importance, and their support with domestic dislocation is least important. Furthermore, I conjecture that migrants try to recruit their most trustworthy helpers, likely close relatives, for the most important types of assistance but may rely on other types of helpers for less crucial assistance types. Lastly, it seems likely that migrants’ own characteristics influence the probability of social network assistance: I expect that migrants who face the least problems accessing support, for example, due to membership in a large migration network, or migrants who are least capable of securing non-network based means are most likely to rely on social network assistance. DATA To empirically test my hypotheses on social network assistance, I will rely on a small sample of Mexican migrants and non-migrants from my own field research and two secondary databases on Mexican migration. Conducting own interviews with Mexican migrants and non-migrants was very helpful in understanding the migration process and comprehensively quantify relevant types of assistance and helpers. The secondary databases were added to improve representativeness, even though they do not cover all assistance types and helpers.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Original Data Set I conducted field research in two small peasant communities (“ranchos”) and one small suburban community in the winter of 2005/2006. All three communities are situated in the central-western Mexican state of Michoacán - a traditional migration state (Bustamante et al. 1998) – and are between 10 and 60 driving minutes away from the state capital Morelia. The airline distance to the Mexico - California border is about 1200 miles. I selected the two ranchos because I had previously lived and worked in these communities, offering me background knowledge and local contacts to support the study. The third community was included upon a hint by an interviewee that migration to the U.S. would still be uncommon in this community, which is a local exception. Appendix A-I provides a brief description of the socio-economic background of the three communities.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Social Network Assistance in Mexico – U.S. Migration

21

In each community I drew representative sub-samples of the population. From a total of approximately 275 households in the first rancho, 300 households in the second rancho, and 75 households in the third community, I selected 21, 31 and 19 households, respectively. In each household I conducted a semi-structured interview with a male adult member, typically the household head (see Massey and Zeteno 2000).27 There were 29 non-migrants and 42 migrants of which I excluded one who had been a legal resident at the time of his first U.S. trip. I started each interview with questions about general socioeconomic characteristics and migration facts (if applicable), such as total number of U.S. trips, trip destinations and durations, legalization status, and reasons to migrate. For each migrant, I reconstructed the individual’s egocentric network at the time of his first labor migration trip to the U.S.: I first constructed a genealogical tree and a list of best friends at the time of his first U.S. trip and then inquired about the migration timing of all network members relative to his first trip.28 Then I made use of my migration milestones (see previous section) to assist the migrant’s often vivid recalling of his first U.S. trip, noting social network assistance at each step. Having previously reconstructed the migrant’s egocentric network at the time of his first U.S. trip proved to be helpful for the individual’s recalling of the obtained types of assistance and facilitated the interview, thus improving the accuracy of the data.29

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

27

I decided to focus on male migrants as female migration was very uncommon. 28 I followed Marsden’s (1990) recommendation to use name generators to elicit egocentric networks. Recalling kinship ties proved to be relatively easy for most interviewees. In order to elicit friendship ties I obtained interviewees’ class lists during primary or secondary school from the local school director or the education department’s state archive. Class lists were not available for the low-migration community. In addition, I further asked interviewees to list their best friends from sport clubs, bands, their neighborhood, or other. 29 Knowing the names of the migrant’s kin and friends made it much easier for me to associate names during the migrant’s narrative and detect possible omissions or errors. For example, some interviewees with strong family migration networks corrected their initial report of no assistance upon verification.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

22

Mexican Labor Migrants and U.S. Immigration Policies

Secondary Data Sets

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

The “Encuesta sobre Migración en la Frontera Norte de México” (EMIF) is a representative survey of the Mexican migration flow to the U.S. under the auspices of two Mexican government institutions and a Mexican university. The survey has monitored the flow of Mexican migrants at important travel points, including public bus terminals, train stations and airports in the seven most frequent Mexican border crossing cities since 1993.30 I pool all observations of Mexicans that state a desire to migrate to the U.S. to work without employment authorization between 1996 and 2003, giving a total of 13,766 observations.31 For a first sample, EMIF1, I further take out repeat migrants to get 9,380 observations of undocumented first-time labor migrants, representing a population of 1,000,047 migrants (once weighted). As first-time migrants arriving at the border have completed only part of the migration process, I construct a second sample, EMIF2, in which interviewees that arrive at the border on a repeat trip report on a previous undocumented labor migration trip to the U.S. EMIF2 is made up of 2,104 observations of undocumented repeat labor migrants that represent a population of 568,663 migrants. Furthermore, I use data from the Mexican Migration Project (MMP), which I present in detail in chapter 3. For the purpose of this chapter, I use information on household heads only, since this subset contains all information I need. In total, 18,083 household heads were surveyed in the MMP107, out of which 6,399 have migrated to the U.S. at some point. I restrict the sample to migrant household heads which have migrated to the U.S. to work without employment authorization since 1980; 2,768 observations remain.

30

According to EMIF’s methodological notes and a personal conversation with an EMIF statistician, the provided probability weights warrant that the weighted sample is representative for the entire Mexican migration flow to the U.S., except for border residents. 31 I start with 1996 as earlier waves used a slightly different methodology.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Social Network Assistance in Mexico – U.S. Migration

23

In the MMP most information on social network assistance refers to migrants’ last U.S. trip. Rather than excluding all migrants who have made more than one U.S. trip in their life time to obtain estimates on migrants’ first U.S. trip - this would shrink the sample by more than half - I decided to include the last migration trip of all undocumented labor migrants, first-time or repeat. Descriptive statistics on migrants’ characteristics and their social networks for the EMIF1, EMIF2, MMP and my own small sample are given in Table C.3 in Appendix C. SOCIAL NETWORK ASSISTANCE IN MEXICAN MIGRATION: EMPIRICAL FINDINGS Original Data Set

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Prevalence of Social Network Assistance: In my 41 interviews with undocumented Mexican first-time migrants I inquired about 12 different forms of social network assistance, grouped according to the major migration steps elicited above: information on migration, initial financing, domestic dislocation, border crossing, initial U.S. sustenance, and first U.S. employment (for more details, see Appendix A-I). Only two unlucky migrants reported that they had not received any help during the first U.S. trip; both had migrated at times when migration networks were still underdeveloped in their community. This implies that 95 percent of undocumented first-time migrants received some form of assistance by social networks. For 10 assistance types, 70 percent or more report social network help on their first undocumented U.S. trip (see Table 2.1).32 The two exceptions with low support both fall into to the area of domestic dislocation, which turned out to be a relatively simple challenge. According to my small sample, social network assistance during Mexicans’ first undocumented migration trip to the U.S. is thus very prevalent.

32

In calculating relative frequencies I took out unknown or non-applicable responses: Four migrants that could legally enter into the U.S. and two migrants that had a free guide were excluded from the smuggling related assistances; 13 migrants who stated that they crossed the border without overnight stay were excluded from the border lodging help; five border residents were excluded from all domestic dislocation assistances.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

24

Mexican Labor Migrants and U.S. Immigration Policies

Hierarchy of Assistance Types: As a first attempt to elicit the relative importance of different types of social network assistance during undocumented Mexicans’ first U.S. trip, I group them by frequency (see Table 2.1). Help with information on migration, initial financing, border crossing, and initial U.S. sustenance is most frequent (above 80 percent), support with first U.S. employment ranks in the middle, and assistance with domestic dislocation is least frequent. As a second criterion to judge the relative importance of each type of assistance, I asked migrants for which assistances they had known, with certainty prior to departure, who would help them (“pre-departure assurance”), assuming that high pre-departure assurance indicates migrants’ effort to ascertain help.33 If assisted, more than 80 percent of undocumented first-time migrants knew with certainty who would help them with border crossing and initial U.S. sustenance, but only 59 percent knew their job search helper prior to departure (see column 1 in Table 2.2). If I multiply the frequency of an assistance type with its pre-departure assurance to get a joint criterion - possibly the most meaningful one help with border crossing and initial U.S. sustenance dominate again (see column four). Lastly, I asked migrants and non-migrants to rank the types of assistance by importance (actual or hypothetical, depending on the case). 34 Help with initial U.S. sustenance was listed first, followed by border crossing or job assistance. Support with initial financing and domestic dislocation came last.35

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

33

It was impossible to elicit pre-departure assurance for initial information gathering, initial financing, and domestic travel means as these steps necessarily needed to be solved prior to departure. 34 Selecting the most important assistance type(s) proved difficult for some interviewees. On first thought many replied that all assistances would be important; 20 and 40 percent of migrants and non-migrants, respectively, failed to specify one or two single most important assistances. I took these out. Furthermore, there was a slight tendency towards common-sense ranking, favoring initial U.S. lodging and food assistances. 35 Initial information gathering and travel companionship were not included in the self-ranking. However, there were indications that initial information gathering was not a bottleneck. Only three non-migrants stated that they had no general migration knowledge and 75 percent or more could list all major migration steps.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

2 4 8 8 20 6 6 6 7 4 7 7

5% 10% 20% 20% 56% 14% 15% 18% 20.0% 10% 17% 17%

78%

85% 78%

85% 79% 74%

42% 68%

80%

95% 90% 78%

3% 18% 0% 3% 6% 5% 5%

15 35 35 27 26 35 32

5%

0%

32

32

0% 0% 2%

38 36 32

2

2 2

0 1 2

1 4

0

0 0 1

Relative and absolute frequency Networks Other

0

0 0

0 6 6

5 18

0

0 0 0

N/A

0

0 0

0 1 0

0 1

1

1 1 0

Unknown

Source: Author's interviews with 41 undocumented Mexican migrants (first U.S. trip). Notes: “N/A” indicates that assistance is not applicable (see footnote 32). Networks include kin, friends and acquaintances. “Other” includes local recruiters and smugglers. Relative frequencies exclude unknown and "N/A" responses. For further details see Appendix A-I.

(I) Information on migration Source of general migration knowledge Conversation about migration intentions Received invitation/ called U.S. contact (II) Initial finance Financial assistance in Mexico (III) Domestic dislocation Getting to border Lodging in border (IV) Border crossing Travel companionship Finding a smuggler Cash-On-Delivery/ U.S. pick-up (V) Initial U.S. sustenance Initial U.S. housing Initial U.S. food provision (VI) First U.S. job Obtaining first U.S. job

None/ Self

Table 2.1: Migration Process and Social Network Assistance (Original Data)

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

N/A 79% N/A 88% 85% 81% 82%

Initial financing Getting to border Lodging in border Travel companionship Finding a smuggler Cash-On-Delivery/ U.S. pick-up Initial U.S. housing Initial U.S. food Obtaining first U.S. job

II

III

IV

V

VI 6

2 5

2 4 8

10 9

4

1

Frequency

6

3 4

N/A 1 2

N/A 5

N/A

Predeparture assurance N/A

6

2 3

N/A 1 4

N/A 5

N/A

4

1 1

N/A 4 3

6 7

7

Ranking based on Frequency Self& ranking Assurance (migrants) N/A N/A

3

1 2

N/A 4 4

7 7

6

Selfranking (Stayers) N/A

5.2

1.8 2.8

N/A 2.8 4.2

7.7 6.6

5.7

N/A

Average rank

Source: Author's interviews with 41 undocumented Mexican migrants (first U.S. trip). Notes: Column 1 shows percentage of assisted migrants that knew, with certainty prior to departure, who would help them. Column 4 is ranking based on multiplication of relative frequency and pre-departure assurance. Column 8 is average over all shown rankings. 30 migrants and 15 non-migrants were asked to name the most important social network assistance for columns 5 and 6, respectively; multiple responses were possible. “N/A” indicates that help was not included. For further detail see Appendix A-I.

59%

N/A

Information on migration

I

Type of Assistance

Predeparture assurance N/A

Table 2.2: Relative Importance of Types of Social Network Assistance (Original Data)

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Social Network Assistance in Mexico – U.S. Migration

27

When I average across the different rankings, social network’s help with initial U.S. sustenance appears most important for an undocumented Mexican’s first U.S. trip, closely followed by border crossing help. Assistance with obtaining a first U.S. job and initial financing rank in the middle, and support with domestic dislocation is clearly the least important assistance of social networks during an undocumented Mexican’s first U.S. trip. Hence, the results confirm my hypothesized hierarchy of assistance types. Hierarchy of Helpers: In Table 2.3 I find confirmation for the generally agreed upon hierarchy of helpers within social networks (e.g., Massey et al. 1987): Undocumented Mexican first-time migrants are most likely to receive assistance from their nuclear family, followed by extended family, friends, and lastly others.36 Help from extended family members is almost as common as help from nuclear family members. Based on pre-departure assurance, assistance by kin, both nuclear and extended, is least risky. It is more than twice as certain as assistance by friends and other helpers. Since kinship help dominates initial U.S. sustenance and border crossing - which I identified to be the two most important assistances - this supports my hypothesis that undocumented first-time migrants will try to recruit their most trusted helpers for their most crucial migration steps. Support by friends is important. Friends provide one-fourth of all assistances in my sample. Friendship networks further seem to specialize. It appears that the share of friendship help is highest if it is of least disadvantage or even necessary that help-giver and helpreceiver are at similar stages in their life time migration cycle - which friends likely are – or if a joint solution to the problem makes sense. This applies to talking about migration intentions, traveling together, and searching jointly for a smuggler and job. Typically undocumented Mexican first-time migrants had one main helper, for example, the person who invited them, but also spread assistances over more than one helper: Only 30 percent relied for more than half of their assistance types on the same helper - commonly their father or brother - but about 60 percent drew on one helper for more than three assistance types.

36

“Other” includes acquaintances, paisanos, local smugglers and local labor recruiters.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Initial financing

II

34% 31%

23% 29%

34% 29%

29% 31% 36%

24% 33%

21%

26% 25%

21% 17%

47% 31% 18%

41% 17%

9%

17% 15%

16% 20%

1% 14% 11%

12% 44%

6%

75% 91%

91% 92%

N/A 100% 90%

N/A 100%

N/A

88% 90%

92% 80%

N/A 89% 89%

N/A 100%

N/A

22% 43%

50% 17%

N/A 64% 75%

N/A 33%

N/A

50% 43%

67% 13%

N/A N/A N/A

N/A N/A

N/A

Pre-departure assurance (B) Nuclear Extend. Friends Other family family N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A

Source: Author's computation based on 42 migrant interviews. Notes: Panel A shows relative frequency by assistance type and helper. Panel B shows percentage of cases in which migrants were certain about aid receipt and specific helper upon leaving. If no assistance was received or assistance was non-applicable or unknown the observation was dropped (applies to both panels).

Obtaining first U.S. job Average

29% 34%

Initial U.S. housing Initial U.S. food

V

VI

21% 24% 36%

Travel companionship Finding a smuggler Cash-On-Delivery/ U.S. pick-up

24% 6%

65%

IV

Getting to border III Lodging in border

General migration knowledge Talk about migration intentions Received invitation/ called contact

I

Type of Assistance

Frequency of assistance (A) Nuclear Extend. Friends Other family family 28% 24% 18% 30% 39% 24% 33% 4% 32% 35% 26% 6%

Table 2.3: Frequency and Assurance of Network Assistance by Network Type

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Social Network Assistance in Mexico – U.S. Migration

29

Secondary Data Sets

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Prevalence of Social Network Assistance: The two secondary databases confirm that overall social network assistance to undocumented Mexican migrants is very common (see Table 2.4). According to the MMP, 95 percent of undocumented Mexican migrants received some help during their last U.S. trip. EMIF reports that “only” 75 percent received some help on their last U.S. trip, mainly because migrants seem to have fewer U.S. contacts - 22 percent report no U.S. contact; virtually all U.S.-connected migrants received some assistance.37 For the most important migration steps, seven or more out of 10 undocumented Mexican migrants were aided: 71 and 86 percent received initial U.S. lodging assistance, and 71 and 69 percent received travel companionship according to the EMIF and MMP, respectively. Hierarchy of Assistance Types: Regrettably, for the secondary databases, only observed frequencies are available as criterion to judge the relative importance of different assistance types (see Table 2.4). Based on observed frequencies, initial U.S. lodging help ranks first in both databases. For the MMP, job assistance (73 percent), travel companionship (69 percent), and smuggler payment (42 percent) follow. The EMIF ranks travel companionship second (71 percent), followed by job assistance (59 percent), initial financing for domestic expenses (58 percent), and border city lodging (44 percent). Hence, except for the low assistance level with smuggler payment, the secondary databases confirm my hypothesized ranking of assistance types. 38

37

The MMP somewhat oversamples traditional migration regions, in which migration networks should be stronger. 38 I cannot explain the low MMP estimate for help with smuggler payment. If I restrict the sample to first trips or more recent years this does not increase the estimate markedly.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

30

Mexican Labor Migrants and U.S. Immigration Policies

Hierarchy of Helpers: The MMP contains a breakdown of social network assistance by type of helper (see Table C.3 in Appendix C). Help by kin is most frequent: 55 percent of assistance to undocumented Mexican first-time migrants is provided by kin, 35 percent by friends, 2 percent by other community members, and 8 percent by employers and others. Friends appear to specialize on assisting with travel companionship (59 percent) and job search (41 percent). Table 2.4: Frequency of Social Network Assistance by Assistance Type (Secondary Databases) Assistance Type (II) Finance Financing domestic trip (III) Domestic dislocation Lodging in border (IV) Border crossing (+) Travel companionship Paying smuggler (V) Initial U.S. sustenance Initial U.S. lodging (*) (VI) U.S. job Obtaining first U.S. job Any Assistance (#)

EMIF Yes No

MMP (†) Yes No

58%

42%

N/A

N/A

44%

56%

N/A

N/A

71% N/A

29% N/A

69% 42%

31% 58%

71%

29%

86%

14%

59%

41%

73%

27%

75%

25%

95%

5%

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Source: Encuesta sobre Migración en la Frontera Norte de México (I and II uses EMIF1, else EMIF2), Mexican Migration Program (see text for details) Notes: (†) unknown observations are taken out; "No" includes non-network support (employer and other); (+) only migrants that used smuggler for MMP (IV); (*) lodging and food assistance for EMIF (#) any assistance in US for EMIF

Social Network Assistance and Beneficiaries’ Characteristics: It seems likely that reliance on social network assistance depends on the beneficiary’s own characteristics: Migrants who have an elevated need for a certain assistance type or are able to obtain it at a lower cost, compared to others, should be more likely to draw on this help. In Appendix B-III) I present a simple economic model of migrants’ receipt of social network assistance and show that a probit model may be used to estimate it. Hence, I estimate four independent probit

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Social Network Assistance in Mexico – U.S. Migration

31

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

regressions with data from the MMP, one each for assistance with travel companionship, smuggler payment, initial U.S. lodging, and finding U.S. employment. All regressions include documented and undocumented Mexican migrants and control for various individual characteristics and time-invariant community effects. Table 2.5 reports the parameter estimates and significance levels of the four probit regressions. Consistently, I find that migrants whose nuclear family members, extended family members, or friends have migrated - an indication of migrants’ ease to access network support - are more likely to receive assistance.39 Nuclear migration networks are the most significant determinant of migrants’ receipt of social network help. In most cases, extended migration networks only become statistically significant if nuclear migration networks are excluded (results not shown), consistent with the explanation that extended and nuclear family networks are substitutes. Friendship and kinship migration networks seem to be complements. Furthermore, I find confirmation that migrants’ need to receive help is important: Migrants who do not work in the U.S. or are undocumented are significantly more likely to report social network assistance. In this respect, Massey and Espinosa’s (1997) argument that migrants’ human and social capital are substitutes is also relevant. I find some support for their hypothesis. The parameter estimate of migrants’ number of prior U.S. trips, a measure of migration-related human capital, is negative but has mixed statistical significance. Educational measures perform very poorly.

39

It should be noted that I avoid Manski’s (1993) reflection problem in a “technical,” thus not fully satisfying, way: the probit model is nonlinear in parameters, but Manski’s critique applies to linear models (see Brock and Durlauf 2001).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Travel Company 0.003 0.176* 0.143* 0.393* N/A -0.022*** -0.162** 0.077 -0.297 -0.254 0.101 0.069 N/A 0.009*** 4,179

Coefficient Smuggler Initial U.S. Payment Lodging 0.524*** 0.392*** -0.017 0.132 0.355* 0.119 0.908** 0.498** N/A 0.527*** -0.010 -0.008* 0.230* 0.076 0.052 0.136 0.153 0.174 -0.367 -0.107 0.940** 0.213 0.603* -0.086 N/A -0.016** 0.022*** 0.037*** 1,225 5,525

U.S. Employment 0.191*** 0.040 0.213*** N/A 0.374*** -0.010*** -0.134 -0.022 -0.074 -0.032 0.264* -0.012 -0.010 0.024*** 5,135

Source: Mexican Migration Project (documented and undocumented migrants, all years). Notes: “***,” “**,” and “*” significant at the 1%, 5% and 10 % level, respectively. Standard errors (not shown) are robust to clustering at year level. Columns 1 and 2 refer to first U.S. trip. Column 2 excludes migrants that did not use a smuggler. Columns 3 and 4 refer to last U.S. trip. All regressions include dummies for community of origin.

Dummy if any nuclear family members have migrated Dummy if any extended family members have migrated Dummy if any friends have migrated Dummy if migrant does not work (unemployed or not searching) Dummy if migrant is undocumented (entered illegally into U.S.) Age at point of U.S. trip Dummy if individual has completed 6-8 years of education Dummy if individual has completed 9-11 years of education Dummy if individual has completed 12-14 years of education Dummy if individual has completed >14 years of education Dummy if individual is female Dummy if any land was owned prior to migration trip Number of U.S. trips Year of U.S. trip Observations

Variable

Table 2.5: Explaining the Receipt of Various Network Assistance Types (MMP)

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Social Network Assistance in Mexico – U.S. Migration

33

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Comparison of Original Data Set, Secondary Data Sets, and Other Studies The findings of my small original sample and the two large secondary samples are consistent in that social network assistance to undocumented Mexican migrants is very frequent overall. The MMP and my original sample show that, in traditional Mexican migration regions with strong migration networks, almost every undocumented migrant receives some form of assistance (95 percent). Even the more conservative estimate from the EMIF states that three out of four undocumented Mexican migrants receive some help while in the U.S. For most assistance types, my original sample reports 10 to 15 percent higher assistance frequencies than the EMIF, with the MMP estimates lying somewhere in-between. There is also consistency in terms of ranking types of assistance. Help with border crossing and initial U.S. sustenance is most important, assistance with initial financing and U.S. employment search are of intermediate importance, and support with domestic dislocation is least important. Notably, social network assistance with the most important migration steps is very high (70 percent or more). Both MMP and my own small sample show that help by kin is most common. Their share is about 50 to 60 percent. Yet, they also agree that friends’ share of total assistance is significant (25 to 29 percent) and that in friends’ areas of specialization - information exchange, travel companionship, and job search – their share might reach 50 percent or more. In comparison to other studies, my estimated assistance levels prove to be upper bound estimates but are not unreasonable (see Table C.2 in Appendix C).40 Besides the fact that other studies use small samples with different regional focuses, observation periods (mostly earlier), migrant types (generally including documented migrants) and wording of questions, there may also be another explanation for their lower assistance estimates. None of the other studies focused on quantification of social network assistance, except for Espinosa (1997), which could imply that the authors overlooked some instances of social network assistance. My own interviewing experience is that the 40

The striking differences in border crossing assistance compared to Espinosa (1997) are likely due to her inclusion of legal migrants and other non-smuggler users.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

34

Mexican Labor Migrants and U.S. Immigration Policies

reconstruction of migrants’ ego-centric migration networks at the time of their U.S. trip and the usage of migration milestones aids migrants’ recalling and interviewers’ recording of social network assistances considerably.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

THE UNRIVALLED IMPORTANCE OF SOCIAL NETWORKS IN MEXICAN MIGRATION This chapter has tried to assess whether, and if so, why, social networks play an important role in supporting the annual labor migration of several hundred thousand undocumented Mexicans to the U.S. While the interdisciplinary migration literature abundantly acknowledges the importance of social networks in undocumented Mexican migration, it has not comprehensively quantified social network assistance nor explained why social networks are such an important facilitator of migration compared to other migration conduits. Based on my reading of other Mexican migration studies and my own interviews with Mexican migrants, I structure the Mexican migration process into six major steps: information gathering for decision making, initial financing, domestic dislocation, border crossing, initial U.S. sustenance, and first U.S. employment. A microeconomic analysis of the challenges faced by an undocumented first-time migrant at each of these steps indicates aggravated problems of information asymmetries between migrants and lenders, smugglers, and employers. This makes trust crucial, disabling marked-based exchanges but empowering social network assistance. Hence, social network support to undocumented Mexican migrants should be very frequent, particularly for border crossing and initial U.S. sustenance; migrants should try to recruit their most trusted helpers for these crucial steps. Descriptive analysis of my own sample and two large secondary databases on Mexican migration supports these hypotheses. Between 75 and 95 percent of undocumented Mexican migrants received social network assistance. For the most important migration steps, 69 to 86 percent were helped. In disagreement with the economic literature on social interactions (e.g., Behrmann et al. 2002; Miguel and Kremer 2003) and previous migration studies (Banerjee 1984; Espinosa 1997; Munshi 2003), I do not find that migration networks’ most important help is the provision of information. Providing employment information proved to be only of intermediate importance, but

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Social Network Assistance in Mexico – U.S. Migration

35

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

“concrete” help in crossing the border and surviving initially in the U.S. was paramount. Migrants used their most trusted helpers for these crucial types of assistance. I find support for the commonly assumed hierarchy of helpers: nuclear family, extended family, friends, and other (e.g., Lomnitz 1977). However, nuclear and extended family networks appear to be good substitutes. Agreeing with Espinosa (1997), friends should not be ignored, as is often done. They provided a quarter of all support and specialized on assisting with information exchange, travel companionship, and smuggler and job search. However, friends’ assistance was regarded as less than half as certain as help by kin. Policy makers should be aware that Mexican migration networks have been very effective (providing the most needed help), resilient (spreading help across various shoulders), and innovative (as the evolving cash on delivery contract shows). Popular U.S. migration management techniques that try to make migrants’ U.S. lives harder in order to induce them to stay in Mexico or ‘self-deport,’ as former Californian governor Pete Wilson opined (cited in Cornelius 2005), tend to underestimate “the principal strength that migrants brought to the 'battle at the border' [and beyond], their tradition of mutual aid” (Spener 2009, p. 199). The upshoot is not to try to uproot Mexican migration networks - a task that would be extremely difficult and disruptive given their extent and density - but rather to make use of these networks. The strong embeddedness of Mexican migrants into home-town based networks indicates that they are still strongly attached to their communities of origin. Recognizing that discouragement of undocumented migration needs to take place at the origin, the U.S. and Mexican governments should consider social networks as a conduit to funnel resources to migrant-sending regions and should support migrants in leveraging these resources to productive uses. This could help to provide Mexicans with a more attractive option to stay at home (see final chapter).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved. Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

CHAPTER 3

Emigrant or Sojourner? Migration Intensity and Its Determinants

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

INTRODUCTION When thinking about socio-economic development, policy makers have often considered out-migration as an indication of failed policies. Migration has rarely been regarded as having the potential to contribute to the development of the sending country. Indeed, its relationship to development was sometimes depicted as harmful, creating “dependency” but no growth impulses (e.g., Reichert 1981). However, researchers, international organizations, and governments have recently (re)discovered the potential of migration to spur domestic development. The potential of remittances, knowledge transfers, and “tapping the diaspora” prompted the former United Nations Secretary General to initiate a “High-Level Dialogue on Migration” in late 2006. Yet since economic theories of migration often ignore important aspects and interdependencies of migration behavior, they offer little guidance as to the conditions under which migration is most beneficial or regarding which types of migration are preferable. A great share of the literature on international migration has restricted itself to one specific aspect of migration behavior: the choice of whether or not to migrate. I term this “extensive migration behavior.” What this reduction of migration behavior to one choice misses is that a migrant also chooses a level of “migration intensity,” or how strongly to reorient himself to the destination and relinquish economic and social ties to his place of origin. In an attempt to fill this gap, this chapter develops the concept of migration intensity, defined as

37

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

38

Mexican Labor Migrants and U.S. Immigration Policies

the degree to which a migrant shifts his attachment, association and engagement from his place of origin to the migration destination. The chapter is structured as follows. The next section reviews the inter-disciplinary literature on international migration behavior to motivate the concept of migration intensity. The third section develops the concept of migration intensity, and the fourth section proposes a general technique to measure it. In the fifth section, this is applied to the Mexican-U.S. case, and the determinants of variations in migration intensity are examined. The final section provides a summary and conclusions. CONNECTING THE INTERDISCIPLINARY LITERATURE

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Migration, Return and Remittance Decision International migration is a complex, multi-faceted phenomenon. As a result, researchers have commonly divided the analysis of international migration behavior into various broad questions, mostly according to disciplines, and have then considered each separately, generally ignoring possible interconnections. Much migration research has focused on the question of what determines an agent’s decision to migrate or not. While migration theories differ with regard to motive be it migration-cost deducted wage differentials (Sjaastad 1962) or joint household strategies in response to risk and credit market failures (Stark and Bloom 1985) - they typically restrict their focus to the “stayversus-migrate” decision. Many empirical analysts, particularly proponents of regression analysis, have shared this focus on what I shall refer to as “extensive migration behavior.” Extensive migration behavior captures only one policy-relevant dimension of migration behavior. It misses other important economic and social decisions that migrants face: which migration strategy to choose, such as “emigration” or “seasonal migration;” whether to remit money or not and if so, how much; whether to invest at home or abroad. Most fundamentally, migrants must decide whether and to what extent they will maintain economic, social, and cultural ties to their home communities.41 Some of these questions have been addressed by distinct parts of the migration literature. Economists have explained return migration as 41

Throughout this book “home” refers to the community of origin.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Emigrant or Sojourner? Migration Intensity and Its Determinants

39

resulting from either failed migration aspirations or the fulfillment of a pre-set saving target (e.g., Djajic and Milbourne 1988). Remittances, a focal point for economists, have been attributed to altruism, selfishness, or reciprocal exchanges (see Rapoport and Docquier 2006). Others have analyzed migrants’ “localized investment” choices, such as small-scale enterprises (Woodruff and Zeteno 2007), housing (Mooney 2004), or social networks (Lomnitz 1977; Massey et al. 1987). Treating return behavior, remittance behavior, and localized investment behavior independently of each other ignores the connections among these decisions. They are related in that each decision requires the migrant to make a choice how strongly he wants to “be invested” at home or abroad.42 Hence, it would be sensible for him to consider them jointly. Migration Types and Assimilation

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Anthropologists and sociologists have long observed that migrants have diverse motives, characteristics, and strategies. They have tried to summarize these differences by postulating certain migration types.43 Temporary and permanent migrants are commonly separated. Many scholars further distinguish target earners (or “return migrants”) and recurrent migrants (sometimes also “seasonal” or “circular” migrants). Recently, some have proposed an additional category, “transmigrants,” to describe individuals who engage simultaneously in origin and destination societies (Schiller et al. 1992). While most anthropologists and sociologists agree that different migration types exist, no unified typology has emerged and existing classifications are often defined in an ad hoc fashion. One difficulty is that migrants cannot be neatly classified into distinct types. There is a multitude of possible criteria and boundaries are arbitrary; for example, when does “temporary” become “permanent?” For these reasons, continuous and multi-indicator scales would be preferable. Moreover, it would be desirable that the scales have a clear foundation in a related theory. 42

Lucas (2005, p. 9) remarks, “certainly return migration and the propensity to remit are linked.” In Chapter 4 I develop a mathematical model that shows how these decisions might be linked in a joint decision framework. 43 I strictly focus on international labor migration, excluding other types of migration.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

40

Mexican Labor Migrants and U.S. Immigration Policies

The sociological literature on assimilation, acculturation, and incorporation focuses on how well the migrant adapts to his new host environment. The orthodox theory, due to Gordon (1964) and others, describes a staged process by which the migrant gradually incorporates into the destination society. Pluralists, transnationalists, and others have criticized the unilinear and automatic fashion by which migrants are assumed to assimilate, pointing to the fact that migrants continue to rely on ethnic ties. Socio-economic and medical paradoxes of reverse assimilation are mentioned.44 The determinants of adaptation and integration remain largely unknown; various individual and structural factors have been proposed, but empirical studies point to strong historical specificity. Economists have mirrored orthodox assimilation theory in studies of educational and employment catch-up, mainly focusing on the closure of the nativeimmigrant wage gap. Recent evidence suggests that inter-generational catch-up occurs, but that significant ethnic wage gaps can persist across generations (Borjas 2006). While the concept of assimilation is related to migrants’ socioeconomic reorientation, the literature has focused exclusively on the destination country. Implications for the sending country are largely unexplored and would be uncertain. For example, the counterpart of assimilationists’ gradual integration could be gradual detachment from the home community.45 Yet, transnationalists would counter that even apparently settled migrants remain incorporated at home (Schiller et al. 1992; Portes et al. 2005).46 Furthermore, questions such as language acculturation or the closure of the wage gap are of secondary importance to the development economist who is interested in the welfare effects of migration in the country of origin.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

44

Portes and Rumbaut (1996) have argued that assimilation depends on the “modes of incorporation” which an ethnic group faces, resulting in “segmented assimilation.” 45 Indeed Massey et al. (1987, p. 284) argue that “progressive integration into U.S. society brings a gradual shift of orientation away from migrants’ communities of origin to the Unites States.” 46 Waldinger and Fitzgerald (2004) question the theory of transnationalism as such, contending that the implied generalizations as to a condition of being ("transnationalism") or a class of people ("transmigrant") cannot be made. See also the response from Glick Schiller and Levitt (2006).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Emigrant or Sojourner? Migration Intensity and Its Determinants

41

Welfare Migration can have profound effects on sending countries. Outmigration constitutes a resource loss to the sending society, not only in labor power but also in financial, social, and political capital. On the other hand, migrants often generate various positive resource transfers, such as money remittances. The net welfare effect of migration on sending communities has been debated, but recent research suggests that the overall effect often appears to be positive.47 The question generally left unanswered is what determines the sign and size of the cost-benefit calculation? The few studies that have addressed this question have emphasized dependence on local conditions, such as investment opportunities (Taylor 1999; Durand et al. 1996a; Lindstrom 1996). They miss the possibility that migrants’ migration strategies and the degree of their continuing attachment to the community of origin could also strongly impact their willingness to engage in positive transfers. As Mines and Massey (1985, pp. 115-116), comparing two similar communities, remark: “The standard of living in Guadalupe is far above that in Los Animas. Larger migrant contributions have allowed village leaders to make more improvements in roads, schools, churches, water systems and sewerage than in Las Animas. [...] These differences in levels of living relate directly to the different kinds of migrancy that predominate in each community. The legal shuttle migrants prevalent in Guadalupe

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

47

Taylor et al. (2005) report that remittances have contributed to poverty reduction in Mexico. Multiplier effects (Adelman et al. 1988), capitalization of micro-enterprises (Woodruff and Zenteno 2007), and “collective remittances” – group-organized transfers that benefit community projects at home (Goldring 2004) - can further enhance migrants’ positive impacts. The World Bank (2006b), using a global general equilibrium model, suggests that the benefits of free labor movement would greatly surpass the benefits of freer trade. On the other hand, some researchers have blamed migration for labor shortages, “dependency,” conspicuous consumption, and inflated prices (Wiest 1984). Whether migration worsens inequality and how parental migration affects children’s educational attainment and health are still debated. The World Bank (2006e) surveys several welfare consequences of migration.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

42

Mexican Labor Migrants and U.S. Immigration Policies still view themselves as town residents [..] they see migration as a ‘temporary’ necessity intending to return to the home community [..] therefore investing earnings in ways destined to improve living standards in their home community. They improve their houses […] and are willing to organize and contribute toward town projects […] On the other hand, the beginner and permanent settlers that typify Las Animas are less attached to the home community. They stay in their U.S. schools and are often U.S. home owners. Their life is ‘in the north’ [.] Their money is spent on furnishing houses in California, leaving little left over to improve living conditions in the home village.” [Emphasis added.]

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Lucas (2005), summarizing the migration literature, notes that male, temporary, undocumented, non-family accompanied migrants without tertiary education remit most (see also Amuedo-Dorantes et al. 2005; Dustmann and Mestres 2010; Durand et al. 1996b; Hagan et al. 2008; Mooney 2004; Reyes 2004). These patterns suggest that migrants with a stronger attachment to their community of origin may be more willing to engage in positive financial transfers. At the same time, greater attachment may mitigate the loss in human, financial and political capital that migration imposes on sending communities. Destination countries often prefer (regularized) circular migration, as exemplified in a recent surge of interest in policies to promote this type of migration (e.g., IOM 2006b; MPI 2006; Portes 2008; World Bank 2007). Circular migrants constitute a flexible labor supply and may be less likely to make use of public health, education, and welfare programs. On the other hand, some immigration policies aim to promote the fuller integration of immigrants. In either case, an index that captures the “intensity of migration” would be useful for policy analysis in both sending countries and destination countries.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Emigrant or Sojourner? Migration Intensity and Its Determinants

43

THE CONCEPT OF MIGRATION INTENSITY The concept of migration intensity refers to the degree to which a migrant shifts his attachment, association and engagement from his place of origin to the migration destination. Migration intensity is reflected in a migrant’s decisions regarding socio-economic reorientation – choice of migration pattern, remittance behavior, and localized investment behavior – choices that are likely to be highly complementary. 48 Migration intensity succinctly summarizes these decisions, making it a variable of considerable policy relevance. Extensive migration behavior is clearly related to intensive migration behavior. However, the concepts are distinct. While extensive migration behavior refers to whether a non-migrant decides to become a migrant, migration intensity refers to how completely those who do migrate choose to disconnect themselves from their origin. Some individuals choose to migrate at the lowest possible migration intensity - keeping all their assets, social ties, and most leisure consumption at home. Others decide to emigrate –cutting most or all of their home-town relationships. Migration intensity is related to assimilation, but again distinct. Assimilation compares immigrants with natives, whereas migration intensity compares migrants with each other. The concept of migration intensity assumes that socio-economic reorientation choices are complementary. Remittance behavior, choice of migration pattern, and localized investment behavior are likely to be correlated as they share a joint dependence on location.49 Sending

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

48

This concept differs from the Consejo Nacional de Población's (CONAPO) use of the term “migration intensity” (intensidad migratoria). CONAPO (2000) uses different indicators to describe migration prevalence by location, which I consider a measure of extensive, not intensive, migration behavior. 49 Carefully controlling for various disturbing factors, such as household and time fixed effects or reversed causality, Dustmann and Mestres (2010) consistently find that the intention to return home and the location of spouse and children significantly affect the remittance behavior of immigrants in Germany. Changes in these variables, indicating changes in home attachment, reduce the amount remitted strongly. Similarly, DeSipio (2000) finds evidence that migrants, who are "more acculturated", as indicated by the relative amount of time a migrant spends in the U.S. or family reunification in the U.S., are less likely to remit (see also Grieco 2003).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

44

Mexican Labor Migrants and U.S. Immigration Policies

remittances makes most sense if a migrant remains attached to his place of origin and plans to return. For example, sending remittances to hire somebody to tend to one’s assets or to further one’s odds to inherit (see de Laat 2005 and Lucas and Stark 1985, respectively) will make sense only if the migrant intends to return. At the same time, remitting is likely to increase the value of returning home (see also Rapoport and Docquier 2006 and Chapter 4). Similarly, since most investment choices for migrants regarding physical, human and social capital are tied to a choice of location, one can expect the choice of where to invest to be correlated with the return and remittance behavior of a migrant. MEASURING MIGRATION INTENSITY: A GENERAL FRAMEWORK Although migration intensity is not directly observable, choices that reflect migration intensity can be observed and these can be used to create an Index of Migration Intensity (IMI). In doing so, one faces two methodological issues: which indicators to select50 and how to aggregate them into an index. Indicators of Migration Intensity

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

I have defined migration intensity as the degree to which a migrant shifts his attachment, association and engagement from his place of origin to the migration destination. Hence, any observable choice that indicates either an important association to the origin or the destination is relevant. This includes both economic and social variables, since for migrants decisions about their economic life (e.g., employment) are inseparably bound to their social life (e.g., living together with their family or not). In general, purchases of localized consumption goods and investments in localized assets can be taken as valuable signals. “Localized” here indicates that the value or utility of the good depends on the location in an important way. 50

Since the identification of the set of indicators will circumscribe the latent phenomenon, it is crucial to select a balanced set that neither favors only one façade of migration intensity (e.g., remittances or language) nor is congruent to measures of related concepts (e.g., extensive migration behavior or assimilation).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Emigrant or Sojourner? Migration Intensity and Its Determinants

45

A good example of localized consumption is “home-time consumption.” This refers to migrants’ decisions to return home to spend time together with family and friends or just to be in their homeland. The proportion of time (e.g., days per year) that a migrant spends at home is one indicator of home-time consumption. Return frequency is a second indicator of home-time consumption since it approximates the expenditure a migrant is willing to afford in order to spend time at home. Monetary or other transfers sent home constitute another signal of a spatial tie. Total remittances are not a good indicator since these are affected by other factors such as wealth. The proportion of income remitted is a preferable indicator (in the absence of income data, another option is to use a dummy variable indicating if any money was transferred). Communication with the community of origin could be another signal for migration intensity, particularly if communication is expensive. Finally, evidence that a migrant has changed his consumption habits as to strongly value goods which are only available at the destination could be taken as signs of elevated migration intensity. I consider investments as local if they can only be transferred at a considerable loss in value, e.g., because markets are absent or weak. The higher the loss in value, the stronger is the signal of spatial preference. The set of localized investments should include not only local physical assets but also local human and local social capital goods. Often migrants possess the same type of good at home and abroad. Therefore it is best to use indicators of proportional ownership: the proportion of item X owned at the destination over total ownership of item X. A higher score implies higher migration intensity. For localized physical assets, business ownership is one useful indicator, since migrants typically work in their enterprises and cannot easily transfer their full value abroad. House ownership can also be used, particularly if the housing market is weak or transaction costs are high. Again, it is best to use proportional rather than absolute indicators to attenuate wealth effects. Localized human capital goods include skills that are costly and valued differently at the origin and destination. A good example is immigration documentation, worthless in the home labor market but highly valuable abroad. The ability to speak the destination language is another good indicator if this knowledge is uncommon at home and not valued in the home labor market. If there are indications that work

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

46

Mexican Labor Migrants and U.S. Immigration Policies

experience is localized this may be used. Human capital indicators sometimes assumed to signal low migration intensity, like “ethnic skills” or “ethnic education” (Chiswick 2009), are less likely to be useful since migrants often continue to depend on these skills when abroad or their acquisition is too customary to signal an intentional tie. Localized social capital may be approximated by the location of social networks. This can provide useful information about migration intensity under either of two conditions: first, if the migrant actively invested in gaining membership to the network, e.g., by marrying a native or joining a social or professional organization, or second, if he influenced the (re)localization of the network itself by influencing the decisions of other members as to whether to stay or to leave. The location of the spouse and minor children is a good indicator for household heads in many societies.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Aggregation of the Migration Intensity Indicators: Principal Components Analysis When several indicators are combined to form an index, linear aggregation is common, and this is the procedure that will be used here. This requires that one chooses an appropriate set of weights. Theory does not suggest specific weights and choosing some weight structure, such as equal weights, without empirical support would require strong assumptions. Instead I use a statistical technique for aggregation, principal components analysis.51 Principal components analysis (PCA) is part of a set of multivariate statistical techniques that are referred to as factor analysis techniques. One of PCA’s uses is dimensional reduction. The idea is that several observed variables are somehow interdependent or correlated, such that most of their information may be expressed by a set of fewer composite variables, called components. Each component is a linear combination of the original variables. PCA’s main advantage lies in the fact that it is the most efficient technique to accrue as much as possible of the original information to as few as possible components. Consecutive components carry less and less additional information, such that latter components might be dropped at little cost.

51

A more formal description of principal components analysis is presented in Appendix B-I, drawing on Basilevsky (1994). See also Kline (1994).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Emigrant or Sojourner? Migration Intensity and Its Determinants

47

Technically PCA computes the eigenvalues and eigenvectors of the observed variables’ covariance matrix, which are then used to compute the “factor score coefficients” (or “weights”). The solution is unique for a given data matrix. To obtain an individual’s index score (commonly referred to as “factor score”), his observed indicator values are “weighted” by their corresponding factor score coefficients and summed. Depending on how many principal components one decides to retain, the Index of Migration Intensity (IMI) will have one or more dimensions.52 MIGRATION INTENSITY: THE U.S. - MEXICAN CASE Data Source This chapter uses data from the Mexican Migration Project (MMP), a publicly accessible, collaborative research project based at Princeton University and the University of Guadalajara. Since 1982, the MMP has surveyed migrants and non-migrants in Mexican communities during the months of November to February to take advantage of the regular return visits of migrants. Each community was visited once, and about five communities were surveyed in a given year.53 The MMP database for the years 1982 to 2006 includes data from a total of 114 communities, in each of which roughly 170 households were randomly chosen and interviewed. This gives a total of 19,003

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

52

Deciding on how many components to include is a tricky decision and authors have combined statistical techniques (see Basilevsky 1994) as well as “rule of thumbs;” for example, to retain only components whose eigenvalue is at least one. Furthermore, authors commonly choose to construct the components or factors in such a way as to achieve maximal interpretability. Therefore, components are often rotated. However for data reduction, which is my interest, rotation seems less common (see Kline 1994). 53 The MMP includes cities as well as towns and ranchos. At the beginning the MMP personnel delimit a “survey site,” which, in the case of ranchos, normally correspond to the whole rancho, in the case of a city, refers to a neighborhood or a part of the city. Within the survey site all dwellings are then enumerated in order to create a sampling frame from which households are randomly selected, and sampling weights are then created. All of my analyses in Chapter 3 will incorporate these weights.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

48

Mexican Labor Migrants and U.S. Immigration Policies

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

surveyed households. I use the information on household heads, since only this subset contains all necessary information. As only a very small fraction of the household heads are female (4.7 percent), I restrict the analysis to males. There are 6,389 households that are headed by a current or former male migrant and 4,315 of them have complete data. Since communities are selected non-randomly, the MMP is strictly representative only for the 114 surveyed communities. Fortunately, the selected communities reflect a broad range of sizes and socio-economic conditions. According to Massey and Durand (2005), comparison of the MMP to a nationally representative survey has shown great accordance. The geographic focus, favoring western Mexican states (southern and northern states have been added recently), seems to be the only major bias. Representativeness within the communities themselves is often difficult to achieve in communities with significant out-migration. The MMP deals with this problem in several ways. First, surveying in Mexico is executed in the traditional months of return. Second, spouses frequently remain in Mexico and are interviewed in the absence of the household head. Third, emigrated households are tracked by “snowball” sampling techniques and interviews are conducted in the U.S. A total of 972 households were interviewed in the U.S., along with 5,778 interviews of migrant-headed households in Mexico. Despite these problems, no other data set comes as close as the MMP to obtaining a representative image of Mexico’s former migrants, current migrants, and emigrants. I first construct a cross-sectional sample in which I include the most recent observation of each migrant household head (i.e., the survey year). I then assemble a panel data set using individuals’ retrospective account from age 18 to the survey year.54 Due to the panel’s retrospective nature and the fact that different communities were interviewed at different points in time, the age of the interviewee at the survey point and the community to which he belongs determine the calendar years for which his life-time history is available. Hence, it would not be appropriate to use calendar years as fixed points around

54

Recall bias is therefore an issue. However, Smith and Thomas (1997) report that recall bias is least problematic for salient events, such as important migration moves. Moreover, the results would be biased only if recall errors were systematic, for which I have no indications.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Emigrant or Sojourner? Migration Intensity and Its Determinants

49

which to define the panel. Instead each individual is sampled in five year time-steps, ending with the survey year.55 Mexican Migration Intensity: Indicators The following indicators are used to construct the Index of Migration Intensity: 56 Home-Time Consumption: I construct a continuous indicator of the proportion of total time that Mexican migrant household heads spend abroad during a five-year interval surrounding the observation year.57 Remittances: Using MMP data on average monthly remittances during the last U.S. trip, I construct a continuous variable of the proportion of income remitted. Localized Physical Capital: My measure of localized physical assets is an index of U.S. home ownership. Unfortunately, while information on home ownership is broadly available, only one-third of the sample contains information as to the location of the house. However, the location of the interview, U.S. or Mexico, is very accurate in predicting the location of the house: for 99 percent of all home owners, the location of the house is correctly predicted in the sub-sample for which the information is available. Hence, I combine ownership data with interview location. Non-owners were assigned a “zero,” clearly distinguishing them from Mexico owners (-1) or U.S owners (1), these latter two “being invested” as opposed to former. Localized Social Capital: I construct a continuous variable of the proportion of the non-head household members (spouse and children) who have migrated during the last three years. Wives or children born in the U.S. are considered equivalent to those who have immigrated as an indicator of investment in U.S. social capital. 58

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

55

I choose period lengths of five years since several of the indicators vary only infrequently over time. The results are not sensitive to different period lengths. 56 Except for remittances, higher indicator scores relate to higher levels of migration intensity. 57 The results are not sensitive to different period lengths such as 2, 3, or 7 years. 58 In cases where the household head is single or divorced, or has no children, this is taken as an indication of absent U.S. ties.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

1.72 0.40

English knowledge (d) Long-term U.S. visa (e)

SD

1.49

4.59

1.31

2.09

0.64

1.28

1.00

4.00

1.00

1.00

0.99

1.00

Max

Min

0.00

0.00

0.00

-1.00

0.00

0.00

Q99

1.00

4.00

1.00

1.00

0.84

1.00

Q90

1.00

4.00

1.00

1.00

0.48

1.00

Q75

1.00

3.00

1.00

0.00

0.30

0.83

Q50

0.00

1.00

0.00

0.00

0.11

0.33

Notes: Cross-section includes last year of observation of all migrant household heads. (a) Values are averaged over 5 year period. (b) Mexican (U.S.) asset ownership is registered as minus one (one); zero if no asset is owned (see also text). (c) Proportion of non-head household members (spouse and kids) who migrated to U.S. during last three years or were born in U.S. (d) Index of English knowledge: “Neither speak nor understand” (0), “Do not speak, but understand some” (1), “Do not speak, but understand much” (2), “Speak and understand some” (3), “Speak and understand much.” (e) Dummy whether individual has U.S residency or citizenship.

Human capital

0.34

-0.30

% Household in U.S. (c)

Social capital

0.18

% Income remitted House index (b)

0.43

Mean

Physical capital

% Time in U.S. (a)

Home-time

Remittances

Indicator

Category

Table 3.1: Migration Intensity Indicators for U.S. - Mexican Case (Cross-Section)

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Emigrant or Sojourner? Migration Intensity and Its Determinants

51

Localized Human Capital: I construct a dummy of long-term U.S. immigration documentation (residency or citizenship) to represent localized human capital. Short-term immigration documentation is excluded since it constitutes no enduring investment. I further construct an ordinal variable of migrants’ ability to understand and speak English, ranging from “neither speaks nor understands” (0) to “speaks and understands much” (4).59 Table 3.1 presents descriptive statistics on the migration intensity indicators in the cross-section (that is, in each migrant’s survey year). Indicator averages and median values tend to be in the lower ranges of the variables’ sample distributions, which indicates that Mexican migrant household heads tend to have a stronger association with Mexico than with the U.S. Mexican Migration Intensity: Aggregation of Indicators Since the IMI is intended to succinctly summarize several decisions that indicate a migrant’s socio-economic reorientation, the underlying indicators should be correlated. For the present sample, I find that mutual correlation coefficients are statistically significant and strong with the expected signs, which are positive except for pairs that include a remittance indicator, where one would expect a negative correlation, since higher remittances reflect lower migration intensity (see Table 3.2).

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

59

I also create the following alternative indicators: I construct an index that measures the number of years during which the migrant did not return home (in the five-year period surrounding the observation year). I construct an index whether the migrant owns a business and where, in an analogous way as the home ownership index is constructed. Furthermore, I construct a dummy whether the migrant’s spouse has migrated during the last three years and a continuous variable that measures the proportion of resident minor children that have migrated during the last three years. These variables are not included in my main analysis but their correlations and alternative PCA results are shown in Tables C.6 and C.7 in Appendix C, respectively.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Networks English Documents

% Household in U.S. English knowledge Long-term U.S. visa 0.57

0.55

0.63

0.61

-0.22

-

Time

-0.28

-0.25

-0.41

-0.31

-

Remit

0.45

0.50

0.60

-

House

0.56

0.51

-

Networks

0.52

-

English

Pearson Correlation Coefficient

-

Documents

Notes: Table reports Pearson correlation coefficients. All correlations are statistically significant at the 1% level (H0=no correlation). Coefficients greater than 0.4 are in bold. Except for remittances, signing is such that higher values indicate stronger association to U.S.

Human capital

Social capital

House

% Income remitted House index

Remit

% Time in U.S.

Home-time Remittances

Physical capital

Time

Indicator

Category

Table 3.2: Linear Correlation Coefficients of Mexican Migration Intensity Indicators

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Emigrant or Sojourner? Migration Intensity and Its Determinants

53

I apply PCA to this set of indicators as described above. I first construct a cross-sectional IMI in which I include information from the last year of observation, and then I apply PCA to the retrospectively created panel to obtain a person and time-varying IMI (see Table 3.3). Time-variant information on English knowledge and remittances is not available, and both variables are dropped in the panel. Looking at the cross-section and panel PCA results, one can see that the first principal component is general in nature, loading strongly and with expected signs on all indicators. Moreover, the first principal component summarizes a large share of the indicators’ information, 57 and 63 percent of the variation in the cross-section and panel, respectively. In both cases there is a clear drop in explanatory power after the first component.60 Hence, one can conclude that the indicator variables are strongly correlated, supporting the hypothesis that they all reflect different facades of migration intensity. Given the generality of the first component, its strong explanatory power, and the drop thereafter, I conclude that, for Mexican male migrants, migration intensity can be summarized by a unidimensional Index of Migration Intensity.61

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

60

The eigenvalues of the second, third, and fourth principal component in the cross-section (panel) are 0.86 (0.75), 0.56 (0.4), and 0.5 (0.35), respectively; thus, explaining 14% (19%), 9% (10%), and 8% (9%) of the standardized variables’ variance, respectively. 61 I show in Appendix D-I that the construction of my index is robust to changes in the specification, such as using a different set of indicators or a different factoring technique, or switching from the panel to the cross-sectional specification.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

0.23 0.22

House index % Household in U.S. English knowledge Long-term U.S. visa

Physical capital

Social capital

0.57

0.22

0.00

0.02

-0.96

0.37

0.02

Q1

0.76

0.03

0.05

-0.75

0.19

0.07

Q2

1.51

0.31

0.08

-0.53

0.20

0.28

Q3

2.69

0.77

0.65

-0.05

0.10

0.76

Q4

3.65

1.00

0.97

0.72

0.02

0.86

Top 5%

(I) Cross-section Means in Sample Ordered With IMI

0.63

2.50

0.33

-

0.34

0.24

-

0.34

(II) Panel Factor Score Coefficient

Notes: Factor score coefficients are from first principal component. The IMI is the linear combination of all indicators, standardized by their mean and variance, and weighted by the factor score coefficients. Information on remittances and English knowledge is not available in panel.

3.39

Eigenvalue

0.23

-0.14

% of variance explained

Human capital

0.25

% Income remitted

Home-time Remittances

0.24

% Time in U.S.

Factor Score Coefficient

Indicator

Category

Table 3.3: Mexican Index of Migration Intensity (PCA Results)

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Emigrant or Sojourner? Migration Intensity and Its Determinants

55

Mexican Migration Intensity and Its Determinants: The CrossSection The cross-sectional IMI of Mexican migrant male household heads ranges between -0.56 and 0.62, and it is standardized to have a mean of zero and variance of one (see above). A higher value indicates a stronger socio-economic reorientation to the U.S., i.e., a higher level of migration intensity. To illustrate the distribution of the values that the cross-sectional IMI takes, I estimate its probability density function. As I do not know the functional form of the IMI’s probability density function, I use kernel density estimation, choosing a Gaussian kernel function and Silverman’s (1986) “rule of thumb” to find the optimal bandwidth (for more details see Appendix B-I). As Figure 3.1 depicts, the peak of the cross-sectional IMI’s estimated probability density function is strongly skewed to the left, indicating that most migrants in the sample are characterized by very low levels of migration intensity. This is not surprising since around one half of the sample is made up of former migrants, who currently reside in Mexico. The cross-sectional IMI’s estimated probability density function has also a fat right tail. Rather than a few distinct migration types, a continuum of migration strategies exists, each associated with a different level of migration intensity. Figure 3.1: Sample Distribution of Mexican Index of Migration Intensity 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

-0.6

-0.4

-0.3

-0.1

0.0

0.2

0.3

0.5

0.6

Notes: Kernel density estimation of cross-sectional IMI, using a Gaussian Kernel and Silverman’s “rule of thumb” to find “optimal” bandwidth.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

56

Mexican Labor Migrants and U.S. Immigration Policies

To illustrate these findings, I reorder the sample according to ascending IMI values and calculate the means of the IMI indicators in each quartile (see Table 3.3). In the first two quartiles Mexican migrant household heads show little socio-economic reorientation to the U.S. Only in the third quartile does socio-economic reorientation to the U.S. rise noticeably. In the top quartile migrants are more attached to the U.S. than to Mexico: they spend most of their time in the U.S., have moved most of their household members to the U.S., and possess longterm U.S. immigration documentation. Migrants in the top five percentile retain almost no socio-economic ties to the origin. Determinants of the IMI’s Cross-sectional Variation: Are crosssectional differences in migration intensity random or do they vary systematically depending on individual and community characteristics? Lucas (2005) suggests that more educated migrants are more likely to stay longer abroad, to remit less, and to reunify with their family aboard, i.e., they should have a higher level of migration intensity (see also Faini 2006; Amuedo-Dorantes et al. 2005; Dustmann and Mestres 2010). Massey et al. (1987, p. 286) postulate a “process of settlement” and increased “personal, social and economic ties” as migrants accumulate time abroad, implying that migration intensity should rise over time (see also Cornelius 1992). Besides these individual-level determinants (XI), factors that operate at the community or regional level (XC), such as differences in the local economy or public infrastructure, and macro-level factors (XM), such as changes in the U.S. economy or U.S. immigration policy (Waldinger 2008), are also likely to influence an individual’s migration intensity level. To test these hypotheses, I estimate the following crosssectional model of migration intensity with ordinary least squares (OLS):

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

(3-1)

IMIi =α+β1 X I +β 2 X iC +β3 X iM +ei ,

where IMIi is the cross-sectional measure of individual i’s migration intensity level, βi are the parameters to be estimated, α is an intercept term, ei is an independent and identically distributed error term, and all variables are explained above.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Emigrant or Sojourner? Migration Intensity and Its Determinants

57

Since Chapter 5 focuses on the time-variant macro-level determinants of migration intensity, I do not examine them in detail here but include year dummies to control for them.62 I start by analyzing how individual characteristics affect migration intensity. The results are reported in columns 1 and 2 of Table 3.4, which differ in that latter regression controls for community-level effects (XC) by including community dummies. Consistent with the findings of Lucas and others, I find that education, as measured by years of schooling, increases migration intensity significantly. Migration intensity also varies over time: I find evidence that migrants have a tendency to “get stuck” as their experience abroad grows. There is also a life-cycle aspect: Migration intensity is highest when migrants are in their twenties but lowest when they are in their sixties, perhaps because retirees come to value home attachment more highly. Consistently, family networks are a significant and strong determinant of migration intensity. As household size grows (wife and resident children), male household heads tend to reduce their level of migration intensity. On the other hand, family networks in the U.S., as approximated by the total number of family members among the household head’s nuclear or extended family who have ever migrated, tend to increase migration intensity.63 Comparative field studies have long noted differences in migration strategies across communities (e.g., Mines and Massey 1985). To investigate the determinants of this variation, I drop the community dummies and instead include several community characteristics. Throughout I include a control for the size of the community of origin, measured in thousands of inhabitants. Larger community size consistently causes a statistically significant but small increase in migration intensity.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

62

It should be noted that the need to control for time-variant factors in my cross-section is due to the fact that different communities were surveyed in different years. Year dummies should capture all macro-level factors (XM) that vary across time but not across individuals. 63 To reduce the danger of endogenous migration networks due to unobserved household level factors that affect both migrant and network members only non-household members are included in my indicator of U.S. networks.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

-0.119*** 0.012*** 0.004*** -0.013*** 0.025*** 0.002*** -0.002

Aged 60 - 69

Years of education

Years since first U.S. trip

Household size

U.S. networks (nuclear family)

U.S. networks (extended family)

Hectares of land owned

-0.065***

Average rainfall in state

-0.002*

-0.002

0.002***

0.027***

-0.014***

0.004***

0.012***

-0.115***

-0.055***

-0.003

0.024

0.088***

-0.001

0.002***

0.024***

-0.015***

0.004***

0.010***

-0.107***

-0.052***

0.006

0.031**

% below twice minimum wage

Local employment # factories per 1000

0.002 -0.058***

Aged 50 - 59

0.027*

Aged 30 - 39

Variable Individual characteristics Aged 18 - 29

Table 3.4: Cross-Sectional Determinants of Migration Intensity

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

-0.001

0.002***

0.026***

-0.014***

0.004***

0.011***

-0.120***

-0.057***

0.002

0.026*

Coefficient

-0.002

0.002***

0.025***

-0.013***

0.004***

0.012***

-0.122***

-0.057***

0.009

0.031**

-0.001

0.002***

0.025***

-0.013***

0.004***

0.011***

-0.122***

-0.058***

0.008

0.033**

-0.076***

-0.001

0.002***

0.025***

-0.012***

0.004***

0.010***

-0.121***

-0.057***

0.009

0.037**

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Yes

0.28

F-value

Adj. R-Squared

Notes: All regressions include year dummies. *p1). In choosing n and t migrants will balance the marginal benefits of increased home-time consumption against the marginal costs of lost consumption of goods due to reduced income, higher migration costs (in the case of n), and more remittances. The increase in remittances is due to the complementarity of remittances and home-time consumption. Changing n or t has distinct effects. While n is more effective in raising home-time consumption, due to migrants’ homesickness, it causes additional migration costs. Hence, migrants will balance their choice of n and t as well.

willingness to substitute between C and Z is just higher than the willingness to substitute between HT and R without changing the main dynamic of the model.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

74

Mexican Labor Migrants and U.S. Immigration Policies

A COBB-DOUGLAS EXAMPLE I now choose exemplary, but plausible, specifications for the utility function and the HT function in order to solve for a closed-form solution. As a simple specification for home-time consumption I set HT(n,t)=ntα, with 01. Replacing C by the budget constraint, noting that HT and R have to be equal at the optimum, and assuming the continuity of n, two first order conditions remain to describe the functioning of the second stage of my model:79 (4-3a)

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

(4-3b)

1 − δ αnt α−1 n(w * − w) + αnt α−1 = δ Z C 1 − δ t α t(w * − w) + t α + M = δ Z C

Equations (4-3a) and (4-3b) show that migrants need to balance the marginal benefits of more home-time consumption due to longer average stays in Mexico (4-3a) or more frequent returns (4-3b) against the opportunity cost of lost consumption due to missed U.S. income, increased remittances, and additional migration costs (in the case of 43b). Equations (4-3a) and (4-3b) uniquely determine tIM and nIM and thus the only interior solution (for n>1), which I will refer to as “internal migration” strategy (IM) or “repeat migration” (see Table 4.1). For the IM strategy (repeat migration) to be a valid course of action, all constraints of the maximization problem (4-3) have to be fulfilled. This is summarized in Conditions 1 and 2 in Lemma 1 in Appendix B-II. It turns out that Condition 1 not only (partially) insures the validity of the IM solution but, as numerical simulations below 79

Allowing any interior solution (n*, t*) to be non-integer with respect to n is an approximation that reduces mathematical costs considerably, while the comparative statics of (n*, t*) should approximate average effects well. To find exact solutions, the utility at the two integer values closest to n* (and associated t values) would need to be compared.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

76

Mexican Labor Migrants and U.S. Immigration Policies

show, further determines when repeat migration is the utilitymaximizing migration strategy (see Figure C.2 in Appendix C). If Condition 1 fails a return migration strategy (“RM”), the boundary where migrants make exactly one return trip (n=1), becomes a valid and, according to simulations, also optimal migration strategy (see Figure C.2).80 Finally, please note that because returning home is without cost in the model – only (re-) entering into the U.S. is costly – the utility of “emigration” (n=0) will always be at least matched by the utility of some RM strategy.81 Table 4.1: Optimal Solution of Internal Migration Strategy t IM =

n IM =

αM

HT IM =R IM =

*

(1 − α )(w − w)

δ(E-AD) D

⎛ δ ⎞ CIM =δt −α ⎜ E+AD ⎟ ⎝ 1-δ ⎠

δ(E-AD) α

t D

Notes: D = M + t α and E = 1 − δ w *Tt α . 1-α

δ

Comparative Statics of the Optimal Migration Strategy: One may now examine how marginal changes in parameters affect migrants’ optimal migration strategy.82 Higher migration costs increase the opportunity cost of home-time consumption. Hence, migrants consume more goods and less home-time. The reduction in home-time consumption lowers the marginal utility, and thus the equilibrium value, of complementary remittances. Moreover, in trying to reduce the price shock to traveling, repeat migrants (migrants with an IM strategy)

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

80

RM is also optimal for parameter sets for which migrants would want to make less than one, but more than zero, return trips, such as n=0.5, which is ruled out by assumption. 81 This may imply a “degenerated” RM strategy, in which t is infinitesimal or even zero; latter case perfectly duplicates emigration (except for n ≠ 0). To avoid this peculiarity, the simulations below classify any migrant with a RM strategy and an equilibrium HT value close to zero ( that is, below a certain minimum value) as “emigrant.” 82 The mathematical details behind the cited results are given in Appendix B-II.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

A Comprehensive Model of International Migration Behavior

77

make fewer trips but stay longer.83 Since return migrants cannot further save on traveling expenses, they work more in the U.S., which causes total U.S. time to rise as migration costs ascend. For repeat migrants the change in total U.S. time is analytically indeterminate, but simulations suggest that it rises as well (see Section II in Figure 4.1). In sum, increasing migration costs strongly intensifies migration strategies, migrants make fewer trips, are likely to increase total U.S. time, remit less, and some switch to more permanent migration strategies.84 Ceteris paribus, a rise in the real U.S. wage rate has, for most parts, the opposite effect on repeat migrants: their migration intensity falls.85 A higher U.S. wage rate enlarges migrants’ total income. As a consequence, repeat migrants consume more goods but also remit more and travel more. Since the opportunity costs of t rise, stays in Mexico shorten. Nevertheless, due to more frequent traveling, home-time consumption and complementary remittances rise overall, if the nonstringent Condition A-2c holds (see Appendix B-II). While it is analytically unclear how total U.S. time varies, simulations suggest that it falls, adding to the de-intensification effect of an increased U.S. wage rate. Heterogeneity in tastes matters. Individuals with a stronger desire to consume goods (higher δ) consume more goods. This is due to both a substitution effect and an income effect, as consumption-hungry individuals work more in the U.S. Since the substitution effect dominates remittances and home-time consumption fall. Individuals who disdain living abroad (low A) act just the opposite. They cut back on consumption of goods but enjoy more home-time. The impact of a change in α is tricky, since a reduction in α increases not only migrants’ homesickness but, ceteris paribus, also lowers the utility of home-time consumption.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

83

Technically, for t* to rise with M, Condition A2b must hold (see Appendix BII). According to numerical simulations, Condition A2b virtually always holds if the internal solution is valid (compare Figures C.3 and C.4 in Appendix C). 84 Let C1 be as in Lemma 1 in Appendix B-II, then ∂C1 α 1 α = C1 − (A + t α ) − Dt α ∂M M δ δM

is negative if C1 is close to zero. Hence, some repeat migrants will switch to a return migration strategy as migration costs rise. 85 It is analytically tricky to determine how a rising U.S. wage rate affects return migrants in my model except that consumption is likely to go up.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

78

Mexican Labor Migrants and U.S. Immigration Policies

Staying or Migrating?

So far, the discussion of the Cobb-Douglas example has focused on the second stage of the model, the choice of the optimal migration strategy, and has skipped the first stage. In order to find out whether staying (non-migrating) might be preferable to migrating, one has to compare the utility of staying to the utility of the optimal migration strategy; that is, to the utility of the optimal IM or RM strategy. Regrettably, the resulting mathematical expression is rather difficult to interpret. Hence, I resort to numerical simulation below. As Figure 4.1 shows, depending on the parameter set, either staying or migrating could be preferable. However, it is possible to determine analytically how marginal changes in parameters affect the decision whether or not to migrate (for mathematical details see Appendix B-II): As expected, higher migration costs or a reduced wage gap increase agents’ likelihood to stay, as the net benefit of migration is smaller. Individuals with an increased preference to consume goods are likelier to migrate, as their willingness to accept the implied loss in home-time consumption is now higher. Agents who disdain living abroad more (lower A) are less likely to migrate. Policy Discussion86

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

How does individuals' behavior change if policy relevant parameters vary? By increasing migration costs, more stringent external border enforcement has two distinct and opposing impacts (see Proposition 1). On the one hand, reinforcing border controls might cause deterrence due to a net income reduction that applies to all migration strategies. On the other hand, heightened external border enforcement leads to an intensification of optimal migration strategies: Migrants try to avoid the cost shock to border crossing by adapting more permanent migration strategies.

86

Other studies have analyzed the host-country's optimal allocation of resources to border enforcement (Ethier 1987, Guzmann et al. 2007) or the optimal distribution between internal enforcement, border control and amnesty programs (Chau 2001; Epstein and Weiss 2009). All focus on the “stay-versus(e)migrate” decision.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

A Comprehensive Model of International Migration Behavior

79

Proposition 1: The Impact of External Border Enforcement I.

Deterrence effect a. Increases in migration costs deter some individuals from migrating. b. However, if non-migrants benefit from other migrants’ transfers, then this deterrence effect is weakened or, at worst, could reverse. II. Intensification effect a. Increases in migration costs lead to an intensification of migrants’ optimal migration strategies (increased migration intensity). b. Return migrants (n=1) spend more time in the U.S. Repeat migrants (n>1) make fewer but longer trips and their total U.S. time is likely to rise. Remittances and home-time consumption fall. Some repeat migrants switch to a return migration strategy.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

It should be noted that the deterrence effect of external border enforcement, which is commonly stressed, depends on the (unspoken) assumption that non-migrants’ fallback is unaffected by changes in border enforcement. However, in labor-exporting regions, such as many parts of Mexico, a large share of non-migrants benefit from other migrants’ transfers (R). This is due to both direct income transfers as well as indirect multiplier effects of spent remittances.87 Recalling that remittances fall if migration costs rise, due to both the intensification and deterrence effect, we note that non-migrants’ income falls as migration costs mount, or dR/dM1), “RM” to return migrant (n=1), and “EM” to emigrant (n=0). Return migrants who spend less than one tenth of their time in Mexico are classified as emigrants. Panel II depicts the equilibrium values of the optimal migration strategy, which is either repeat migration (light area) or return migration (dark area). Throughout, all nonvariant parameters are equal to default values (see Table C.8). If the wage ratio varies then the U.S. wage rate (w*) is fixed and the Mexican wage rate (w) varies.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

86

Mexican Labor Migrants and U.S. Immigration Policies

Former U.S. President Bush and several U.S. Congress members have proposed to expand U.S. guest worker programs. Some bills proposed to U.S. Congress have further included programs of “earned legalization.”93 Section I-B in Figure 4.1 attempts to illustrate the possible consequences of such programs. First of all, it needs to be noted that predicting the net impacts of such policies is difficult due to conflicting effects: On the one hand, migration costs (M) are likely to fall if a migrant obtains a visa. On the other hand, his comfort of living in the U.S. (A) may also improve.94 Since guest worker programs are generally temporary, but becoming comfortable abroad requires time and long-term planning assurance, it seems reasonable to expect that their main effect would be a reduction in M. This would de-intensify the migration strategies of the beneficiaries. However, if legalization causes a substantial increase in A, which is most likely in the case of permanent residency visas, the beneficiaries could be intensified. LOCALIZED INVESTMENTS: INVESTING IN U.S. COMFORT

For a migrant, not only the choices of remittance sending and return and repeat migration are interdependent, the focus of the model presented so far, but other location-dependent choices are also linked, in particular “localized investment behavior.”95 This section presents a simple extension to indicate how such localized investment behavior could be incorporated into the Cobb-Douglas model. I start with the migrant’s decision problem, as described in equation (4-3), and assume that he now possesses the ability to affect his displeasure of living abroad. More specifically, I assume that the migrant may invest f units into activities or assets that increase his relative comfort of living in the U.S. The migrant’s relative comfort of

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

93

“Earned legalization” refers to the offering of long-term permanent residency to undocumented migrants based on criteria, such as continuing employment, absence of criminal history, or payment of tax obligations (CRS 2006a). 94 Program details, such as visa duration, the conditions for renewal, possible family reunification, labor market effects, or possible housing provisions will also be important. 95 In Chapter 3 I provide a definition for “localized investments.” Furthermore, I find empirical evidence that for Mexican migrants, various forms of localized investments, such as home ownership, social network membership, or language knowledge are correlated to migration patterns and remittances.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

A Comprehensive Model of International Migration Behavior

87

living abroad (A) becomes a function of f or A(f). For example, the migrant might induce family members to move to the U.S., improve his U.S. home, or join a club in the U.S. I assume that A(f) is twice continuously differentiable, increasing in f (Af>0), and strictly convex (Aff 1 . If C1=0 then n

IM

= 1.

2) Condition 2: If 1 AD ⎫ ⎧ ⎫ ⎧ C2=D ⎨ A + Ttα −1 ⎬ − E = w* {1 − α ( 1 − δ )} − w + α ⎨tα −1 + δ α ⎬ δ Tt ⎭ ⎩ ⎭ ⎩ IM IM is positive then n t < T . If C2=0 then

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

n IM t IM = T .

Inserting C1 and C2 into the equilibrium values given in Table 4.1 proves Lemma 1. Condition 1 ensures that trip frequency is larger or equal than one.156 Condition 2 guarantees that repeat migrants (migrants that use an internal migration strategy) do not spend more time at home than their total time endowment T. Numerical simulations show that a failure of Condition 2 is highly unlikely for the U.S. Mexican case, as the wage gap would need to vanish (see Figure C.3 in Appendix C). Return Migration (RM): The partial derivation of the Lagrangian associated with equation (4-3) by n, at the boundary where n=1, proves that C1 of Lemma 1 has to be non-positive for the Lagrangian multipliers not to become negative. The partial derivation by t can be simplified into the implicit function G: 156

Condition 1 has an intuitive interpretation. If either the relative desire for home-time consumption is increased (lower δ), the cost of absence is increased (lower A), or the relative migration costs are lowered (M/w*) then more repeat migration strategies (IM) are preferred over one-trip RM strategies.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Appendix B

159

{

}{

}

{

}

1 − δ α−1 G(.,t)= w * + αt α−1 A + t α − α t Tw * − (tw * + M + t α ) = 0 , δ

which defines tRM; as nRM is equal to one, the solutions of HTRM and CRM follow easily. 2.) Optimal Migration Strategy: Partial Derivatives

Repeat Migration (IM): Let X= (nIM, tIM, RIM, CIM, tIM*) be the optimal internal migration solution as defined in Table 4.1. Then the partial derivates are (with superscripts suppressed): ⎧ ⎧ α ⎫ α ⎫ + t α ⎬ (E-AD)+E ⎪ ⎨ ⎪ (< 0) − α (1 ) M ⎩ ⎭ ⎪-δ ⎪ α 2 ⎪ ⎪ t D ⎪ ⎪ t ⎪ ⎪ (> 0) ⎪M ⎪ ∂X ⎪ ⎪ = ⎨ δE ⎬ (< 0) ∂M ⎪- 2 ⎪ D ⎪ ⎪ ⎪ δA ⎪ (> 0) ⎪ tα ⎪ ⎪ α ⎪ (> 0)b 2 α-1 * ∆ α α δ + αδ Tt w+ t (1)w AD ⎪ ⎪ ⎪ ⎪ δ(1-α)∆w(E-AD) ⎩ ⎭ ⎧ ∂n ⎫ ⎪ ∂α ⎪ ⎪ ⎪ t ⎪ ⎪ ⎪ α(1 − α) ⎪ (> 0) ⎪ ⎪ , * α ∂X ⎪ (1 − δ)w MTt ⎪ (> 0) # =⎨ ln t ⎬ ∂α ⎪ (1 − α)D 2 ⎪ ⎪ δMA ⎪ (< 0) # ⎪⎪ ln t ⎪ (1 − α)t α ⎪ * ⎪ ∂t ⎪ ⎪ ⎪ ⎩ ∂α ⎭

}

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

{

,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

160

Appendix B ⎧ δ ⎧ DE α ⎫ ⎫ (E − AD) ⎬ ⎪ α 2⎨ *+ ⎪ ( > 0) w ∆ ⎭ ⎪t D ⎩ w ⎪ ⎪ t ⎪ ⎪⎪ (< 0) ⎪ ∆w ⎪ α ⎪⎪ ∂X ⎪⎪ (1 − δ)Tt ⎧ t∆w α⎫ * c − tw + t ⎬ =⎨ ⎨ ⎬ ( > 0) 2 * α D ⎩ ⎭ ∂w ⎪ ⎪ ⎪ Tt α + At ⎪ ⎪δ ⎪ ( > 0) α t ⎪ ⎪ ⎧⎪ ⎫⎪ ⎪ Tt α 1 t ⎪ (< 0)c,d αt α ADt DE (E AD) − − − + ⎬ ⎪ ⎪ 2 ⎨ ∆w w* ⎭⎪ ∆w ⎭⎪ ⎩⎪ δ(E − AD) ⎩⎪ ⎧ δ ⎫ (< 0) ⎪ − α ⎪ t ⎪ ⎪ (= 0) 0 ⎪ ⎪ ∂X ⎪⎪ − δ ⎪⎪ (< 0) =⎨ ⎬ ∂A ⎪ δD ⎪ α ⎪ t ⎪ (> 0) ⎪ Tt α D 2 ⎪ ⎪ ⎪ 2 ⎩⎪ δ(E-AD) ⎭⎪ (> 0)

,

,

M 1− δ * α where ∆w=w * - w , D = w Tt . Superscripts + t α and E = 1-α δ “b)”,”c)”, and ”d)” require that for the indicated sign to be valid, the non-stringent Condition A2-b, A2-c or A2-d holds, respectively; #) requires tIM>1 for the indicated sign.

Condition A2: b)

AD ( 1 − α )∆ wAD ⎫ ⎧ w* {1-α ( 1 − δ )} > w − ⎨α 2 t α −1 + αδ α + ⎬ E − AD ⎭ Tt ⎩

c)

w* (1-α ) ≥ w − α tα −1

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

d)

δ AD ⎫ ⎧ w* δ ≥ w − ⎨α t α −1 + α ⎬ Tt ⎭ ⎩

B) Return Migration: Let G(.) be the implicit function that defines . Then, the partial derivative of G with respect to tRM is

RM

t

∂G 1− δ 1− δ ⎫ 1− δ ⎫ ⎧ ⎫ ⎧ 2( α−1) ⎧ = α(α − 1)t α− 2 ⎨ A − (Tw * − M) ⎬ + α∆wt α−1 ⎨1 + α ⎬ + α (2α − 1)t ⎨1 + ⎬ ∂t δ δ ⎭ δ ⎭ ⎩ ⎭ ⎩ ⎩

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Appendix B

161

(with superscripts suppressed). The first term in braces must be negative for G(.)=0 to hold. Hence, the partial derivative will be positive as long as α>0.5. Since this is not a restrictive assumption for a return migrant (see Table 4.2), I assume that it holds when indicating the signs of the following partial derivatives, making use of the implicit function theorem:157

 

1 − δ α−1 t δ (< 0) ∂G ∂t RM 1 α (Tw * − M)t α−1 − α∆wt α + αt 2α−1 ∂t (< 0) =− 2 ∂G ∂δ δ ∂t ⎧ 1− δ ⎫ A + t α−1 ⎨α (t − T) + t ⎬ ∂t RM δ ⎩ ⎭ ( = ?) =− * ∂ G ∂w ∂t ∂t RM =− ∂M

α

∂t RM ∆w + αt α−1 =− (< 0) ∂G ∂A ∂t

Please note that ∂HT RM ∂R RM ∂t RM ∂ (nt)RM ∂t RM = = αt α−1 , , = ∂X ∂X ∂X ∂X ∂X

and

∂t*RM ∂t RM =− ∂X ∂X

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

for return migrants (where X may be any parameter). Finally, the partial derivatives of CRM are:

157

Please remember that as α falls, the degree of homesickness rises. Numerical simulations show that an agent for whom α is below 0.5 is characterized by such a strong level of homesickness that he would not choose a return migration strategy.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

162

Appendix B ∂CRM ∂t =− ∆w + αt α−1 − 1 (< 0) # ∂M ∂M

{

}

∂CRM ∂t =− ∆w + αt α−1 (> 0) ∂A ∂A

{

}

∂CRM

{

}

∂t =T−t− ∆w + αt α−1 (> 0) # ∂w * ∂w * ∂CRM ∂t =− ∆w + αt α−1 (> 0) ∂δ ∂δ

{

}

where ∆w=w * - w and “#” denotes the fact that δ, αδ, α > 0.5 is a sufficient condition to obtain the indicated sign. 3.) Migrating or Staying

Let UNM, URM, and UIM be the maximal utility of non-migration, return migration, and repeat migration, respectively. Let ∆NMRM=UNM – URM and ∆NMIM=UNM – UIM. Then, the following is true:

 ∂ (∆ NMRM ) ∂M

=

A+ tα (> 0) F

∂ (∆ NMRM ) 1 = − 2 log(A + L) − log(A + t α ) (< 0) ∂δ δ

{

∂ (∆ NMRM ) ∂w *

=−

}

(T − t)(A + t α ) (< 0) F

∂ (∆ NMRM ) 1 ⎧ (1 − δ) (1 − δ) ⎫ = ⎨ − ⎬ (< 0) ∂A δ ⎩ A + L A + tα ⎭   1− δ ∂ (∆ NMIM ) = ∂M

E(E − AD) + AD(AD + (

δ

E)

1− δ E + AD)(E − AD 2 ) δ

(> 0)

∂ (∆ NMIM ) 1 = − 2 log(A + L) − log(A + nt α ) (< 0) ∂δ δ

{

∂ ( ∆ NMIM ) Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

∂w *

}

∂C =−

∂R ∂w * − 1 − δ ∂w * (< 0)c C δ A + nt α

∂ ( ∆ NMIM ) 1 ⎧⎪ (1 − δ) (1 − δ) 2 ⎫⎪ = ⎨ − ⎬− ∂A δ ⎪⎩ A + L A + nt α ⎪⎭

∂C

∂A* (< 0) C

where F = (Tw * − t(w * − w) − M − t α )(A + t α ) ; c) requires the nonstringent Condition A2-c) to hold for the indicated sign.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Appendix B

163

III) A SIMPLE MODEL OF MIGRANTS’ RECEIPT OF SOCIAL NETWORK HELP

In each situation in which a migrant i might need some assistance type j (see Chapter 2), I assume that he has two options: either he can draw on help by others (H) or not (NH). Furthermore, led the utilities UH and UNH be associated with the choices H and NH, respectively. I assume that the difference in these two utilities, ∆U = UH - UNH, is made up of a random and a structural part, latter of which is driven by two main factors: a migrant’s need or desire to receive help (XN) and his costs of obtaining help (XC).158 I expect that a migrant’s likelihood to receive assistance increases if his need is higher, examples are: having no immigration documentation, being a dependent, or having less migration related human capital (see Massey and Espinosa 1997). I further expect that an individual is less likely to draw on help if his costs of obtaining assistance are high, chiefly, if he has limited access to migration networks. It is impossible to observe the utility difference ∆U. Yet, one may observe the discrete random variable, yij, recording if the migrant i received help for assistance type j. This discrete random variable should take on the value of one if UH > UNH holds and zero else: (B-5)

{

} {

}

yij = 1 ∆Uij > 0 = 1 β1jXijN + β2 jX ijC + εij > 0 .

If eij, the random part of ∆Uij, follows a standard normal distribution, which I assume, it can be shown that the probability that individual i chooses help for assistance type j is

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

(B-6)

(

)

Pij = F α + β1jX ijN + β2 jX ijC ,

where F(.) is the standard normal cumulative distribution function, β are the parameters to be estimated, and α is an intercept (Greene 2005). Equation (B-6) can then be used to obtain the joint probability density function of the sample, which can be used to estimate the parameters β by means of maximum likelihood estimation (Greene 2005).

158

This is known as a “random utility model.” See Chapter 2 in Train (2003).

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved. Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

APPENDIX C

Additional Tables and Figures

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

I) ADDITIONAL TABLES

165

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Networks & employment Impact of migration Networks & employment Border smuggling Migration patterns European emigration

Aguilera & Massey (2003) Alarcón (1992)

Networks & female migration Migration theories

Bastida (2001)

Boyd (1989)

Prior info & migration.

Banerjee (1984, 1990)

Baines (1995)

Avila et al. (1998)

Amuedo & Mundra (2007) Andreas (2001)

Ethnic brokers

Adler (2002)

Eremittances carry info

F - specific vs. general, motivational L

E - letters & stories

L - “Nortenization”

General Information (i) Study Research topic Information & risk

L

Eremittances finance trips

E

Initial financing

E/S–sustain at border

E - protection

L - border crossing

L - COD, refer smuggler E - travel company

E - COD

Type of Assistance (ii) Domestic Border crossing/ dislocation smuggler

Table C.1: Summary of Literature on Social Network Assistance

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

L

E

E - search housing

Initial U.S. lodging

L

E - pre-arrange / refer / recruit for job E - search job

F

E - refer / recruit job F - refer / search job L

U.S. employment

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Dynamic migration model Traditional health care Housing & settlement MX - U.S. migration

Job aid

E - stories

F - specific job info, expectations L

Migration & L assistance Cohen et al. (2003) Transnationalism Colussi (2003) Migration decision Cuecuecha (2008) Social assistance Durand (1994) Mexican E - “rules of migration game,” costs & objectives Epstein (2002) Location choice migration is signal

Choldin (1973)

Chavez (1992)

Chavez (1990)

Carrington et al. (1996) Chavez (2003)

Caces (1986)

General Information (i)

E - COD, refer / prearrange smuggler, travel company

S E

E - border sustenance,

E - finance

S S - housing

E - travel company; S - sustenance

E - COD, refer/ prearrange smuggler

F/S - LT housing E

L

E

L

Type of Assistance (ii)

Table C.1 Cont.: Summary of Literature on Social Network Assistance

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

E - search / recruit for job

L - search job

E/S - prearrange, refer / recruit for job S - search

L

search job (model)

F - info / teach / search job

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

International migration Adaptation strategies MX internal migration Migration theories Network assistance Migration case study MX -U.S. smuggling

Graeme Hugo (1981) Graves & Graves (1974) Kemper (1977)

Lomnitz (1976, 1977) Lopez-Castro (1986) Lopez-Castro (1998)

Lindstrom (1996)

Transnationalism

MX migration history Network development Co-residence

Migration & social capital

Goldring (1992)

Glick (1999)

Espinosa (1997) Massey & Espinosa (1997) Fernández-Ruiz (2003) Garcia (2005)

General Information (i)

Lmotivational L - border crossing

L

E - cheap housing E - migration is signal

E - costs, feasibility, migr. culture E -“mental condition.”,

E

L

L

E

E - initial finance

S

L - info,

E - share costs

E - border sustenance

L - evading INS, travel company F - refer smuggler, COD

E/S - search / refer / teach on job E - search / recruit for job F/S -refer / search job

L

E - free ST housing E

E - free housing L

E

E - refer / teach on job E - refer job

E - search/ recruit for job L

E - search job

F/S - LT housing E - access to E - refer / search housing job L L

E /S - refer & search E /S - ST or smuggler; L- COD, LT housing travel company E - refer smuggler, E COD L

Type of Assistance (ii)

Table C.1 Cont.: Summary of Literature on Social Network Assistance

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Sassen (1995)

Rodriguez (2001)

Papail (1998)

Munshi (2003)

Morawska (1990) Networks & employment Migration & networks Mexican migration Labor market (general)

Migration

E- jobs

L - job vacancy

Marriage & networks MacDonald & Italian chain EMacDonald (1964) migration opportunities Massey, et al. Mexican (1987) migration Menjivar (2000) Migration & L - risk networks reduction Mestries (1994) Labor migration L - job vacancy Mines (1981, Migration case E - job advice 1984) study Monto (1994) Labor migration

Luke et al. (2004)

General Information (i)

S - financing

E/S - pay transport

S

E

E - border sustenance

E - refer smuggler

S - COD

E/S

E- COD, L- travel company L - info, COD E - protection E - COD, refer smuggler

E

L - refer job

S - free ST housing E

E - info / refer / recruit / search job

S - search / prearrange job E - refer job

E - info / refer/ recruit for job E E - info / refer / search on L E –search / recruit / teach on job E/S - LT E/S - info / teach housing on job S E - refer / search, recruit for job E - pool E - refer job resources F - search job

E

L E - COD

Type of Assistance (ii)

Table C.1 Cont.: Summary of Literature on Social Network Assistance

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Migration & return

Migration and family Networks & employment Networks & relocation Migration & networks

U.S. – MX smuggling Mexican migration Migration decision Networks & migration

Border crossing

F - job type & vacancy F - job vacancy E - general vs. specific knowledge L - risk reduction

L - migration culture info and risk (model) Lmotivation

L - border crossing

L

L

L

L - refer smuggler, COD

L

F - info on/ refer smuggler F-COD, pre-arrange smuggler L - refer smuggler L -protection

Type of Assistance (ii) E - moral support F - border sustenance

L

L

L

L - free ST housing

S

L

L - pre-arrange / search / refer job

F - refer / info on job E- search / recruit for job L - info on job

L - search job

S

L

L- search job

Notes: All listed studies relate to social networks and migration, mostly Mexico – U.S. migration. Only migration process from decision making to first U.S. employment is considered (see text). Indicates if a study focused on a certain type of assistance (F), gave further examples or explanations for it (E), provided statistics on it (S), or merely listed it (L). “COD” refers to “cash on deliverance.”

Winters, et al. (2001), Davis & Winters (2001) Zahniser (1999)

Wilson (1994)

Trigueros & Rodriguez (1988) Waldinger (1997)

Tilly & Brown (1967)

Taylor (1986)

Singer & Massey (1998) Spener (2001, 2004) Tapia (2003)

General Information (i)

Table C.1 Cont.: Summary of Literature on Social Network Assistance

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Appendix C

171

Table C.2: Social Network Assistance in Other Mexican Migration Studies Frequency of assistance (in percent) Initial Finding Smuggler U.S. Job Study financing smuggler payment lodging search > 75 75 Chavez (1990) (*) 80 Cohen et al. (2003) 15 17 23 64 64 Espinosa (1997) (+) 70 Garcia (2005) Genicot & Senesky 47 70 60 (2004) 40 Glick (1999) (*) 9 40 Massey et al. (1987) 27 59 Mines (1981) (*, #) Papail (1998) 29 95 75

Notes: All entries are in percent and rounded. (*) Lodging assistance refers to long-term housing. (+) Assistances 3, 5 and 6 report “information on border crossing,” “help finding a place to live,” and “information on available jobs,” respectively. (#) Shows only reported assistance levels since 1970.

Table C.3: Social Network Assistance by Type of Helper (MMP) Relative frequency of assistance Type of Assistance

Kin

Friends

Community

Other

Travel companionship

36%

59%

0%

6%

Paying Smuggler

75%

15%

0%

10%

Initial U.S. housing

56%

26%

7%

11%

Obtaining first U.S. job

51%

41%

0%

4%

Average

55%

35%

2%

8%

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Notes: If no assistance was received, or assistance was not applicable or unknown observation was dropped.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

172

Appendix C

EMIF2 96 31 71 57 4 49 31 8 9 3

MMP 93 42 100 67 38 28 20 9 4 2

3

N/A

N/A

2

16

N/A

N/A

12

4

N/A

N/A

7

U.S. Employment (2)

Unemployed (%) Commerce (%) Service and Retail (%) Agriculture (%) Manuf. & Industry (%) Other or unknown (%)

N/A N/A N/A N/A N/A N/A

0 2 26 33 13 26

N/A N/A N/A N/A N/A N/A

0 0 25 20 46 9

Origin (3)

Urban (>2500 inh.) (%) Migration states (%) (I) North. Border States (%) (II) Various A (%) (III) Interior States (%) (IV) Various B (%) (V) South-West. States (%) (VI)

48 100 0 0 0 0 0

43 26 11 18 12 18 15

43 39 12 14 12 17 8

90 47 13 24 4 7 8

California (%) Texas (%) Illinois Arizona (%)

N/A N/A

29 26

56 19

N/A

14

5

53 14 12 2

Networks (1)

Individual Characteristics (1)

EMIF1 95 28 81 72 4 47 39 9 2 1

Destinati on (2)

Table C.4: Profile of Mexican Migrants and Social Networks Variable OWN Individual is male 100 Age of individual 36 Individual is household head 74 Individual is married N/A < 6 years of education (%) 20 6-8 years of education (%) 32 9-11 years of education (%) 27 12-14 years of education (%) 5 > 14 years of education (%) 7 Number of U.S. trips 1

# of nuclear family members with U.S. trip # of extended family members with U.S. trip # of friends with U.S. trip (*)

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

#

Notes: Table reports mean or percentages (“%”). (1) Refers to point of interview, except for U.S. trips and networks in own sample, which are at time of trip; in own sample, 10 percent of interviewees did not report educational attainment (*) only best friends for own sample. (2) For EMIF1, intended U.S. job and intended U.S. destination. (3) I Guanajuato, Michoacán, Jalisco and Colima; II - Baja California, Sonora, Chihuahua, Coahuila, Nuevo León and Tamaulipas; III-Sinaloa, Durango, Nayarit, Zacatecas, San Luis Potosí and Aguascalientes; IV - Federal District, State of México, Querétaro, Hidalgo and Tlaxcala; V - Oaxaca, Guerrero, Puebla and Morelos; VI - Veracruz, Tabasco, Chiapas, Campeche, Yucatán and Quintana Roo.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

7.70 9.98

# of siblings and parents of migrant who have ever migrated 4.75 # of uncles, aunts, cousins, nephews and siblings-in-law of 15.27 migrant who have ever migrated hectares of land owned prior to first U.S. trip 0.33 # of factories in municipality per 1000 inhabitants % in municipality that earns less than twice minimum wage Average rainfall in state since 1941 (liters) Dummy variable indicating if community has a clinic # of sport fields in community per 1000

U.S. networks (nuclear family)

# factories per 1000

% below 2x minimum wage

Average rainfall in state

Clinic in community

# sport fields per 1000

Hectares of land owned

U.S. networks (extended family)

# of household members (wife and children)

Household size

0.73

0.42

0.74

0.33

1.97

4.83

18.34

2.44

1.46

0.74

0.51

0.00

55.70

0.00

0.00

0.20

0.07

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

-0.55

Min

10.12

8.75

36.19

12.38

0.96

1.09

1.28

# years since first trip to U.S.

6.12

0.12

0.16

0.25

Time since first U.S. trip

Aged 50 - 59

1.37

Dummy variable indicating if individual is aged 60-59

Dummy variable indicating if individual is aged 50-59

Aged 40 - 49

0.31

1.00 1.09

Years of education

Dummy variable indicating if individual is aged 40-49

Aged 30 - 39

0.00 0.16

SD

Years of education

Dummy variable indicating if individual is aged 30-39

Aged 18 - 29

Mean

Aged 60 - 69

Index of Migration Intensity Dummy variable indicating if individual is aged 18-29

IMI

Description

Variable

Table C.5: The Cross-Sectional Determinants of Migration Intensity: Descriptive Statistics

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

5.00

1.00

1.52

0.96

43.83

115.00

181.00

12.00

19.00

64.00

28.00

1.00

1.00

1.00

1.00

1.00

0.62

Max

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

0.33 0.86

% of households in community with a phone Dummy variable indicating if community has a post office Air-line distance of state capital to U.S. border (miles, thousands) Dummy variable indicating if street from community to next highway is paved # of inhabitants in community (millions)

Post office in community

Distance to U.S. border

Source: MMP 114

Community population

Highway access paved

0.82

0.74

0.93 0.12

0.43

1.02

0.53

4.57

0.38

1.27

1.09

0.84

% with phone

# crimes per 1000

Culture of return

0.81

SD

0.92

Dummy variable indicating if community has a central plaza Dummy variable indicating if migrants usually return home for the community’s patron saint's day # of crimes in municipality per 1000 inhabitants

Plaza in community

Mean

Description

Variable

Table C.5 Cont.: The Cross-Sectional Determinants of Migration Intensity: Descriptive Statistics

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

0.00

0.00

0.01

0.00

0.00

0.00

0.00

0.00

Min

1.65

1.00

0.60

1.00

0.86

14.36

1.00

1.00

Max

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

SD

HT

Remit

PC

SC HC

1 2 3 4 5 6 7 8 9 10 % of time in US 1 0.43 1.28 HT 2 0.35 1.22 0.95 # years w/o return % of Income remitted 3 0.18 0.64 -0.22 -0.21 Remit Remittances (dummy) 4 0.76 1.29 -0.32 -0.31 0.48 House Index 5 -0.30 2.09 0.61 0.58 -0.31 -0.34 PC Business Index 6 -0.11 1.42 0.41 0.37 -0.17 -0.24 0.43 % HH in US 7 0.34 1.31 0.63 0.60 -0.41 -0.45 0.60 0.38 8 0.33 1.43 0.61 0.57 -0.41 -0.46 0.56 0.37 0.90 SC Wife in US % res. kids in US 9 0.31 1.32 0.55 0.52 -0.34 -0.36 0.53 0.35 0.86 0.68 English knowledge 10 1.72 4.59 0.55 0.49 -0.25 -0.24 0.50 0.28 0.51 0.49 0.44 HC Long-term Permit 11 0.40 1.49 0.57 0.48 -0.28 -0.29 0.45 0.30 0.56 0.52 0.51 0.52 Notes: Table reports Pearson correlation coefficients in cross-section. Coefficients that are greater than 0.4 and 0.8 are in bold and italics, respectively.

Mea

Table C.6: Correlation Coefficients of IMI Indicators (Alternative Indicators)

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

0.96 11 1 5 1,601

4 1 4 4,324

Dummy freeb

0.94

Panela

1 7 4,324

13

0.87

2 9 4,322

4

0.97

Cross-Section Alternative Large indicatorsc setd

1 6 4,324

0

0.99

1 6 4,324

0

0.99

MLF

Factoring PAF

e

Panel

1 4 1,519

0

0.97

1960s

1 4 1,519

0

1.00

1980s

2000s sample observed inf

.

Notes: For each column a separate factor analysis is carried out and factor scores are computed to order sample. For columns one to six, results are compared to cross-sectional IMI (see Table 3). Last two columns compare independent PCA results of panel (see ‘f’). a Uses only indicators that are also available in panel (excludes indicators for remittances and English knowledge). b All non-interval indicators are replaced by continuous substitutes or dropped. c Alternative indicators for business ownership, return behavior, migration behavior of wife, migration behavior of children, and migrants’ remittance behavior are used; due to lack of alternatives both human capital indicators are kept. d Uses mix of standard cross-sectional and alternative indicators; factor scores are from first principal component. e Uses different factoring techniques: “MLF” and “PAF” refer to maximum Likelihood and principal axis factoring, respectively. f Compares factor scores and rank of 2000s sample weighted with “correct” weights (2000s weights) with factor scores and rank when weighting 2000s sample with weights of different decade (that is, weights taken from PCA of 1960ies or 1980ies sample, respectively)

Spearman rank correlation coefficient % more than 0.5 x SD difference in factor scores No. of components (Eigenvalue > 1) No. of indicators No. of observations

Evaluation criterion

Table C.7: Robustness of the IMI

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

w w M T s L α A δ

Mexican wage

Migration costs (per trip)

Total time

$-Time scale

HT of non-migrants

Degree of homesickness

Comfort of living in U.S.

Desire for consumption

*

Symbol

U.S. Wage

Parameter/ Variable

Default Value

120

N/A

1

120

1.5

.25

1.2

0.65

15

0.65

Taste Parameters

--

$/months

months

1000 $

1000 $

1000 $

System Parameters

Unit

Table C.8: Simulation Baseline Parameter Values and Ranges

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

[0.35, 0.95]

[3, 45]

[0.35, 0.95]

--

[0.2, 5]

120

[0.25, 15]

[0.125, 0.5]

[0.6, 2]

Range

Model calibration

Constant > max {HT of migrants}

See footnote 91

10 years

MMP (smuggler fee + MX wage)

MMP (recent median)

MMP (recent median)

Source/ Explanation

Stay/ migrate % village in U.S. state j Stay/ migrate Stay/ migrate # Apprehensions # Apprehensions

Carrion & Sorensen (2006)

Gathmann (2004)

Genicot & Senesky (2004)

Hanson & Spilimbergo (2001)

Hanson & Spilimbergo (1999)

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Bivariate vector autoregression

OLS/ IV (Levels & First Differences)

Cox's proportional hazard model

Linear probability model (OLS/ IV)

Nested logit model (OLS/ IV)

Linear probability model (OLS/ IV)

Specification Econometric model

Dependent variable

Angelucci (2005)

Study

240

342

12,458

71,060

17,631

19771996

19681996

19671995

19781998

19762001

19721993

Data Period

135,260

Size

No

Yes (less)

No

No

Mixed

Yes (less) 1.09

-0.46

0.53

-0.12

1.29

0.01

0.61

---

-0.01 -0.08 -0.02

-0.93

-0.03 -0.01 -0.01

---

Defense budget, elections, other

---

Nominal budget of DEA

Border states' political clout

Real DEA budget

U.S. Border Enforcement Coefficient Instruments Significant? Min Max Best

[-0.53, -1.292]

[-0.43, -3.05]

[- 0.65, -1.26]

Deterrence elasticity

Table C.9: Econometric Studies on the Effect of U.S. Border Enforcement on the Flow of Mexican Migration

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

178

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Duration in Mexico

# Apprehensions Stay/ migrate % village that migrates

Kossoudji (1992)

Orrenius& Zavodny (2003)

Reyes et al. (2002)

Richter et al. (2005)

335

1,860

Size

1,759

19802002 Yes (more)

No

Mixed

Max

0.17 (a)

0.35 (a)

-0.49 0.58*

Significant? Min Yes (more)

Coefficient

0.01

0.24 (a)

---

Best 0.36

---

---

Defense budget, past enforcement

---

Instruments

U.S. Border Enforcement

+3.77

[-0.58, 0.49]

Deterrence elasticity

Notes: Column 5 reports if border enforcement’s coefficient was significant at 5% level and caused more or less migration. Studies use line-watch hours of U.S. Border Patrol agents if not otherwise stated. Carrion & Sorensen use line-watch hours divided by the mileage of sector's borderline (in logs). Kossoudji uses number of apprehensions. Richter et al. use change in Border Patrol budget. Columns 7 to 9 report border enforcements’ estimated coefficient (may not be comparable across specifications); negative signs indicate deterrence (except for Hanson & Spilimbergo). Other independent variables are not shown. (a) Sample includes only repeat migrants. (b) Reported results are from second stage.

Dynamic model (OLS)

19701998

19691996

19741979

Period

Data

2-stage model: logit 11,241/ (1st) & OLS (2nd) 28

OLS/ IV (first differences)

Duration model

Specification Dependent Econometric model variable

Study

Table C.9 Cont.: Econometric Studies on the Effect of U.S. Border Enforcement on the Flow of Mexican Migration

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

180

Appendix C

Table C.10: Determinants of Mexican Migration Behavior: Descriptive Statistics

Baseline independent Variables

Dependent variables

Variable IMI: Index of Migration Intensity

0.0

0.4

-0.3

0.9

ME: Dummy if person starts new illegal U.S. trip in year t

0.0

0.2

0.0

1.0

2.9

1.9

1.2

9.8

4.2

3.2

0.3

45.9

1.2

0.3

0.5

3.8

0.1

0.3

0.0

2.0

0.2

0.2

0.0

1.0

38.8

35.0

4.5

131.8

-0.1

0.1

-0.3

0.5

0.2

0.4

0.0

1.0

-0.5

0.5

-1.0

1.0

0.1

0.3

0.0

1.0

0.2

0.4

0.0

1.0

1.5

0.6

0.8

3.7

Line-watch hours of U.S. Border Patrol agents at the Southern U.S. Border (*) Predicted monthly wages in Mexico (in constant pesos of 2000, thousands) Predicted monthly wages in U.S. (in constant dollars of 2000, thousands) # of parents of migrant that have gone to U.S. before year t Number of green cards available to Mexicans (*) Yearly change in Mexican CPI

Total enforcement hours of U.S. Border Patrol agents by year and Mexican state (*) (see Appendix A) Index of real earnings in Mexican manufacturing Weekly earnings in U.S. construction (in constant 2000 dollars)

Alternative visas

Regional enforc.

Proportion of time spend in U.S. in 5 years surrounding year t House index: house in Mexico (-1), no house (0), house in U.S. (1) Proportion of dependents that have gone to U.S. in last 3 years Dummy of long-term U.S. immigration documentation

Alt. wage

IMI indicators

Average of community members' predicted IMI in years t-2 to t-6

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Mean SD Min Max

Non-immigrant visas granted to Mexicans (*) Temporary worker permits granted to Mexicans (thousands)

116.3 31.1

72.3 164.0

599.2 19.7 568.2 644.8 1.5

0.9

0.4

4.5

18.5

28.6

0.0

136.6

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Appendix C

181

Min Max

Additional individuallevel controls

Age of individual 38.2 Sex of individual (1= male; 2 = female) 1.2 Dummy if individual is married 0.7 Dummy: 6-8 years of education 0.3 Dummy: 9-11 years of education 0.1 Dummy:12-14 years of education 0.1 Dummy: >14 years of education 0.1 Hectares of land owned 2.1 Number of businesses owned 0.2

16.0 65.0 1.0 2.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 3600.0 0.0 4.0

Dummy if U.S. president was 0.7 0.5 0.0 Republican in past period Dummy if majority in Senate was 0.4 0.5 0.0 Republicans in past period Dummy if majority in House of Repre0.2 0.4 0.0 sentatives was Republican in year t-1 DEA Budget: Staffing & 676.1 341.0 237.7 Appropriations, all sources (constant 1995 dollars) (*) Number of Representatives from Southern border states in U.S. House of 77.6 6.9 73.0 Representatives in year t Highest within-party seniority of senators from Southern border states in 23.9 7.0 11.0 year t Number of Representatives from Southern border states in House 9.7 2.9 5.0 appropriations committee in year t Number of Representatives from 7.6 1.4 6.0 Southern border states in House judiciary committee in year t

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Add. macro controls

Variable Mean SD Growth in Mexican real per capita GDP 0.0 0.0 (index) Unemployment rate in Mexico (in 0.0 0.0 percent) Unemployment rate in U.S. (in percent) 6.6 1.3 Dummy if year is subsequent to IRCA 0.7 0.5

Instruments for border enforcement

Table C.10 Cont.: Determinants of Mexican Migration Behavior: Descriptive Statistics

13.5 0.4 0.5 0.4 0.3 0.3 0.2 36.8 0.4

-0.1

0.1

0.0

0.1

4.0 0.0

9.7 1.0

1.0 1.0 1.0 1645.8

96.0

38.0

14.0

10.0

Notes: Statistics are based on ME panel, but are similar for IMI panel. (*) Measured in millions.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

182

Appendix C

Table C.11: First Stage Regression Results of Extensive Migration Behavior First stage coefficient

Variable Corresponds to table 5.1, column

1 2 3 4 Political Political DEA Political majority majority budget clout

President Republican (past period)

0.3885

Senate majority Republican (past period)

0.0014

-0.1393

House majority Republican (past period)

4.3065

4.1216

***

***

0.4698

***

DEA Budget (const. 2000 Dollars, millions)

0.0051 ***

Senate seniority (of senior

-0.1250

senator from border states)

***

Size of house Delegation

0.7508

(from border states)

***

# Members in House appropriations Committee (from border states)

-0.1894

# Members in House judiciary

0.3109

committee (from border states) Observations

346,545 346,545 346,545 290,392

Adj. R-squared w/o instruments

0.26

0.12

0.12

0.10

Adj. R-squared w/ instruments

0.76

0.78

0.80

0.05

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Notes: ***", "**" significant at the 1% and 5% level, respectively. The dependent variable is line-watch hours. Regressions include all second stage regressors.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Appendix C

183

Table C.12: Intensive Migration Behavior: Relative Magnitudes of Coefficients Coefficient from

Independent variable is IMI

% UStime

U.S. home

Networks

Visa

Predicted Mexican wages

-0.10

-0.22

-0.23

0.04

-0.18

Predicted U.S. wages

2.66

8.16

5.08

0.30

0.74

Migration networks

7.44

4.41

3.92

5.16

5.70

Green Cards

5.94

7.55

6.88

5.08

7.05

Mexican inflation

0.03

0.04

0.03

0.02

0.03

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Notes: Table shows magnitudes of coefficients, taken from column 4 of Table 5.3 and columns 4 to 7 of Table 5.4, relative to border enforcement’s coefficient from same tables.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

0.0790

Green cards (millions granted to Mexicans) 0.0071

0.0318 ***

Migration networks (# of parents ever in US)

Mexican inflation (annual change in CPI * 10^3 )

0.0200 ***

Predicted U.S. wages (const. 2000 Dollars, thousands)

0.0038 ***

-0.0021 ***

 

2

3

 

   

0.0048

3.2705

0.0323 ***

0.0153 ***

-0.0022 ***

0.0051 **

  -0.0020

-4.2288

0.0336 ***

0.0174 **

-0.0030 ***

0.0015 *

Extensive migration behavior 15a Visas Recall

1

Predicted Mexican wages (const. 2000 Pesos, thousands)

Border enforcement squared (millions of line-watch hours)

Border enforcement (millions of line-watch hours)

Variable

0.4561

0.1207 ***

0.1677 ***

0.1632 ***

-0.0030

-0.0115 ***

0.1504 ***

Enforce ment

4

 

5

6

7

0.3764

0.1590 ***

0.2014 ***

0.0859 **

-0.0020

0.0420 ***

 

0.5624

0.1639 ***

0.2032 ***

0.0684 **

-0.0036

0.0411 ***

 

 

0.1810

0.1074 ***

0.1450 ***

0.0455

-0.0050 **

0.0179 ***

Intensive migration behavior 15a Visas Recall

Coefficient

Table C.13: Determinants of Mexican Migration Behavior: Additional Specification Tests

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

 

 

0.3838 *

0.0929 ***

0.1381 ***

0.0709 ***

-0.0026

0.0112 ***

No Weights

8

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

2

NO 346,545 0.58 N/A

0.33 N/A

 

 

0.18 N/A

YES 232,701

YES

N/A 16.83%

NO 24,948

YES

Enforce ment

4

5

6

7

N/A 21.68%

YES 24,948

YES

-0.0422 **

   

NO 24,948 N/A 22.78%

YES

-1.2727

   

YES 17,664 N/A 8.22%

YES

Intensive migration behavior 15a Visas Recall

Coefficient

YES 24,948 N/A 6.83%

YES

No Weights

8

Notes: ***", "**", "*" significant at the 1%, 5% and 10 % level, respectively. Standard errors in parentheses are robust to clustering at year level. Regional enforcement is millions of total enforcement hours per year and Mexican state (see Appendix A). Columns 3 and 7 include a maximum of 15 years of recall. Column 8 uses no MMP sampling weights. (#) Normalized change in IMI: change in IMI due to standard deviation increase in border enforcement divided by IMI’s standard deviation. All IV regressions use political majority instruments.

Enforcement Elasticity ∆ IMI if ∆-SD in LWH (#)

YES

-0.2268 * YES 346,545

 

YES

-0.0053 *

LWH instrumented

Observations

3

Extensive migration behavior 15a Visas Recall

1

Individual dummies

Labor Contract Visas (millions granted to Mexicans)

Non-Immigrant Visas (millions granted to Mexicans)

Variable

Table C.13 Cont.: Determinants of Mexican Migration Behavior: Additional Specification Tests

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

186

Appendix C

II) ADDITIONAL FIGURES Figure C.1: Robustness of the IMI to Changes in Extensive Migration Behavior 0.6 0.4 0.2 0 -0.2 1966

1971

1976

1981

1986

1991

1996

2001

-0.4

Only Migrants Active in Current Period

Active and Inactive Migrants

Notes: Figure shows yearly average of IMI using two different panels (see Appendix D-I for details).

Figure C.2: Simulation of Optimal Migration Strategy Wages & Migration Costs Preferences W ages &Mig ration C osts Prefernces I

2 4  ww* 6 w* 8 w 10

2

RM I IM 10 8 6 4 M

RM IM M 40 30 A 20

0.8 0.6 δ 10

0.4 04

Prefernces IIII Preferences RM

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

40 30 A20 10

IM 0.4

0.8 0.6 α

Notes: Graphs shows whether repeat migration (“IM”) or return migration (“RM”) strategy is the utility-maximal migration strategy. Light area marks where Condition 1 holds. All parameters are equal to their default values.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Appendix C

187

Figure C.3: Simulation of Condition 1 and 2 Condition 1: Internal Migration versus Return Migration Baseline scenario I 400 2 200 0 40 30 20A

0.4 .4 0.6

d δ

0.8

10

Baseline scenario II

0.4 .4 0.6

400 200 2 10 0

δd 0.8

8 6 M Mx €€ 4 €€€€€€ * ww* 2

WCS 400 200 0 0.4 .4 0.6

δd 0.8

5

20 15 10M

Condition 2 Baseline scenario I 1 0.75 0.5 0.25 0 0.8 A0.6

0.4

2 1.8 1. 1.6 w* w*  1.4 € €€€€€€€ w 1.2 w 

WCS I

0.8 d 0.6

WSC II (wage-ratio of 1.5)

0.8

0.6 α a δ 0.4 0.4 04

1 0.75 0.5 0.25 0

0.8 δd0.6 0.4

0.4 04

0.8 0.6 αa

1 0.75 0.5 0.25 0

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Notes: Graphs in top and lower panel show if Conditions 1 or 2 hold, respectively. Light area marks where the respective condition fails. In baseline scenario all parameters are set equal to default values. In worst-case scenario (“WCS”) parameters are set so as to make repeat migration most unlikely.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

0.4

0.4

2 1.8 1. 1.6w*  w* €€€€€€€€ 1.4 € w  w 1.2 12

2 1.5 1 0.5 0 0.8 α a 0.6 0.4

2

WCS (A2c)

€€€€€€€€ w  w

3 w*  w*

4

4 * 3 ww* €€€€€€€w € 2 w

0.4

0.8 dδ 0.6 0.6aα

0.6 0.8 δd

WCS (A2d)

0.4

2 1.5 1 0.5 0 0.4

1 0.75 0.5 0.25 0 0.8

WCS II (wage ratio of 1.5)

Notes: Graphs show if Conditions A2b-d) hold. Light gray marks area where the respective condition fails. In baseline scenario all parameters equal default values. In worst-case scenario (“WCS”) parameters are set so as to make internal migration most unlikely.

2 1.5 1 0.5 0 0.8 α a 0.6

1 0.75 0.5 0.25 0 0.8 0.6aα

WCS I

A2b

0.4 A2c (left, middle) and A2d (right)

2 0.8 1.8 1. 1.6 w* δd 0.6 w* 1.4 €€€€€€€€ w  1.2 w 0.4

Baseline scenario (A2c)

1 0.75 0.5 0.250 0.8 a 0.6 α

Baseline Scenario

Figure C.4: Simulation of Additional Conditions

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

APPENDIX D

Additional Specifications and Robustness Tests

I) ROBUSTNESS OF THE IMI

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Cross-Sectional IMI

To ascertain that the cross-sectional IMI is structurally stable, I perform additional factor analyses and compare the results to results of Table 3.3. Table C.7 confirms the structural stability of the IMI: factor patterns are very similar (one dominant principal component), few observations have strongly differing factor scores (see second row), and Spearman rank correlation coefficients are high (mostly above 0.9). When I apply the specification of the panel IMI, which excludes indicators for remittances and English knowledge, to the cross-section, I find great accordance to the cross-sectional IMI (see column 1). Both PCA results favor the use of only one principal component, the Spearman rank correlation coefficient is 0.94, and 96 percent of the observations have a less than one standard deviation difference in factor scores. I take this as an indication that the cross-sectional IMI and panel IMI are good substitutes. Replacing the indicators out of which the cross-sectional IMI is formed by alternative indicators does only cause minor changes. Even if all indicators are replaced simultaneously, a very drastic modification, the PCA results remain largely unchanged, as indicated by a 0.87 Spearman rank correlation coefficient (see column 3). If indicators are replaced individually the changes are even less (not

189

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

190

Appendix D

shown; results available from author). If I enlarge the set of indicators to nine, the sample rankings remain also very comparable (see column 4). Strictly seen, PCA is only mathematically exact for continuous variables (Basilevsky 1994). Hence, I replace all non-continuous variables by continuous substitutes or drop them, if no substitute is available. I find little divergence (see column 2). Using different factoring techniques, instead of principal components factoring, such as maximum likelihood or principal axis factoring, has almost no effect (see columns 5 and 6, respectively). It could be that the variation in the IMI is driven by changes in extensive rather than intensive migration behavior. For example, individuals may enter and exit the migration flow without my panel adjusting for these changes.159 The pertinent question is, in how far do these variations in extensive migration behavior affect the IMI? Figure C.1 in Appendix C depicts the panel IMI against an alternative panel version, in which only active migrants are included, that is, only individual-year observations of migrants who reside in the U.S. in the current period. Despite the considerable difference in the two panels, both indexes progress almost identical over time (the Pearson correlation coefficient is 0.94). Since the active-migrant panel is by construction immune to compositional changes in extensive migration behavior, I take this as an indication that the Panel IMI is robust to changes in extensive migration behavior.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Panel IMI

In the MMP database the sampled migrants and communities vary nonrandomly over time. Hence, the panel IMI may be affected by sample selection. In order to assess this, I split my panel into several subpanels, perform independent PCAs on each, and compare the results. First, I group individual-year observations into four cohorts according to their community’s survey year. I find that the factor score coefficients of the first principal component (i.e., the “weights”) change only little across cohorts (not shown; results available from author). Secondly, I conduct a thought experiment, trying to determine what would happen if the intensive migration behavior of a new-millennium 159

Irrespective of when an individual migrates for the first or last time, I use his full working-age life-time history in the standard panel.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Appendix D

191

migrant would be observed in the 1980s or 1960s: I compare the factor scores and ranking of the 2000s sample, when using the “correct” weights from the 2000s, with its factor scores and ranking, when using the weights from the 1960s or 1980s (see columns 7 and 8 of Table C.7). I find almost no differences. II) ADDITIONAL SIMULATIONS FROM CHAPTER FOUR Simulation of the Optimal Migration Strategy (Stage 2) Figure C.2 in Appendix C shows that whenever Condition 1 holds, the light gray area, the internal migration strategy as defined in Table 4.1 dominates return migration (“RM,” n=1). Since (nIM, t IM) is the unique extreme point of equations (4-3a) and (4-3b), it constitutes the optimal migration strategy (if Conditions 1 and 2 hold). If Condition 1 fails, the dark gray area, the internal migration strategy is not valid and return migration dominates.

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Simulation of Conditions 1, 2, and A2 I simulate Conditions 1 and 2, and the additional Conditions A2. I start by de-picting the “baseline scenario” (“BS”), in which all parameters are set equal to their default values. Then, I twist parameters within their ranges, as given in Table C.8 in Appendix C, so as to make internal migration most unlikely, i.e. the “worst-case scenario” (“WCS”). Areas in light gray mark where the condition under consideration fails. Condition 1 verifies whether internal migration or return migration is valid and optimal. Migration costs relative to the U.S. wage, the comfort of living abroad, and the preference for consumption are most important for this calculation (see Figure C.3 in Appendix C). Condition 1 holds in the baseline scenario. However, if either A, δ, or M becomes large, return migration dominates. Return migration strongly dominates in the worst-case scenario. Condition 2 insures that the total time in Mexico, nt, does not exceed the total time endowment T. Figure C.3 confirms that Condition 2 is mainly determined by the wage ratio. As long as the wage ratio is above 1.5, which is very likely for the U.S. – Mexican case, Condition 2 holds.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

192

Appendix D

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

I further simulate the additional Conditions A2, needed to establish the signs of the partial derivatives of the repeat migration strategy. Condition A2-b) checks whether higher migration costs prolong average stays in the U.S. As shown in Figure C.4 in Appendix C, Condition A2-b) is virtually identical to Condition 2 and, thus not an additional constraint. Remittances rise with U.S. wages if Condition A2c) is fulfilled, which is true as long as neither the wage ratio is very low, nor α very high. Higher U.S. wages shorten U.S. trips if Conditions A2-d) and A2-c) hold, which is very likely according to Figure C.4.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

References

Acosta, Pablo, Cesar Calderón, Pablo Fajnzylber, and Humberto Lopez (2008). “What is the Impact of International Remittances on Poverty and Inequality in Latin America?,” World Development , 36(1). 89-114. Adelman, Irma, Edward Taylor, and Stephen Vogel (1988). "Life in a Mexican Village: A SAM Perspective," Journal of Development Studies, 25(5). 5 24. Adler, Rachel (2002). “Patron-Client Ties, Ethnic Entrepreneurship and Transnational Migration: The Case of Yucatecans in Dallas, Texas,” Urban Anthropology, 31, 129-161. Aghion, Philippe, Diego Comin, Peter Howitt, and Isabel Tecu (2009). “When Does Domestic Saving Matter for Economic Growth?,” Working Paper 09-80, Boston, MA: Havard Business School. Aguilera, Michael, and Douglas Massey (2003). "Social Capital and the Wages of Mexican Migrants: New Hypotheses and Tests," Social Forces, 82(2). 671-701. Akerlof, George (1970). "The Market for 'Lemons': Quality Uncertainty and the Market Mechanism", Quarterly Journal of Economics, 84 (3). 488-500. Alarcón, R (1992). “El Proceso de Nortenicazacion,” in Thomas Calvo and Gustavo Lopez (eds) Movimientos de la Populacion en el Occidente de Mexico, Zamorah, MX: El Colegio de Michoacan. Alba, Richard and Victor Nee (1997). "Rethinking Assimilation Theory in a New Era of Immigration." International Migration Review, 31 (4), 826874. Alwin, Duane (1973). "The Use of Factor Analysis in the Construction of Linear Composites in Social Research," Sociological Methods and Research, 2, 191-214.

193

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

194

References

Amuedo-Dorantes, Catalina, Cynthia Bansak, and Susan Pozo (2005). “On the Remitting Patterns of Immigrants: Evidence from Mexican Survey Data,” Federal Reserve Bank of Atlanta, Economic Review, 1, 37-58. Amuedo-Dorantes, Catalina and Mundra Kusum (2007). “Social Networks and Their Impact on the Earnings of Mexican Migrants,” Demography, 44 (4), 849-863. Amuedo-Dorantes, Catalina and Francesca Mazzolari (2010). “Remittances to Latin America from Migrants in the United States: Assessing the Impact of Amnesty Programs,” Journal of Development Economics, 91, 323–335. Andres, Peter (2001). “The Transformation of Migrant Smuggling Across the U.S.-Mexican Border,” in David Kyle and Rey Koslowski (eds.) Global Human Smuggling: Comparative Perspectives, Baltimore, MD: John Hopkins University Press. Angelucci, Manuela (2005). “U.S. Border Enforcement and the Net Flow of Mexican Illegal Migration,” IZA Discussion Papers No. 1642. Avila, Jose Luis, Carlos Fuentes, and Rodolfo Tuiran (1998). “Migracion Temporal de Adolescentes y Jovenes, 1993 – 1997,” Mexico, DF: Consejo Nacional de Poblacion. Baines, Dudley (1995). Emigration from Europe 1815 - 1930, Cambridge, UK: Cambridge University Press. Banerjee, Biswajit (1984). “Information Flow Expectations and Job Search,” Journal of Development Economics, 15, 239-257. Banerjee, Biswajit (1990). “The Determinants of Migrating with a PreArranged Job and of the Initial Duration of Unemployment,” Journal of Development Economics, 36, 337-351. Basilevsky, Alexander (1994). Statistical Factor Analysis and Related Methods: Theory and Applications, New York, NY: Wiley. Basok, Tanja (2007). Canada's Temporary Migration Program: A Model Despite Flaws, Washington, DC: Migration Policy Institute. Bastida, Elena (2001). “Kinship Ties of Mexican Migrant Women on the United States/Mexico Border,” Journal of Comparative Family Studies, 32 (4), 549 -70. Bauer, Thomas, Gil Epstein, and Ira Gang (2005). "Enclaves, Language and the Location Choice of Migrants," Journal of Population Economics, 18 (4), 649–662.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

References

195

Bean, Frank, Rodolfo Corona, Rodolfo Tuirán, and Karen Woodrow-Lafield (1997). “The Quantification of Migration between Mexico and the United States,” in U.S. Commission on Immigration Reform (ed) Migration Between Mexico and the United States: Binational Study. Washington, DC: U.S. Commission on Immigration Reform. Beine, Michel, Frederic Docquier and Hillel Rapoport (2001). “Brain Drain and Economic Growth: Theory and Evidence,” Journal of Development Economics, 64 (1), 275-89. Behrman, Jere, Hans-Peter Kohler, and Susan Cott Watkins (2002). “Social Networks and Changes in Contraceptive Use over Time: Evidence from a Longitudinal Study in Rural Kenya,” Demography, 39(4), 713-738. Bhagwati, Jagdish (1984). "Incentives and Disincentives: International Migration," Weltwirtschaftliches Archiv, 12(4), 678–701. Bowles, Samuel (2004). Microeconomics: Behavior, Institutions, and Evolution, Princeton, NJ: Princeton University Press. Borjas, George (1994). “The Economics of Immigration,” Journal of Economic Literature, 32, 1667–1717. Borjas, George (2006). “Making it in America: Social Mobility in the Immigrant Population,” NBER Working Papers No. 12088. Borjas, George (2009). “The Analytics of the Wage Effect of Immigration,” NBER Working Papers No. 14796. Bourdieu, Pierre (1986). “The Forms of Capital,” in John Richardson (ed.) Handbook of Theory and Research for the Sociology of Education, New York, NY: Academic Press. Boyce, James, Andrew Klemer, Paul Templet, and Cleve Willis (1999). “Power Distribution, the Environment, and Public Health: A State-Level Analysis,” Ecological Economics, 29, 127-140. Boyd, Monika (1989). "Family and Personal Networks in International Migration: Recent Developments and New Agendas," International Migration Review, 12 (3). 638-670. Brock, William and Steven Durlauf (2001), “Interactions-Based Models,” in James Heckman and Edward Leamer (eds.) Handbook of Econometrics, Volume 5, Amsterdam: Elsevier. Brownell, Peter (2005). “The Declining Enforcement of Employer Sanctions,” Washington, DC: Migration Information Source. Brücker, Herbert, and Philipp Schröder (2005). “International Migration with Heterogeneous Agents: Theory and Evidence,” IZA Discussion Paper No. 2049.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

196

References

Bruhn, Miriam, and Inessa Love (2009). “The Economic Impact of Banking the Unbanked: Evidence from Mexico,” Policy Research Working Paper No. 4981, Washington, DC: World Bank. Bustamante, Jorge A., Guillermina Jasso, J. Edward Taylor and Paz Trigueros Legarreta (1998). "Characteristics of Migrants: Mexicans in the United States," in U.S. Commission on Immigration Reform (ed.) Migration Between Mexico and the United States: Binational Study, Washington, DC: U.S. Commission on Immigration Reform. Caces, Fe (1986). “Immigrant Recruitment into Labor Force: Social Networks among Filipinos in Hawaii” Amerasia Journal, 13, 23-38. Card, David (2009), “Immigration and Inequality,” American Economic Review, 99 (2), 1‐21. Carrington, William and Enrica Detragiache (1999). “How Extensive Is the Brain Drain?” Finance and Development, 36(2), 46–49. Carrington, William, Enrica Detragiache, and Tara Vishwanath (1996) “Migration with Endogenous Moving Costs”, American Economic Review, 86 (4): 909-30. Carrión-Flores, Carmen and Todd Sorensen (2006). “The Effects of Border Enforcement on Migrants Border Crossing Choices: Diversion, or Deterrence?,” available at: http://client.norc.org/jole/soleweb/7195.pdf (accessed on August 6, 2010) CBO (2006). Cost Estimate: S. 2611 - Comprehensive Immigration Reform Act of 2006, Washington, DC: Congressional Budget Office. CBO (2007). The Impacts of Unauthorized Immigrants on the Budgets of State and Local Governments, Washington, DC: Congressional Budget Office. Cerrutti, Marcela, and Douglas Massey (2001). “On the Auspices of Female Migration from Mexico to the United States.” Demography, 38, 187-200. Chamberlain, Gary (1980). "Analysis of Covariance with Qualitative Data," Review of Economic Studies, 47(1), 225-38. Chavez Blancarte, Marina ( 2003). "Enfermarse y Curarse en Chicago: Las Practices de Autoatencion," in Lopez Castro (ed) Diáspora Michoacana, Zamora, Mexico: El Collegio de Michoacan. Chavez, Leo (1990). "Coresidence and Resistance: Strategies for Survival among Undocumented Mexicans and Central Americans in the US," Urban Anthropology, 19(1-2), 31-61. Chavez, Leo (1992). Shadowed Lives: Undocumented Immigrants in American Society, Fort Worth, TX: Harcourt Brace College Publishers. Chiswick, Carmel (2009). “The Economic Determinants of Ethnic Assimilation,” Journal of Population Economics 22 (4), 859-880.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

References

197

Choldin, Harvey (1973). “Kinship networks in the migration Process,” International Migration Review, 7(2), 163 – 175. Cohen, Jeffrey, Alicia Sylvia Gijón-Cruz, Rafael G. Reyes-Morales and Garry Chick (2003). “Understanding Transnational Processes: Modeling Migration Outcomes in the Central Valleys of Oaxaca, Mexico,” Field Methods, 15 (4), 366-385. Coleman, James (1988). "Social Capital in the Creation of Human Capital", American Journal of Sociology, 94, 95-120. Colussi, Aldo (2004). “An Estimable Model of Illegal Mexican Immigration,” Ph.D. Dissertation, Philadelphia, PA: University of Pennsylvania, Department of Economics CONAPO (2000). Índice de Intensidad Migratoria México-Estados Unidos, 2000, Mexico, D.F.: Consejo Nacional de Población. Conning, Jonathan and Christopher Udry (2005). “Rural Financial Markets In Developing Countries,“ in Robert Evenson and Prabhu Pingali (eds.) Handbook Of Agricultural Economics, Vol. 3, Amsterdam: Elsevier. Constant, Amelis and Klaus Zimmermann (2003). "The Dynamics of Repeat Migration: A Markov Chain Analysis," Center for Comparative Immigration Studies Working Paper No. 85. Cornelius, Wayne (1989). “Impacts of the 1986 U.S. Immigration Law on Emigration from Rural Mexican Sending Communities,” Population and Development Review, 15(4), 689-705. Cornelius, Wayne (1992). “From Sojourners to Settlers: The Changing Profile of Mexican Migration to the United States,“ in J. Bustamante, C. Reynolds, and R. Hinojosa-Ojeda (eds) U.S.-Mexico Relations: Labor Market Interdependence, Stanford, CA: Stanford University Press. Cornelius, Wayne (2001). “Death at the Border: Efficacy and Unintended Consequences of US Immigration Control Policy,” Population and Development Review, 27, 661–685. Cornelius,Wayne (2005). “Controlling ‘Unwanted’ Immigration: Lessons from the United States, 1993–2004,” Journal of Ethnic and Migration Studies, 31, 775–794. Cornelius, Wayne and Jessa Lewis (2007), “Impacts of Border Enforcement on Mexican Migration: The View From Sending Communities,” La Jolla, CA: Center for Comparative Immigration Studies. Cornelius,Wayne and Marc Rosenblum (2005). “Immigration and Politics,” Annual Review of Political Science, 8, 99–119. Cornelius, Wayne and Idean Salehyan (2007). “Does Border Enforcement Deter Unauthorized Immigration? The case of Mexican Migration to the United States of America,” Regulation & Governance, 1, 139–153.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

198

References

CRS (2006a). Immigration: Policy Considerations Related to Guest Worker Programs, Washington, DC: Congressional Research Service. Cuecuecha, Alfredo (2008). Mexican Migration to the US: Evidence and Theory, Saarbrücken, Germany: Vdm Verlag Dr. Müller. Dasgupta, Partha (2003). “Social Capital and Economic Performance: Analytics,” in E. Ostrom and T.K. Ahn (eds.) Foundations of Social Capital, Cheltenham, UK: Edward Elgar. Da Vanzo, Julie (1983). "Repeat Migration in the United States: Who Moves Back and Who Moves On," The Review of Economics and Statistics, 64(4), 552-559. Davila, Alberto, Jose Pagan, and Gokce Soydemir (2001). “The Short-Term and Long-Term Deterrence Effects of INS Border and Interior Enforcement on Undocumented Immigration,” Journal of Economic Behavior and Organization, 49(4), 459-472. Davis, Benjamin and Paul Winters (2001). “Gender, Networks and Mexico–US Migration,” Journal of Development Economics, 38 (2), 1-26. Davis, Benjamin and Paul Winters (2002). “Domestic and International Migration from Rural Mexico: Disaggregating the Effects of Network Structure and Composition,” Population Studies, 56, 291-309. De Jong, Gordon and Robert Gardener (1981). “Migration Decision Making: Multidisciplinary Approaches to Micro-level Studies of Migration in Developed and Developing Countries,” New York, NY: Pergamon Press. De Laat, Joost (2005). “Moral Hazard and Costly Monitoring: The Case of Split Migrants in Kenya,” paper presented at Northeastern University Development Conference 2005. Delgado, Raul and Luis Eduardo Guarnizo (2007). “Migration and Development: Lessons from the Mexican Experience,” Feature Story, Washington, DC: Migration Policy Institute. DHS (2010). Estimates of the Unauthorized Immigrant Population Residing in the United States: January 2009, Washington, DC: U.S. Department of Homeland Security. Djajic, Slobodan and Ross Milbourne (1988). "A General Equilibrium Model of Guest-Worker Migration: A Source-Country Perspective,” Journal of International Economics, 25, 335–351. Dolfin, Sarah and Garance Genicot (2010). “What Do Networks Do? The Role of Networks on Migration and ‘Coyote’ Use,” Review of Development Economics, 14 (2), 343-359. Donato, Katharine, Jorge Durand, and Douglas Massey (1992). “Stemming the Tide? Assessing the Deterrent Effects of the Immigration Reform and Control Act,” Demography, 29(2), 139-57.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

References

199

Donato, Katharine, Jonathan Hiskey, Jorge Durand and Douglas Massey (2010). “Continental Divides: International Migration in the Americas, ”The ANNALS of the American Academy of Political and Social Science Series, Sage Publications: Thousand Oaks, CA. Durand, Jorge (1994). Mas Alla de la Linea, Mexico, DF: Consejo Nacional Para la Cultura y las Artes. Durand, Jorge, William Kandel, Emilio Parrado, and Douglas Massey (1996b). "International Migration and Development in Mexican Communities," Demography, 33(2), 249-64. Durand, Jorge, Emilio A. Parrado and Douglas Massey (1996a). “Migradollars and Development: A Reconsideration of the Mexican Case,” International Migration Review, 30(2), 423-444. Durlauf, Steven and Marcel Fafchamps (2005). “Social Capital,” in Philippe Aghion and Steven Durlauf (eds.), Handbook of Economic Growth (Volume 1), Amsterdam: Elsevier. Dustmann, Christian (2000). “Temporary Migration and Economic Assimilation,” Swedish Economic Policy Review, 7, 213-244. Dustmann, Christian (2001). “Why Go Back? Return Motives of Migrant Workers,” in Slobodan Djajic (ed) International Migration: Trends, Policy and Economic Impact, London, NY: Routledge. Dustmann, Christian and Josep Mestres (2010). “Remittances and Temporary Migration,” Journal of Development Economics, 92 (1), 62-70. EMIF (2006). Encuesta sobre Migración en la Frontera Norte de México – Procedentes del Sur 1996 – 2003, Mexico, DF: El Consejo Nacional de Población. Epstein, Gil (2002). "Informational Cascades and Decision to Migrate," IZA Working Paper No. 445. Epstein, Gil and Ira Gang (2006), “The Influence of Others on Migration Plans,” Review of Development Economics, 10 (4), 652-665. Eschbach, Karl, Jacqueline Hagan, and Nestor Rodríguez (2003). "Deaths During Undocumented Migration: Trends and Policy Implications in the New Era of Homeland Security," In Defense of the Alien, 26, 37-52. Escobar Latapí, Agustín, and Susan Martin (2006). Mexico – U.S. Migration Management: A Binational Approach, Executive Summary, Washington, DC: Georgetown University. Espenshade, Thomas (1990). “Undocumented Migration to the United States: Evidence from a Repeated Trials Model," in Frank Bean, B. Edmonston, and J. Passel (eds.) Undocumented Migration to the United States, Washington, D.C.: The Rand Corporation.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

200

References

Espenshade, Thomas (1994). "Does the Threat of Border Apprehension Deter Undocumented US Immigration?" Population and Development Review, 20(4), 871-892. Espinosa, Kristin and Douglas Massey (1997). “Undocumented Migration and the Quantity and Quality of Social Capital,” Soziale Welt, 12, 141-162. Espinosa, Kristin (1997). “Helping Hands: Social Capital and the Undocumented Migration of Mexican Men to the United States,” Ph.D. Dissertation, Chicago, IL: University of Chicago, Department of Sociology. Ethier, Wilfred (1986). “Illegal Immigration: The Host-Country Problem”. American Economic Review, 76(1), 56-71. Faini, Riccardo (2006). "Remittances and the Brain Drain," IZA Discussion Papers No. 2155. Fernández-Ruiz, Guillermo (2003). "Crónica Sincrónica de la Migracción Michoacana," in Gustavo Lopez-Castro (ed) Diáspora Michoacana, Zamora, Mexico: El Collegio de Michoacan. Filmer, Deon and Lant H. Pritchett (2001). “Estimating Wealth Effects without Expenditure Data or Tears: An Application to Educational Enrollments in States of India,” Demography, 38(1), 115-132. Fussell, Elizabeth (2004). “Sources of Mexico’s Migration Stream: Rural, Urban, and Border Migrants to the United States,” Social Forces, 82(3), 937–67. Gans, Herbert (1997). "Toward a Reconciliation of ‘Assimilation’ and ‘Pluralism’: The Interplay of Acculturation and Ethnic Retention," International Migration Review, 31(4), 875-92. GAO (2006). Illegal Immigration, GAO-06-770, Washington, DC: General Accounting Office. Garcia, Carlos (2005). “Buscando Trabajo: Social Networking among Immigrants from Mexico to the United States,” Hispanic Journal of Behavioral Science, 27 (2), 3-22. Gathmann, Christina (2004). "The Effects of Enforcement on Illegal Markets: Evidence from Migrant Smuggling along the Southwestern Border," IZA Working Paper No. 1004. Glick, Jennifer (1999). “Economic Support from and to Extended Kin: A Comparison of Mexican Americans and Mexican Immigrants,” International Migration Review, 33(3), 745-765. Glick Schiller, Nina and Peggy Levitt (2006). "Haven’t We Heard This Somewhere Before? A Substantive View of Transnational Migration Studies by Way of a Reply to Waldinger and Fitzgerald", CMD Working Paper # 06-01, Princeton, NJ: Princeton University.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

References

201

Goldring, Luin (1992). “Diversity and Community in Transnational Migration: A Comparative Study of two Mexico-U.S. Migrant Circuits,” Ph.D. Dissertation, Ithaca, N.Y.: Cornell University, Department of Rural Sociology. Goldring, Luin (1998). “The Power of Status in Transnational Social Spaces,” in Michael Smith and Luis Guarnizo (eds) Transnationalism From Below, New Brunswick, N.J.: Transaction Publishers. Goldring, Luin (2002). “The Mexican State and Transmigrant Organizations: Negotiating the Boundaries of Membership and Participation,” Latin American Research Review, 37(3), 55–99. Goldring, Luin (2004a). “Family and Collective Remittances to Mexico: A Multi-Dimensional Typology,” Development and Change, 35(4), 799– 840. Goldring, Luin (2004b), “Remesas y Microbancos,” Migración y Desarrollo, 3, 92-98. Gordon, Milton Myron (1964). Assimilation in American Life: The Role of Race, Religion, and National Origins, New York, NY: Oxford University Press. Granovetter, Mark (1973). “The Strength of Weak Ties,” American Journal of Sociology, 78, 1360-80. Graves, Nancy and Theodore Graves (1974). "Adaptive Strategies in Urban Migration," Annual Review of Anthropology, 3, 117-151. Greene, William (2005). “Econometric Analysis-Fifth Edition,” New York, NY: Pearson Education. Grieco, Elizabeth (2003). The Remittance Behavior of Immigrant Households: Micronesians in Hawaii and Guam, New York, NY: LFB Scholarly Publishing LLC. Hagan, Jacqueline, Karl Eschbach, and Nestor Rodriguez (2008). “U.S. Deportation Policy, Family Separation, and Circular Migration,” International Migration Review, 42 (1), 64-88. Handlin, Oscar (1973). The Uprooted, 2nd ed., Boston: Little, Brown and Company. Hanson, Gordon (2006). “Illegal Migration from Mexico to the United States,” Journal of Economic Literature, 44 (4), 869 - 924. Hanson, Gordon H., Raymond Robertson and Antonio Spilimbergo (2002) “Does Border Enforcement Protect U.S. Workers from Illegal Immigration?” The Review of Economics and Statistics, 84(1), 73-92. Hanson, Gordon and Antonio Spilimbergo (1999). “Illegal Immigration, Border Enforcement, and Relative Wages: Evidence from Apprehensions at the U.S.-Mexico Border,” The American Economic Review, 89(5), 1337-57.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

202

References

Hanson, Gordon, and Antonio Spilimbergo (2001). “Political Economy, Sectoral Shocks, and Border Enforcement,” The Canadian Journal of Economics, 34 (3), 612-638. Harris, John and Michael Todaro (1970). "Migration, Unemployment and Development: A Two-Sector Analysis," American Economic Review, 60(1), 126-42. Heckman, James 1979. “Sample Selection Bias as a Specification Error,” Econometrica, 47(1), 153-161. Herzog, Henry and Alan Schottman (1983). “Migration Information, Job Search and the Remigration Decision,” Southern Economic Journal, 50(1), 43–56. Hill, John (1987). "Immigrant Decisions Concerning Duration of Stay and Migratory Frequency," Journal of Development Economics, 23, 221 - 234. Hugo, Graeme (1981). "Village-Community Ties, Village Norms, and Ethnic and Social Networks: A Review of Evidence from the Third World," in Gordon de Jong and Robert Gardner (eds) Migration Decision Making: Multidisciplinary Approaches to Micro-Level Studies in Developed and Developing Countries, New York, NY: Pergamon Press. IFAD (2010). Facts and Figures, International Fund for Agricultural Development, www.ifad.org/remittances (accessed November 22, 2010). IOM (2005). World Migration 2005: Costs and Benefits, Geneva, Switzerland: International Organization for Migration. IOM (2006). “Migration and Development: Opportunities and Challenges for Policymakers,” IOM Migration Research Series 22, Geneva, Switzerland: International Organization for Migration. IOM (2007). Engaging Diasporas as Development Partners for Home and Destination Countries: Challenges for Policymakers,” IOM Migration Research Series 26, Geneva, Switzerland: International Organization for Migration. Jones, Richard (1982). “Channelization of Undocumented Mexican Migrants to the US”, Economic Geography, 58(2), 156-177. Katz, Eliakim and Oded Stark (1986). “Labor Migration and Risk Aversion in Less Developed Countries,” Journal of Labor Economics, 4(1), 134-149. Kearney, Michael (1986). “From the Invisible Hand to Visible Feet. Anthropological Studies of Migration and Development,” Annual Review of Anthropology, 15, 331-61. Kemper, Robert (1977). Migration and Adaptation: Tzintzuntzan Peasants in Mexico City, London, NY: Sage Publications. Kline, Paul (1994). An Easy Guide to Factor Analysis, New York, NY: Routledge.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

References

203

Kossoudji, Sherrie (1992). “Playing Cat and Mouse at the Mexican-American Border,” Demography, 29(2), 159-180. Lee, Ronald, and Timothy Miller (2000). “Immigration, Social Security, and Broader Fiscal Impacts,” American Economic Review, 90 (2), 350 - 54. Levy, Mildred and Walter Wadycki (1973). “The Influence of Family and Friends on Geographic Labor Migration: An Intercensal Comparison,” The Review of Economics and Statistics, 55, 198-203. Lindstrom, David (1996). "Economic Opportunity in Mexico and Return Migration from the United States," Demography 33(2), 357-374. Lomnitz, Larissa (1976). "Migration and Networks in Latin America," in Alejandro Portes and Harley Browning Current Perspectives in Latin American Urban Research, Austin, TX: University of Texas Press. Lomnitz, Larissa (1977). "Networks and Marginality: Life in a Mexican Shantytown," New York, NY: Academic Press. Lopez-Castro, Gustavo (1988). “La Migracion a Estados Unidos en Gomez Farias,” in Gustavo Lopez and Sergio Pardo Migracion en el Occidente de Mexico, Zamorah, Michoacan: El Colegio de Michoacan. Lopez Castro, Gustavo (1998). "Coyotes and Alien Smuggling," in U.S. Commission on Immigration Reform (ed.) Migration between Mexico and the United States: Binational Study, Washington, DC: U.S. Commission on Immigration Reform. Lopez-Castro, Gustavo (2003). Diáspora Michoacana, Zamora, Mexico: El Collegio de Michoacan. Lopez Castro, Gustavo and Sergio Pardo Galvan (1988). Migracion en el Occidente de Mexico, Zamorah, Michoacan: El Colegio de Michoacan. Lucas, Robert (2005). International Migration and Economic Development: Lessons from Low-Income Countries, Northampton, MA: Edward Elgar. Lucas, Robert and Oded Stark (1985). “Motivations to Remit: Evidence from Botswana,” Journal of Political Economy, 93, 901-918. Luke, Nancy, Mark Rosenzweig, and Kaivan Munshi (2004). ”Marriage, Networks, and Jobs in Third World Cities,” Journal of European Economics Association, 2(2-3), 437 – 446. MacDonald, John and Leatrice MacDonald (1974). "Chain Migration, Ethnic Neighborhood Formation, and Social Networks," in Charles Tilly (ed.) An Urban World, Boston: Little, Brown & Company. Manning, Robert (2000). Five Years after NAFTA Rhetoric and Reality of Mexican Immigration in the 21st Century, Washington, DC: Center for Immigration Studies. Manski, Charles, (1995). Identification Problems in the Social Sciences, Cambridge, MA: Harvard University Press.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

204

References

Marsden, Peter (1999). “Network Data and Measurement,” Annual Review of Sociology, 16, 435 – 463. Martin, Philip (2001). “There Is Nothing More Permanent than Temporary Foreign Workers,” Washington, DC: Centre for Immigration Studies. Martin, Jack (2007). The Costs of Illegal Immigration to Illinoisans, Washington, DC: Federation of American Immigration Reform. Massey, Douglas, Rafael Alarcón, Jorge Durand, and Humberto González (1987). Return to Aztlan: The Social Process of International Migration from Western Mexico, Berkley, CA: University of California Press. Massey, Douglas, Joaquin Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino, and Edward Taylor (1994). “An Evaluation of International Migration Theory: The North American Case,” Population and Development Review, 20 (4), 699-751. Massey, Douglas, and Jorge Durand (2005). Crossing the Border. Research from the Mexican Migration Project, New York, NY: Russell Sage Foundation. Massey, Douglas S., Jorge Durand, and Nolan J. Malone (2002). Beyond Smoke and Mirrors: Mexican Immigration in an Era of Economic Integration, New York: Russell Sage Foundation. Massey, Douglas and Kristin Espinosa (1997). “What’s Driving Mexico-U.S. Migration?: A Theoretical, Empirical and Policy Analysis,” American Journal of Sociology, 102, 939-99. Massey, Douglas and Audrey Singer (1998). “The Social Process of Undocumented Border Crossing among Mexican Migrants,” International Migration Review, 32 (3), 561-592. Massey, Douglas and Rene Zenteno (2000). “A Validation of the Ethnosurvey: The Case of Mexico-U.S. Migration,” International Migration Review, 34(3), 766-793. McKinley, James (2010). “An Arizona Morgue Grows Crowded,” New York Times, July 28. McKenzie, David, and Hillel Rapoport (2005). “Migrant Networks, Migration Incentives and Education Inequality in Rural Mexico,” Mimeo, Stanford, CA: Stanford University. Menjivar, Cecilia (2000). “Fragmented Ties: Salvadoran Immigrant Networks in America,” Berkeley, CA: University of California Press. Mestries-Benquet, Francis (1994). “Migración Internacional y Agricultura Campesina en Zacatecas,” in Mario Bassols (ed) Campo y Ciudad en una Era De Transición, Mexico, DF: Universidad Autónoma Metropolitana.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

References

205

Miguel, Edward and Michael Kremer (2003). “Networks, Social Learning, and Technology Adoption: The Case of Deworming Drugs in Kenya,” Mimeo, University of California, Berkley. Mines, Richard (1981). “Developing a Community Tradition of Migration to the United States,” Monograms in U.S.-Mexican Studies No 3, San Diego, CA: University of California. Mines, Richard (1984). "Network Migration and Mexican Rural Development: A Case Study" in Richards Jones (ed) Patterns of Undocumented Migration, Totowa, NJ: Rowman & Allanheld. Mines, Richard and Douglas Massey (1985). "Patterns of Migration to the US from Two Mexican Communities." Latin American Research Review, 20(2), 104-23. MMP (2007). Mexican Migration Project 114, Princeton, NJ: Princeton University and the University of Guadalajara (available at: http://mmp.opr.princeton.edu/). Montgomery, James (1991). “Social Networks and Labor-Market Outcomes: Toward an Economic Analysis,” American Economic Review, 81(5), 1408–1418. Monto, Alexander (1994). The Roots of Mexican Labor Migration, Westport, CT: Praeger Publishers. Mooney, Margarita (2004). "Migrants Social Capital and Investing Remittances in Mexico," in Jorge Durand and Douglas Massey (eds) Crossing the Border, New York: Russell Sage Foundation. Morawska, Eva (1990). "The Sociology and Historiography of Immigration," in Virginia Yans-McLaughlin (ed) Immigration Reconsidered: History, Sociology, and Politics, New York: Oxford University Press. Moses, Jonathon, and Bjørn Letnes (2004). “The Economic Costs to International Labor Restrictions: Revisiting the Empirical Discussion,” World Development, 32(10), 1609–26. MPI (2006). From a Zero-Sum to a Win-Win Scenario? Literature Review on Circular Migration, Washington, DC: Migration Policy Institute. Mulhausen, David (2006). “Building a Better Border: What the Experts Say,” Backgrounder No. 1952, Washington DC: Heritage Foundation. Munshi, Kaivan. (2003). “Networks in the Modern Economy: Mexican Migrants in the U.S. Labor Market,” Quarterly Journal of Economics, 118(2), 549-599. Orozco, Manuel (2009). “Remittances and Social Development: The Latin American Experience,” in Katja Hujo and Shea McClanahan (eds) Financing Social Policy: Mobilizing Resources for Development, New York, NY: UN Research Institute for Social Development.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

206

References

Orrenius, Pia and Roberto Coronado (2007). "Border Crime and Border Enforcement,” in M. Barraza de Anda and T. Fullerton Jr. (eds) Basic Border Econometrics, Cd. Juarez, Mexico: Universidad Autonoma de Cd. Juarez. Orozco, Manuel and Rachel Fedewa (2006). “Leveraging Efforts on Remittances and Financial Intermediation,” Inter-American Development Bank Working Paper No. 24. Orrenius, Pia and Madeline Zavodny (2003). “Do Amnesty Programs Reduce Undocumented Immigration? Evidence from IRCA,” Demography, 40 (3), 437–450. Orrenius, Pia and Madeline Zavodny (2010). "Mexican Immigrant Employment Outcomes over the Business Cycle,”, American Economic Review: Papers & Proceedings, 100 (2), 316–320. Pagan, Adrian and Aman Ullah (1999). Nonparametric Econometrics, New York, NY: Cambridge University Press. Palloni, Alberto, Douglas Massey, Miguel Ceballos, Kristin Espinosa, and Michael Spittel (2001). “Social Capital and International Migration: A Test Using Information on Family Networks,” The American Journal of Sociology, 106 (5), 1262–98. Papademetriou, Demetrios (2005). “Reflections on Restoring Integrity to the United States Immigration System: A Personal Vision,” MPI Insight No. 5, Washington, D.C: Migration Policy Institute. Papail, Jean (1998). "Factores de la Migracion y Redes Migratorias," in U.S. Commission on Immigration Reform (ed) Migration Between Mexico and the United States: Binational Study. Washington, DC: U.S. Commission on Immigration Reform. Passel, Jeffrey (2006). Size and Characteristics of the Unauthorized Migrant Population in the U.S., Washington, DC: Pew Hispanic Center. Passel, Jeffrey, and Rebecca Clark (2008). Immigrants in New York: Their Legal Status, Incomes and Taxes, New York, NY: Urban Institute. Passel, Jeffrey and D’Vera Cohn (2009). Mexican Immigrants: How Many Come? How Many Leave?, Washington, DC: Pew Hispanic Center. Passel, Jeffrey and Roberto Suro (2005). Rise, Peak, and Decline: Trends in U.S. Immigration 1992–2004, Washington, DC: Pew Hispanic Center. Pastor, Manuel and Susan Alva (2004). “Guest Workers and the New Transnationalism: Possibilities and Realities in an Age of Repression,” Social Justice, 31(1–2), 92-112. Perryman Group (2008). An Analysis of the Economic Impact of Undocumented Workers on Business Activity in the US with Estimated Effects by State and by Industry, Waco, TX: The Perryman Group.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

References

207

Pessar, Patricia (1999). “The Role of Gender, Households, and Social Networks in the Migration Process: A Review and Appraisal,” in Charles Hirschman, P. Kasinitz and J. DeWind (eds) The Handbook of International Migration: The American Experience, New York, NY: Russell Sage. Piore, Michael (1979). Birds of Passage: Migrant Labor and Industrial Societies, Cambridge, UK: Cambridge University Press. Portes, Alejandro (2001). “Introduction: The Debates and Significance of Immigrant Transnationalism,” Global Networks, 1(3), 181–193. Portes, Alejandro (2007). “Migration, Development, and Segmented Assimilation: A Conceptual Review of the Evidence,” The Annals of the American Academy of Political and Social Science, 610, 73 - 97. Portes, Alejandro and József Böröcz (1989). “Contemporary Immigration: Theoretical Perspectives on Its Determinants and Modes of Incorporation,” International Migration Review, 23, 606-30. Portes, Alejandro, Cristina Escobar Alexandria, and Walton Radford (2005). “Immigrant Transnational Organizations and Development: A Comparative Study,” Center for Migration and Development Working Paper No 05-07. Portes, Alejandro, and Ruben Rumbaut (2006). Immigrant America: A Portrait, Berkley, CA: University of California Press. Preston, Julia (2010). “Justice Dept. Sues Arizona Over Its Immigration Law,” New York Times, July 6. Pries, Ludger (2004). “Determining the Causes and Durability of International Labor Migration Between Mexico and the United States: Some Empirical Findings,” International Migration, 42(2), 3-39. Rapoport, Hillel and Frédéric Docquier (2006). “The Economics of Migrants’ Remittances,” in J. Mercier-Ythier and S.C. Kolm (eds), Handbook on the economics of giving, altruism and reciprocity, Amsterdam: Elsevier. Reichert, Joshua, and Douglas Massey (1979). “Patterns of U.S. Migration from a Mexican Sending Community: A Comparison of Legal and Illegal Migration,” International Migration Review, 13(4), 599 – 623. Reichert, Joshua (1981). “The Migrant Syndrome: Seasonal U.S. Wage Labor and Rural Development in Central Mexico,” Human Organization, 40, 5666. Reyes, Belinda, Hans Johnson, Richard Van Swearingen (2002). Holding the Line? The Effect of the Recent Border Build-up on Unauthorized Immigration, San Francisco, CA: Public Policy Institute of California.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

208

References

Reyes, Belinda (2004). "Changes in Trip Duration for Mexican Immigrants to the United States," Population Research and Policy Review, 23(3), 235257. Richter, Susan, Edward Taylor, and Antonio Yúnez-Naude (2007), “Impacts of Policy Reforms on Labor Migration from Rural Mexico to the United States,” in George Borjas (ed) Mexican Immigration to the United States, Chicago, US: The University of Chicago Press. Rosenblum, Marc (2004). "Beyond the Policy of No-Policy: Emigration from Mexico and Central America,” Latin American Political Society, 4(1), 91– 125. Rosenzweig, Mark and Oded Stark (1989). “Consumption Smoothing, Migration and Marriage: Evidence from Rural India,” Journal of Political Economy, 97, 905-26. Rozelle, Scott, Edward Taylor, and Alan de Brauw (1999). “Migration, Remittances, and Agricultural Productivity in China,” American Economic Review, 89(2), 287-291. Ruhs, Martin (2005). “Designing Viable and Ethical Labour Immigration Policies,” in IOM (ed) World Migration 2005, Geneva, Switzerland: International Organization for Migration. Rumbaut, Ruben G. (1997). “Assimilation and Its Discontents: Between Rhetoric and Reality,“ International Migration Review, 31(4), 923-960. Sassen, Saskia (1988). The Mobility of Labor and Capital: A Study in International Investment and Labor Flow, New York, NY: Cambridge University Press. Schiller, Nina Glick, Linda Basch, and Cristina Blanc-Szanton (1992). "Transnationalism: A New Analytic Framework for Understanding Migration." In Nina Glick Schiller, L. Basch and C. Blanc-Szanton (eds) Towards a Transnational Perspective on Migration, New York, NY: New York Academy of Sciences. Schiller, Nina Glick and Linda Basch (1995). “From Immigrant to Transmigrant: Theorizing Transnational Migration,” Anthropological Quarterly, 68 (1), 48 – 64. Silverman, Bernard (1986). Density Estimation for Statistics and Data Analysis, New York, NY: Chapman and Hall. Simon, Carl, and Lawrence Blume (1994). Mathematics for Economists, W.W. Norton & Company, Inc.: New York, N.Y. Sjaastad, Larry (1962). "The Costs and Returns of Human Migration," The Journal of Political Economy, 70(2), 80-93. Smith, James and Duncan Thomas (1997). “Migrations in Retrospect: Remembrances of Things Past,” Rand Unrestricted Draft No 1628.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

References

209

Spener, David (2001). “El contrabando de Migrantes en la Frontera de Texas con el Nordeste de México: Mecanismo para la Integración, “Espiral: Estudios sobre Estado y Sociedad, 7(21), 201 -247. Spener, David (2004). “Mexican Migrant-Smuggling: A Cross-Border Cottage Industry,” Journal of International Migration and Integration, 5(3), 293320. Spener, David (2009). Clandestine Crossings: Migrants and Coyotes on the Texas-Mexico Border, Ithaca, NY: Cornell University Press. Stark, Oded (1978). Economic-Demographic Interactions in Agricultural Development: The Case of Rural-to-Urban Migration, Rome: Food and Agriculture Organization. Stark, Oded (1991). The Migration of Labour, Cambridge, MA: Basil Blackwell. Stark, Oded and David Bloom (1985). “The New Economics of Labor Migration,” American Economic Association Papers and Proceedings, 75 (2), 173-178. Stark, Oded and Simon Fan (2007). “On the Analytics of Seasonal Migration,” Economics Letters, 94(2), 304-312. Stark Oded, Helmenstein Christian., Prskawetz Alexia (1997). “A Brain Gain with a Brain Drain,” Economics Letters, 55(2), 227-234. Stark, Oded, and Edward Taylor (1991). “Migration Incentives, Migration Types: The Role of Relative Deprivation,” The Economic Journal, 101, 1163-1178. Tapia, Carlos (2003). “Recorriendo Caminos: La Literature Acerca de la Migracion Michoacana,” in Lopez Castro (ed) Diáspora Michoacana, Zamora, Mexico: El Collegio de Michoacan. Taylor, Edward (1986). “Differential Migration, Networks, Information and Risk,” Human Capital and Development, 4, 147-171. Taylor, Edward (1987). “Undocumented Mexico-US Migration and the Returns to Households in Rural Mexico,” American Journal of Agricultural Economics, 69, 616-638. Taylor, Edward (1995). Micro Economy-Wide Models for Migration and Policy Analysis: An Application to Rural Mexico, Paris, France: Organization for Economic Cooperation and Development. Taylor, Edward (1999). "The New Economics of Labour Migration and the Role of Remittances in the Migration Process," International Migration, 37 (1), 63-84. Taylor, Edward and Alejandro Lopez-Feldman (2010). "Does Migration Make Rural Households More Productive? Evidence from Mexico", Journal of Development Studies, 46 (1), 68–90.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

210

References

Taylor, Edward, Jorge Mora, Richard Adams, and Alejandro Lopez-Feldman (2005). “Remittances, Inequality and Poverty: Evidence from Rural Mexico,” University of California Davis, Department of Agricultural and Resource Economics Working Paper No. 05-003. Tienda, Marta (1980). “Families and Structural Assimilation of Mexican Immigrants in the United States,” International Migration Review, Vol. 14 (3), 383 – 408 Tilly, Charles (1990). “Transplanted Networks,” in Virginia Yans-McLaughlin (ed). Immigration Reconsidered: History, Sociology, and Politics, New York: Oxford University Press. Tilly, Charles, and Harold Brown (1967). “On Uprooting, Kinship, and the Auspices of Migration,” International Journal of Comparative Sociology, (8), 139-164. Topa, Giorgio (2001). “Social Interactions, Local Spillovers, and Unemployment,” Review of Economic Studies, 68 (2), 261-295. Train, Kenneth (2003). Discrete Choice Methods with Simulation, New York, NY: Cambridge University Press. Trigueros, Paz and Javier Rodriguez Pina (1988). "Migracion y Vida Familiar en Michoacan (un Studio de Caso)." in Gustavo Lopez and Sergio Pardo Migracion en el Occidente de Mexico, Zamorah, Michoacan: El Colegio de Michoacan. U.S. Commission on Immigration Reform (1998). “Migration between Mexico and the United States - Binational Study,” Washington DC: U.S. Commission on Immigration Reform. U.S. Department of Defense (2007). National Defense Budget Estimates for FY 2007, Office of the Under Secretary of Defense, Washington, DC: U.S. Department of Defense. Udry, Christopher (2001). “Social Learning Through Networks: The Adoption of New Agricultural Technologies in Ghana,” American Journal of Agricultural Economics, 83(3), 668-73. United Nations (2002). World Population Prospects: The 2002 Revision, Volume I, Department of Economic and Social Affairs - Population Division, New York, NY: United Nations. United Nations (2008). International Migrant Stock: The 2008 Revision, Department of Economic and Social Affairs - Population Division, New York, NY: United Nations. Waldinger, Roger (1997). “Social Capital or Social Closure? Immigrant Networks in the Labor Market,” The Ralph and Goldy Lewis Center for Regional Policy Studies Working Paper Series No 26.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

References

211

Waldinger, Roger and David Fitzgerald (2004), "Transnationalism in Question,” American Journal of Sociology, 109 (5), 1177–95. Wallerstein, Immanuel (1974). The Modern World System. Capitalist Agriculture, and the Origin of the European World Economy in the Sixteenth Century, New York, NY: Academic Press. Weber, Christian (1999). “On the Determinants of Complementarity: Elasticities of Substitution, Income Shares, and Relative Prices,” Research in Economics, 53, 323-336. Wiest, Raymond (1984). “External Dependency and the Perpetuation of Temporary Migration to the United States,” in Richards Jones (ed) Patterns of Undocumented Migration: Mexico And The United States, Totowa, NJ: Rowman & Allanheld. Wilson, Tamar (1998). “Weak Ties, Strong Ties: Network Principles in Mexican migration,” Human Organization, 57(4), 394-403. Winters, Alan, Terrie Walmsley, Zhen, Wang, and Roman Grynberg (2002). “Negotiating the Liberalisation of the Temporary Movement of Natural Persons,” University of Sussex Discussion Papers in Economics No. 87. Winters, Paul, Alain de Janvry, and Elisabeth Sadoulet (2001). “Family and Community Networks in Mexico–U.S. Migration,” Journal of Human Resources, 36(1), 159-84. Wood, Charles (1981). “Structural Changes and Household Strategies: A Conceptual Framework for the Study of Rural-Urban Migration,” Human Organization, 40, 338-344. Woodruff, Christopher, and Rene Zenteno (2007). “Migration Networks and Microenterprises in Mexico,” Journal of Development Economics, 82 (2), 509-528. World Bank (2006a). Mexico 2006–2012: Creating the Foundations for Equitable Growth, Report No. 39993-MX, Washington, DC: World Bank. World Bank (2006b). Global Economic Prospects - Economic Implications of Remittances and Migration, Washington DC: World Bank. World Bank (2006e).The Development Impact of Workers’ Remittances in Latin America, Report No. 37026, Washington, DC: World Bank. World Bank (2007). Migration and Remittances: Eastern Europe and the Former Soviet Union, World Bank, Washington DC: World Bank. World Bank (2010), World Development Indicators, 2006-CDROM edition and www.data.worldbank.org (accessed on November 22, 2010). Yans-McLaughlin, Virginia (1990). Immigration Reconsidered: History, Sociology, and Politics, New York, NY: Oxford University Press. Yoshida, Chisato and Alan Woodland (2005). The Economics of Illegal Immigration, New York, NY: Pelgrave Macmillian.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

212

References

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Zahniser, Steven (1999). “One Border, Two Transitions Mexican Migration to the United States as a Two-Way Process,” American Behavioral Scientist, 42 (9), 1350-1380.

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Index

country of origin see source country cumulative causation 8 data data issues 107 EMIF 22 MMP 47–49 original data set 20, 151–52 Department of Homeland Security 7, 130 destination country 2, 139 development cooperation 150 development partnership 148–50 diaspora 4 distributional effects 139, 140, 141 earned legalization 145 economic efficiency 139, 141 emigrant see migrant types Europe 2, see also destination country eviction 35, 101, 145 extensive migration behavior 37, 43 compositional differences 116 determinants 100–101, 109– 17

assimilation 2, 39, 51, see also migration intensity asymmetric information 19 border 96 border enforcement 2, 130–31, 141 costs 94, 130 impact on migration decision 78–80, 83, 97–99, 109, 125–27 impact on migration strategies 78–80, 83, 99– 100, 117, 121–24 instruments 106, 112 line-watch hours 104–6, 127 net effect 126 regional border enforcement measure 113, 153 U.S. border control 96 Bracero Program 95, 102 brain drain 65 chain migration See social  networks circular migration 146 Comprehensive Immigration Reform Act 145 country of destination see destination country

213

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

214 model 70 propensity to migrate 101, 127 regression model 108 wage gap 100, 112 family and friends effect see  social networks family reunification 49, 87 determinants 124, 131 financial deepening 149 globalization 2 helpers 27–28, 30, 35 High-Level Dialogue on Migration 37 'home attachment' see migration intensity home ownership 49 determinants 124, 131 home-time consumption 49, 69 determinants 82, 124, 131 home-town associations 150 human rights 139, 141 illegal Mexican migration see Mexican migration Immigration and Naturalization Service 95, 104 immigration policies see policy making Immigration Reform and Control Act 61, 95 Index of Migration Intensity see migration intensity, index INEGI 113 intensive migration behavior see migration intensity international migration 1, 68, see also migration models development 1, 37, 41–42, 63, 149

Index understanding labor migration 136–38 International Organization for Migration 42, 150 investment 2, 86–89, 149, see also home ownership model 86 labor export-led growth model 1, 146 labor rights 139, 141 language acculturation 51 legalization efforts determinants 124, 131 marginalization 139 mass migration 4, 7, 95–96 Mexican government see source country Mexican migration 4, 55–56, 61, see also migration process dependency 37, 131 flow of undocumented migrants 4, 7, 109, 127 geography 107, 172 history 95–96 migrant profile 95, 172 remittances 49, 148 stock of undocumented migrants 7, 145 Mexico - U.S. immigration see Mexican migration 'migradollars' see remittance sending migrant types 3, 39, 55, 60, 63 determinants 116, 125 simulation results 81–83 migration decision see extensive migration behavior migration intensity 83 concept 43–44 decomposition 121–24

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

Index determinants 56–61, 100– 101, 117–25 index 51, 55–56, 56, 61 indicators 44–46, 49 model 71–73 regression model 108 wage gap 100, 120 migration management see policy making migration models 37 assumptions 68, 69, 74 Harris and Todaro 66 home-time consumption 69, 72 optimal strategy 75–77, 78, 85 remittance sending behavior 67, 72, 80, 82 return migration 67, 82 simulation 81–86 social network assistance 163 migration networks see social networks migration patterns see migration strategies; migrant types migration process 12, 13 migration steps 14 migration strategy 43, 57, 71–73, see also migrant types NAFTA 96 national identity 139, 149 national security 139, 141 New Economics of Labor Migration 67, 88 NGOs 150 Operation Gatekeeper 96 Operation Hold-the-Line 96 Operation Rio Grande 96 Operation Safeguard 96

215 policy making 3, 35, 63, 78–80, 139–50, see also border enforcement economic development 84 guest worker program 86, 146 legalization 86, 95 policy objectives 139–40 poverty reduction 40, 148 President Bush 86 principal components analysis 46, 155 principal-agent problem 16 public finances 139, 140, 141 public service use see public finances rancho 151 remittance sending 38, 43, 65, 80, 82, see also migration models, remittances multiplier effect 125 ressource outflow 140, 141 return migrant see migrant types; return migration return migration 38, see also migration intensity rule of law 139, 141 seasonal migrant see migrant types settlement 49, 56, 131, 149, see also migration intensity determinants 82, 124 skilled migration 3 smuggler 16 fee 83 social network assistance 7, see also types of social network assistance; helpers

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,

216

types of social network assistance 8 border crossing 16–17 cash on delivery 17 co-residence 9 domestic dislocation 16 information provision 14 initial financing 15 initial U.S. sustenance 18 job search 9, 18 prevalence 24 relative importance 19 travel companionship 9 U.S. border control costs 94, 130 unauthorized Mexican migration see Mexican migration United Nations 1 World Bank 149

Copyright © 2011. LFB Scholarly Publishing LLC. All rights reserved.

beneficiaries’ characteristics 30 model 163 pre-departure assurance 24, 27, 152 prevalence 10, 23–34 social networks 13, 57 measurement 21 migration behavior 101, 112, 120 sojourner see migrant types source country 3, 41, 139 stay-versus-migrate decision See  extensive migration  behavior transaction costs 65, 68 transnationalism 40, 63, 137, see also migrant types trust 15–17

Index

Mexican Labor Migrants and U.S. Immigration Policies : From Sojourner to Emigrant?, LFB Scholarly Publishing LLC, 2011. ProQuest Ebook Central,