c
Lisa Hofheinz
ID 11263342 Supervisor: Dr. Abbey Steele Second reader: Dr. Jana Krause June 2017 Master Thesis Political Science – International RelationsTransnational Streetgangs:
Immigrants from Central America and their
participation in the MS-13 and the M-18
II
Abstract
In the 1990s, transnational street gangs have been identified as a new phenomenon threatening international and public security in targeted countries (Arana, 2005). The Mara Salvatrucha 13 (MS-13) and the 18th Street Gang (M-18), identified as the maras, have evolved into the most violent contemporary examples by reaching the definitional status of a “new urban insurgency” (Manwaring, 2005). The maras’ transnational character builds on their formation in California and the evolution embedded in migration patterns between the U.S. and Central America (CA). This has been addressed by existing scholarship due to the maras’ violent criminal activities and their geopolitical implications. A puzzle identified by the literature is the varying spatial proliferation of the maras to CA.
Previous research has found a covariance between the proliferation of and the participation in the U.S. maras. It suggests that immigrants from El Salvador, Guatemala and Honduras were more likely to join the MS-13 and the M-18 in the U.S. than other immigrant groups from CA. This leads to the main inquiry of this paper, which develops the stated puzzle further and addresses the question of what explains the varying participation of immigrant groups from CA in the maras in the U.S.
This paper provides evidence for the proposed variation in the mara participation. It further finds a covariance between membership in the U.S. maras and the legal status of residency of immigrants from CA. Empirical evidence supports the argument that increased economic exclusion resulting from undocumented residency links these two variables. The theoretical framework is developed by building on existing approaches to gang participation and evolution, which highlight the concepts of multiple marginality and numerous risk factors. It is argued that economic exclusion fortifies these factors influencing the voluntary joining of the MS-13 and the M-18 as well their coercive recruitment. Word count: 17.163 (incl. citations)
III
Abbreviations Central America
Federal Bureau of Investigation
Illegal immigration Reform and Immigrant Responsibility Act Immigration Reform and Control Act Lawful Permanent Resident Mara Dieciocho Mara Salvatrucha Migration Policy Institute Organization of American States Organized Crime Temporary Protected Status Transnational Migration Patterns United Nations United Nations Office on Drugs and Crime United Nations Development Programme U.S. Immigration and Naturalization Services United States of America (CA) (FBI) (IIR IRA) (IRCA) (LPR) (M-18) (MS-13) (MPI) (OAS) (OC) (TPS) (TMP) (U.N.) (UNODC) (UNDP) (INS) (U.S.)
IV
Contents
Abstract ... II Contents ... IV 1. Introduction ... 1 2. Literature Review ... 4 3. Theoretical Framework ... 7 4. Research Design ... 10 4.1. Operational definitions ... 10 4.2. Operationalization ... 12 4.3. Methodology ... 16 4.4. Case selection ... 18 4.5. Data collection and sources ... 20 4.6. Background of the cases ... 24 5. Research Results ... 27 5.1. Expected findings ... 27 5.1.1. Participation in the maras ... 28 5.1.2. Unauthorized population in the U.S. ... 29 5.1.3. Economic exclusion of Central American immigrants ... 31 5.2. Analysis ... 34 5.2.1. Participation in the U.S. maras and legal status of residency ... 34 5.2.2. Economic exclusion of undocumented immigrants ... 37 5.2.3. Economic exclusion as a causal mechanism ... 39 5.3. Discussion ... 43 6. Conclusion ... 48 7. Bibliography ... 51 8. Figures ... 59 9. Tables ... 60
1
1.
Introduction
Figure 1: MS-13 gang member in El Salvador. Source: Gettyimages, available at https://ichef-1.bbci.co.uk/news/624/cpsprodpb/1248/production/_95708640_gettyimages-56730305.jpg (09.05.2017)While it is difficult to determine the proportion of violent criminal activity directly associated with the Mara Salvatrucha (MS-13) and the Mara Dieciocho (M-18), two of the most violent contemporary transnational street gangs, „there can be no doubt that they [now] account for a significant […] share of actual and perceived violence“ (Rodgers, 1999, p.9) in Central America (CA). The MS-13 and the M-18, defined as the
maras1, have become a variable in the political systems of El Salvador, Honduras and
Guatemala, commonly referred to as the Northern triangle,2 by evolving into interstitial
organizations “emerging as alternative sources of orders in areas of deficient statehood” (Thrasher, 1927, p.10).
The transnational formation and evolution rooted in the historical migration pattern between CA and the United States (U.S.) make for a complex and unique phenomenon
1 The maras are an interchangeable synonym for the MS-13 and the M-18 street gangs for the purpose of
this paper.
2 These countries are conceptualized as the “Northern triangle” in the literature ((Bruneau et al., 2011;
2 with major geopolitical implications. The need for comprehensive research on the roots and causes of these particular street gangs is essential.3 The MS-13 and M-18 are both Hispanic street gangs, whose origins lie in Los Angeles, USA (Bruneau et al., 2011). The M-18 was founded in the 1960 by Mexican immigrants and significantly grew in scale due to an influx of Central American refugees in the 1980s. Between 1985 and 1987, immigrants from El Salvador established the rival gang MS-13. Both gangs began to control turf in low-class neighborhoods, so-called barrios, in L.A. and engaged in extreme violence (against one another and other rival gangs, as well
as law enforcement) and illegal criminal activities.4
The maras account for a new phenomenon in the gang spectrum, which distinguishes itself from more local gangs by its transnational roots and activities in the international drug trade, its connections to the United State’s prison gangs and the Mexican Mafia
(Valdez, 2011). By 2015, the maras had expanded from California to nearly all 50
American states, Central America, Canada and Europe (Feere and Vaughan, 2008; Miguel Cruz, 2010).
A puzzling observation was addressed in the 1990s, when the Northern triangle started to face mara-related levels of violence “that surpass[ed] those of the revolutionary conflicts that affected the region during the 1970s and 1980s” (Rodgers and Hazen, 2015, p.2). Belize, Costa Rica, Panama and Nicaragua in contrast were experiencing no significant levels of mara activity (Claire Ribando Seelke, 2016; UNODC, 2007, 2012). The questions of how the maras evolved not only over time but also over space became
increasingly relevant (Decker and Pyrooz, 2015; Rodgers and Hazen, 2015)5.
3 This study solemnly focuses on the MS-13 and the M-18 (inclduing their local clikas) and does not
incorporate the presence of other local pandillas and street gangs due to the dominant relevance of the Maras in the political and academic discourse.
4 Gang members have specific symbols, hand gestures and tattoos that openly demonstrate their group
identity to either the M-18 or the MS-13 (and their respective local clikas). Activties range from low-level criminal activity to an increasingly professionalized participation in OC, particularily in regard to drugs (Arias and Goldstein, 2010; Koonings and Kruijt, 2007; Manwaring, 2005; Sullivan and Elkus, 2009) 5 Even within the changing landscape of violence in CA, the maras have to be viewed separately from the more endemic street gangs for the mentioned reasons. The distinctiveness of this “new” phenomenon claims definitional accuracy within both research and literature in order to identify both new levels of gang related violence and the transnational proliferation as well as to understand and address both political and societal implications (Coughlin and Venkatesh, 2003; Rodgers and Hazen, 2015).
3
The existing scholarship argues that migration explains the initial transnational
expansion of the maras from the United States to Central America in the 1990s6
(Abrahamsen et al., 2009; Bruneau et al., 2011; Meyer, 2016; Rodgers and Hazen, 2015). Among others, members of the U.S. maras were deported back to their countries of origin. The literature finds a covariance between a varying spatial proliferation of the maras and a varying participation of Central American immigrants in the U.S. maras but so far has not developed conclusive causal explanations for the evolution and expansion of the phenomenon.
The relevance of a potential variation in gang membership in the U.S. for the broader understanding of the evolution and expansion has been established but not addressed empirically. This leaves a significant gap in the literature. This paper develops the puzzle regarding the variations within the phenomenon further and addresses the research
question of what explains a varying participation of immigrant groups from CA7 in
the maras in the U.S.
Inherently, this requires a measuring of participation, a broader understanding of the explanatory variables for gang participation as well as the underlying dynamics. Building on Rodgers and Baird’s (2015) theoretical argument, a potential covariance between legal status of residency and mara participation will be explored. If supporting evidence is found, the research will analyze the mechanisms linking legal status of residency and gang participation.
As this research is exploratory and the reasons for gang participation are multifactorial (Bordua, 1961; Rodgers, 1999), alternative explanatory mechanisms will be taken into consideration.
The main purpose of this paper is to address the defined gap in the literature by approaching a specific sequence in the distinctive evolution of the maras. Understanding
6 The discourse about the transnational dimension of the “super gangs” MS-13 and M-18 is diverse. Clikas
(local branches) of the maras emerged in some Central American countries in the late 1990s, “rapidly attracted local youths and either supplanted or absorbed local pandillas” (Rodgers and Baird, 2015b, p.6). While local clikas are explicitly affiliated with the maras in the U.S. neither a federal nor a transnational structure or chain of command seems to exist (Bruneau et al., 2011; Rodgers and Baird, 2015b; Valdez, 2011). 7 Central America in this paper encompasses the countries of Belize, Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua and Panama. Mexico will be neglected due to the scale of this paper and its increased complexity related to cartels and other entities involved in OC.
4
the reasons for the expected variation in participation will provide further insight on their proliferation to CA. It builds a foundation for future research regarding evolving dynamics and consequently the implications of the maras’ activities across Central and Northern America and beyond its borders (Medina et al., 2007; UNDP, 2013; Wolf, 2012).
This paper is structured as following. Chapter 2 summarizes how the existing scholarship addresses the puzzle. This leads to the development of a theoretical framework for this thesis and the formulation of two hypotheses. Chapter 4 will elaborate on the design of the research by defining the operationalization of the employed concepts and justifying the research techniques as well as the sources of data, the indicators of measurements and their respective limitations. The research results are summarized and discussed in order confirm or reject the hypotheses and to assess the degree to which they address the research question. Lastly, chapter 6 concludes and illustrates the implications of the findings for the research question, the proposed theoretical framework and future research.
2. Literature Review
This chapter is concerned with the existing scholarship’s approach to the illustrated research question. It further defines the gap in the literature by elaborating on the literature discussing the evolution of the maras.
Research on the maras has picked up from the fields of sociology, criminology and anthropology as well as by political science and international relations scholars in the early 1990s. This was a direct result of the political and economic implications of the maras’ increasingly internationalized illegal violent activities (Miguel Cruz, 2010; Núñez,
1996; Rocha, 2008; Rodgers, 2007).8
8 Regional Research Institutes as well as the Federal Bureau of Investigation (F.B.I.) in the U.S. began to
publish multivolume studies on Central American gangs and annual reports in the 2000s (ERIC et al.). Studies based on primary work by the United Nations (UNODC, 2007; 2012; 2013), Demoscopía (2007), USAID (2006) and National and Congressional Research agencies (Ribando Seelke, 2016; Manwaring, 2005; Medina et al., 2007; Meyer, 2016), mainly from the U.S. and Canada, addressed the “gang threat” (Ribando Seelke, 2016, p.2).
5
The transnational character of the MS-13 and the M-18 had further challenged the
territorial aspect of the classical analytical framework for gang studies9.
The scholarly discourse on the maras has emphasized the relevance of the broader
transnational migration pattern (TMP)10 between Central America and the U.S. for a
comprehensive understanding of expansion. These TMP entail Central American migration to the U.S., the formation and growth of the gang from California to all U.S. states, the deportation of U.S. members of the maras back to their countries of origin in CA and the resulting evolution of the phenomenon (see figure 2). The circular dimension of the development is acknowledged, as evidence shows the re-immigration of formerly deported gang members back to the U.S. (Feere and Vaughan, 2008; Lawrence, 2016; Sullivan, 2006; Zilberg, 2011). Figure 2. Transnational migration pattern relating to the evolution of the maras Sources: Data from Rodgers and Baird (2015), UNODC (2007) 9 Frederic Thrasher’s classic definition of a gang as “an interstitial group, originally formed spontaneously, and then integrated through conflict. It is characterized by the following types of behavior: meeting face to face, milling, movement through space as a unit, conflict, and planning. The result of this collective behavior is the development of tradition, unreflective internal structure, esprit de corps, solidarity, morale, group awareness, and attachment to a local territory” (Thrasher, 1927, p.57) is still considered a classic analytical framework (Bordua, 1961; Cohen, 1955; Miller, 1976; Rodgers and Baird, 2015a).
10 The concept of transnational migration pattern (TMP) serves as a definitional framework that describes
the the dynamics and structures of crossing borders and “belonging to two or more societies and the same time” (Levitt, 2004). For the purpose of this paper, it defines the recurring circular migration of individuals between CA and the U.S.
6
The discourse accommodates the proliferation of the maras by arguing that their relatively simultaneous emergence in CA in the late 1990s was at least partly resulting from the massive influx in criminal deportees from the U.S. (OAS, 2015; Rocha, 2013;
Rodgers and Baird, 2015).11
The U.S. Congress had passed the Illegal Immigration Reform and Immigrant
Responsibility Act (IIR IRA) in 1996, an immigration legislative that entailed an expansion of the list of deportable crimes to minor offenses, such as drunk driving and petty theft (Arana, 2005), and determined that “non-US citizens sentenced to one year or more in prison were to be repatriated to their countries of origin” (Rodgers and Hazen, 2015, p.4). Additionally, many political asylees from CA were no longer granted amnesty after the terminations of civil conflicts and political stability increased in the region in the 1990s (Rocha, 2013). The expiration of visas resulted in the undocumented residence of many immigrants, who were increasingly deported on the grounds of crimes against federal Immigration law (Light et al., 2014).
Resultantly, about 160,000 illegal immigrants and 50.000 convicts were sent back between 1996 and 2000 (United Nations Office on Drugs and Crime (UNODC), 2007, pp.40–42). Many of these deportations were the consequence of drastic measures by law enforcement agencies in L.A. in response to increasing gang violence and crime (Medina et al., 2007; Rodgers and Baird, 2015b). The overwhelming majority of the
criminal deportees were Salvadorians, followed by Hondurans and Guatemalans (Casas-Zamora, 2016; Demoscopía, 2007; OAS, 2015).12
The literature establishes a causality between their proliferation of the maras to the Northern triangle (Bruneau et al., 2011; Vigil, 1988; Wolf, 2012) and the participation of immigrants from El Salvador, Guatemala and Honduras in the U.S. maras (Demoscopía, 2007; UNODC, 2013; Vigil, 2003; Wolf, 2012).
11 A causal correlation between returning migrants and the influx in mara activity was established by the
literature, as studies demonstrated that many of their members in CA were deported from the U.S. and had been previous gang members (Demoscopia 2007).
12 Central American migrants affiliated with gang criminality in the L.A. were most likely members of the
7
Deportees founded local clikas of the MS-13 and the M-18 in their respective Central American “home countries”, remaining physically and symbolically affiliated with the maras in the U.S. (Abrahamsen et al., 2009; Rodgers, 1999; Rodgers and Hazen, 2015). This resulted in the growing of a “foreign problem imported by deportees that has evolved and grown in response to domestic factors and conditions” (Rodgers and Hazen,
2015, p.5)13. With the political and economic instability and the history of violence of the
respective countries serving as fertilizers, the maras of Central America became a powerhouse in the streets of Guatemala, El Salvador and Honduras. (Demoscopía, 2007; García, 2006; Koonings and Kruijt, 2007). The literature accordingly provides further
evidence for the evolution of the maras in CA.14
It becomes clear that the existing scholarship addresses both proliferation and evolution of the maras extensively. The formation of the maras has been studied, as has been outlined in the introduction to this paper, but “gang research […] so far failed to develop causal issues fully” (Ayling, 2011, p.2). Scholarship finds evidence for the emergence of the maras in the U.S. (Valdez, 2011; Vigil, 2014, 2003, 1988) but does not further provide insight on variation in participation.
The identified gap in the literature will be the main inquiry of this paper, as it is considered immanent to the understanding of the proliferation and the evolution of the maras (see figure 2). Based on the outlined discourse, the following chapter will build a theoretical framework to guide this research.
3. Theoretical Framework
The literature review demonstrates that mara membership in the U.S. has been linked contextually with their initial spatial proliferation (Federal Bureau of Investigation (FBI), 2015; Rodgers and Hazen, 2015; Vigil, 1988) and opens up space for further research (see table 7).
Accordingly, the guiding research question of this paper is what explains an expected varying participation of immigrants from Central America in the U.S. maras before 2000.
13 In addition to the already existing gang culture, the phenomenon was fueled by national policies in El
Salvador, Guatemala and Honduras.
14 The latest research focuses on issues of gender (Aguilar Umaña and Rikkers, 2012; Bruneau et al., 2011;
Oettler, 2011), transnational OC (Arnson et al., 2011; Shirk et al., 2015; UNODC, 2007) and societal reintegration dynamics of ex-gang members (Rodgers, 2016; Rodgers and Hazen, 2015).
8
Based on the described covariance, the maras’ proliferation to the Northern triangle could therefore be repatriated to the participation of immigrants from El Salvador, Guatemala and Honduras in the U.S. maras. Logically, immigrants from Belize, Costa Rica, Nicaragua and Panama are expected to have participated in the maras to a significantly smaller degree.
Understanding why some groups from CA were more likely to get affiliated with the mara requires an extensive comprehension of “what accounts for the emergence, decline, spread and evolution of gangs” (Hazen and Rodgers, 2014, p.12).
Rodgers and Baird (2015) find that Nicaraguans “did not participate in the local gang scene to the same extent” (Rodgers and Baird, 2015b, p.482) and Rocha points to the higher probability of legal admittance and naturalization for Nicaragua immigrants in the U.S., compared to other immigrants from CA (Rocha, 2007), as an explanatory factor. Legal immigration pattern of immigrants from the Northern triangle could have differed for several reasons.
Based on this argument, the relevance of the legal status of residency for gang participation will be explored. If Nicaraguans were more likely to obtain a legal status of residency in the U.S. and less likely to participate in the maras, I would expect immigrants from the Northern triangle to be less likely to be granted documented residency in the U.S., as they are expected to have been more likely to participate in the U.S. maras. Inherent to the argument is that undocumented immigrants were more likely to join the maras.
Increased levels of economic exclusion of undocumented immigrants could be the mechanism linking the two variables (López, 2015a; Miguel Cruz, 2010; Passel, 2006). Economic exclusion is a significant and measurable effect of undocumented residency in the U.S. as it limits the possibilities to integrate into the labor market and educational institutions (Rosenblum and Ruiz Soto, 2015; Terrazas, 2011; Warren and Warren, 2013a) and has been identified as a risk factors for gang participation.
Building on the existing theoretical framework in gang studies, I will rely on theories for the formation of street gangs by Vigil (1988, 2003, 2011) and Cruz (2010) to explore how exactly economic exclusion impacts the likelihood of joining gangs. Based on Cruz
9 (2010), the dynamics affecting the voluntary participation in the maras15, the coercive recruitment of the MS-13 and the M-18 in “governed barrios” (Cruz, 2010; p.395) and the coercive recruitment of members in prisons (Miguel Cruz, 2010; Skarbek, 2011) will be elaborated. The economic marginalization resulting from the increased likelihood for undocumented residency for immigrants from the Northern triangle might have increased their vulnerability to join or be recruited by the maras.
An alternative explanation for a variation in gang participation is based on the varying geographical dispersion of immigrant settlements within the U.S (Rodgers and Baird,
2015; Rocha, 2007).16 Alternative explanatory mechanisms for the participation in the
maras based on undocumented residency include aspects of social exclusion, which will be addressed in the Chapter 5.3.
The developed theory provides a framework linking isolated theoretical approaches presented by the existing scholarship in order to address the research question of this paper. Further, it proposes a plausible explanatory mechanism linking Central American immigrants’ participation in the maras with their legal status of residency. To evaluate whether the theory holds, the following hypotheses will be tested. H1: Immigrants from the Northern triangle were more likely to reside in the U.S. without legal authorization than other immigrants from CA. H2: Economic marginalization lead to the increased probability of gang participation. The degree to which the hypotheses hold will affect the validity of this theoretical framework. The results of the research are expected to reveal the implications of the developed theory on future scholarship. Eventually, this paper adresses the gap identified in the literature and provide insight on the variations within the mara
15 It is difficult to assess whether gang participation is voluntary, as coercive and voluntary participation
are identified and distinguished by the subjective psychological perceptions of the involved individuals. These can be ambivalent and are often influenced by the individuals’ interpretation of social pressure, for example.
16 Other potential independent variables will be taken into consideration but the research and analysis
will focus on the explanatory value of undocumented residency for the outcome due to the scope of this paper. Explanatory factors more endemic to the respective countries of origin (like state capacity, socioeconomic backgrounds, history of violence, geographic positioning for drug-trafficking and disarmament after civil war) are considered to be more relevant for the understanding of the varying presence of the maras in CA in the 2000s after they have first established local clikas. This is acknowledged but will be neglected due to a differing theoretical and analytical approach of the research question of this paper.
10
phenomenon in regard to its participating members and proliferation. The design of research employed to test the hypotheses relating to the theory will be elaborated in the following chapter.
4. Research Design
In order to test my hypotheses and relate the results to the developed theory, I will proceed as following. First, the conceptual definitions and the operationalization of the research will be identified. I will then elaborate on the methodology of this thesis and justify the selection of the cases and sources. This chapter will conclude by giving a brief background of the cases of analysis, providing a transition into the next chapter.
4.1.
Operational definitions
The employed operational definitons are clarified in order to increase the understanding of the following analysis and discussion.
The participation of immigrants in the maras measures the membership of immigrant groups from Central American countries in these gangs. Research identifies the maras as Latino-gangs (FBI, 2015; Rodgers and Baird, 2015b) and argues that immigrants from Central America were most likely to join the maras, when getting affiliated with gangs in the U.S. Ethnic exclusion and social marginalization of Hispanic immigrants in the U.S. society and by other street gangs were contributing factors to the initial formation of the gangs based on self-identification and continue to impact gang participation (Bordua, 1961; Miller, 1958, 2001; Vigil, 2014).
The participation in the maras does have perceived benefits for its members, the
mareros, but is generally considered to be physically dangerous because of
violence operated and experienced by the members. Participation in the maras required full-time and often life-long commitment. It limits future possibilities of licit societal integration (Bordua, 1961; Cohen, 1955). Members of the MS-13 and M-18 could be identified more easily than other gang members because of their flamboyant appearance including full-body tattoos (see figure 1) (Aguilar Umaña and Rikkers, 2012; Valdez, 2011; Vigil, 2003).
11 The spatial proliferation addressed in the broader context of the maras is defined as the “rapid increase in the number or amount of something“ (Oxford Dictionary). It entails the maras’ initial emergence in CA in the late 1990s and explicitly does not include their following expansion to other countries in CA. This is important as factors and dynamics relating to the initial proliferation vary from those impacting the following expansion.
Definitions of the term immigrant vary. This paper defines immigrants as the foreign-born population in the U.S. and includes “naturalized citizens, lawful permanent residents, refugees, asylees, legal nonimmigrants (including those on student, work, or certain other temporary visas), and persons residing in the country without authorization.“ (Lesser and Batalova, 2017).
The undocumented residency of individuals defines the legal status of a person “who resides in the United States but who is not a U.S. citizen, has not been admitted for permanent residence and is not in a set of specific authorized temporary statuses permitting longer-term residence and work” (Passel, 2006). Two groups of people account for the vast majority of this population. Those who overstayed their legal visa or violated their terms of admission and remained in the U.S., and those who entered without valid authorization (Warren and Warren, 2013a). The first group includes individuals that are authorized to stay in the U.S.
temporarily because of Temporary Protected Status (TPS)17 or on the grounds of
unresolved asylum claims.18 It is estimated that about 80% of U.S. immigrants stay
after their application for legal residency has expired or was declined (Schulkin, 2012; Warren and Warren, 2013a). The terms undocumented and unauthorized are used interchangeable.
Legal admission and naturalization are concepts used within the data collection
and analysis of this paper to indicate different aspects of the estimated size of the 17 “The United States provides limited humanitarian protection on a temporary basis – known as TPS – to individuals fleeing unsafe or problematic circumstances in their home countries (or who are already present in the U.S. and cannot return home) but who fail to meet the formal threshold required for refugee or asylum status. The Secretary of Homeland Security and Secretary of State can issue TPS for a period of six to 18 months and can extend the period if conditions do not change in the country or origin.“ (Terrazas, 2011).
18 The temporarily authorized individuals are expected to account for not more than 10% of the
12
undocumented population in the context of economic exclusion and deportation. Hence, they will be elaborated shortly. The majority of legal immigrants are admitted to the United States as immediate relatives of U.S. citizens or through the preference system, consisting of family-sponsored and employment-based immigration. Further, authorized immigration is possible through diversity programs. Accepted refugees and asylees are eligible to become immigrants by U.S. law through applying for permanent residency one year after they entered the United States.
Naturalization refers to “the conferring of U.S. citizenship, by any means, upon a
person after birth” (Gibson and Jung, 2006). The process of naturalization requires lawfully admitted permanent residence in the U.S. prior to the application for U.S. citizenship (U.S. Immigration and Naturalization Service, 1997). Undocumented residency became increasingly relevant in the context of Immigration crimes and deportations after the enactment of IIR IRA in 1996 (Arana, 2005). Economic exclusion is an operational concept that entails a „process whereby individuals or groups and the environments in which they live are excluded from the resources and opportunities which are considered the norm in a society.“ This refers to a „scarcity of material resources, [...] lack of opportunities, isolation, discrimination, marginalization from decision-making and from an adequate quality of life“ (O'Hara, 1998). It is closely related to poverty and includes aspects of the definition for social exclusion.
4.2.
Operationalization
This section demonstrates how the research is operationalized by incorporating the defined concepts and stating the employed variables as well as their indicators of measurement. Limitations of the measurement in relation to the data will be addressed in Chapter 4.5.
The dependent variable is defined as the participation of immigrants from C.A. in the U.S. maras. The independent variable of this paper is the legal residency status of immigrants from Central America in the U.S. between 1986 and 2000. This paper aims to explore if and how the independent variable accounts for a variation in the outcome.
13
First, the dependent variable will be measured in order to account for an expected variation in the outcome. Attempting to provide as much explanatory and empirical insight as possible, I choose to observe the participation of Central American immigrants in the U.S. maras between 1980 and 2000.
The measurement of participation in the maras is extremely challenging. The indicator of measurement is defined as qualitative or quantitative evidence on mara membership per country of origin provided by the literature. The outcome is measured nominally as significant participation and no significant participation. This accommodates a lack of precise data on gang members and their respective ethnicity, a lack of alternative indicators, reduces measurement errors and allows
for a variation in the outcome.19 This measurement error entails that the lack of
data for particular immigrant groups could result in no evidence for mara participation for these groups. This does not necessary mean that no participation existed. The measurement incorporates the varying scales of the immigrant population in the U.S. by considering the proportional value of the number of gang members in regard to the total number of immigrants from a particular country. By identifying significant trends, I am trying to increase the validity and the reliability of my measurement. The limitations of the data on mara participation will be elaborated further in section 4.5.
I am expecting to find that immigrants from the Northern triangle were significantly participating in the U.S. maras based on the claimed correlation with the proliferation of the maras to that region. Second, the independent variable had to be measured to confirm or reject the first hypothesis, stating that immigrants from the Northern triangle were more likely to reside in the U.S. without legal authorization than other immigrants from CA. The measurement of the legal status of residency of immigrants from CA between 1986 and 2000 is trying to capture the probability for obtaining a legal status of
19 The existing literature claims that the numbers of criminal deportees from the U.S. provide
insight into the participation in the U.S. This has to be analyzed with extreme caution as deportation polices were highly related to the legal status of residency. An equal relative share of gang participation for all immigrant groups from CA still would have resulted in varying deportation scales due to varying legal statuses. Additionally is no precise data on the specific cause for criminal conviction available, a distinction between criminal deportation based on crimes against federal immigration and gang crimes is not possible.
14
residency during the time period with the greatest influx in immigration from CA. Ideally, the time frame would include the earlier 1980s, which is not possible due to a lack of sufficient data before 1986.
An indicator for the legal status of residency is the estimated undocumented population in the U.S. by country of birth. Further indicators for the residency status will be the statistical evidence on naturalizations and deportations per year. All indicators are measured numerically and considered relatively to the respective scale of immigrant of the cases of analysis.
Naturalization rates shed light on previous permanent residency (as a prerequisite for naturalization), deportation rates could provide further insight on proportions
of undocumented residents from Central American countries.20 Both indicators
provide inference on the likelihood of an individual to be naturalized or deported respectively and could confirm the trends found in the data on the undocumented population. If immigrants from a specific country of origin were more likely to reside in the U.S. undocumented, I would expect to see more deportations (eventually due to Federal Immigration Crimes in the context of IIR IRA [1996]) and less naturalizations in comparison to other immigrants from CA between 1986 and 2000.21
Based on the developed theory, I am expecting a covariance between participation in the maras and legal status of residency. Furthermore, this would confirm the hypothesis that immigrants from the Northern triangle were more likely to reside in the U.S. without legal authorization than other immigrants from CA. Accordingly, immigrants from Belize, Costa Rica, Nicaragua and Panama would have been less likely to stay in the U.S. undocumented. 20 Foreign-born U.S. citizens can be deported when charged with particular felonies. Deportation is
significantly more unlikely than for permanent and temporary residents, as well as asylees and refugees (104th Cong. 1st sess., 1996). If an immigrant from CA was more likely to obtain a legal status and could apply for naturalization, he or she was more likely not deported (Miguel Cruz, 2010; Skarbek, 2011). Therefore the legal status affected the established deportation patterns, independently from its effect on gang participation. If gang participation among CA immigrants was equal, higher deportation numbers to the Northern triangle could still be explained by a higher number of Immigration Crimes by immigrants from the Northern triangle (Light et al., 2014).
21 Further insight on the number of applications for legal residency could be provided by the
statistics on asylum in the U.S. This would be an appropriate indicator because immigrants from CA were most likely going to apply for asylum in the U.S. to obtain a legal residency permit (MPI Data Hub, 2013; Terrazas, 2011).
15
Third, data on economic marginalization would be collected for Guatemalan immigrants and Nicaraguan immigrants in order to find evidence for the second hypothesis of this paper, stating that economic marginalization leads to the increased probability of gang participation. Based on previous findings, the hypothesis will be tested in two steps. First, data on the increased economic exclusion will be collected in order to confirm the claim that increased economic exclusion correlates with undocumented residency. Further, I will test how economic exclusion impacts gang participation.
The logic hereby is, as stated in the theoretical framework, that undocumented residency increases economic marginalization as it limits the possibilities to participate in the licit labor market and educational institutions in the U.S. (Rosenblum and Ruiz Soto, 2015; Warren and Warren, 2013b).
Indicators for economic marginalization are poverty levels, levels of income, levels of education and levels of employment. Measurement of economic exclusion is challenging for the particular purpose of this research. Separate data for the undocumented and documented share of a national immigrant cohort is not available. Thus, the averaged levels of per immigrant group will be used. The available data is based on the surveys by the U.S. Census Bureau, which includes both undocumented and documented immigrants.
For a comparison of Central American immigrant groups, which could have experienced similar levels of economic exclusion, the following predictions for the data can be made. I am expecting higher levels of economic marginalization for a group including a higher number of undocumented immigrants. I am expecting lower levels of economic marginalization for immigrant groups consisting of less undocumented aliens. Overall, I am expecting the undocumented population to be significantly underrepresented in the data. The limitations of the data on immigrant populations and especially in regard to undocumented immigrants will be discussed further in section 4.5.
In order to assess the relevance of the mechanism for the hypothesis and account for the limitations of the data, economic exclusion will be measured in relative trends.
16
Finally, I will test the relevance of economic exclusion for an increased vulnerability of joining or be recruited by the maras. Hence, this will further provide further causal inference on the hypothesized mechanism and confirm or reject the hypothesis that economic marginalization leads to the increased probability of gang participation.
Due to the purpose and scope of this paper, I will isolate a specific independent variable and economic mechanism and test its value for two particular cases. This will hopefully allow the assessment of their value for the proposed theory by employing the following methodology.
4.3.
Methodology
This chapter elaborates on the methods used to collect and analyze data in order to address the hypotheses and resultantly the research question of this paper.
This thesis is an attempt at a comprehensive exploration of a proposed covariance of legal residency and participation in the maras linked through economic exclusion. To test my theory and the presented hypotheses, I will conduct a cross-sectional comparative study of the Central American immigrant groups by country of origin and the respective probabilities to reside in the U.S. without documentation as well their participation in the U.S. maras. The utilization of mixed methods is appropriate as will be elaborated in the following.
First, I am going to employ a textual analysis to outline the historical migration patterns and characteristics from CA to the U.S. between 1970 and 1990. This is appropriate as I am aiming for descriptive inference and will be a first step to a broader analysis and serves as a background to the data collection.
The testing of the first hypothesis (H1: Immigrants from the Northern triangle were more likely to reside in the U.S. without legal authorization than other immigrants from CA) aims for causal inference. The primary concern of this part of the research is to explore a covariance among the variables.
Therefore, I will first measure the mara participation of immigrants from Belize, Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua and Panama in the U.S.
17
maras by collecting empirical evidence through the textual analysis of primary research. The measurement will rely on qualitative data.
Proceeding, I will collect quantitative data on the foreign-born population in the U.S. from CA as well as legally admitted individuals, naturalizations and deportations in order to measure my independent variable.
Estimations on the undocumented population in the U.S. are measured numerically by subtracting the number of legal admissions per year stated by the U.S. Immigration and Naturalization Services of the Department of Homeland Security (INS) from the increase of the total foreign-born population per year conducted by the U.S. Census Bureau. The latter includes undocumented individuals in its survey, as it is not directly concerned with legal immigration matters, whereas the INS’ immigration statistics only entail legally admitted aliens. The calculated difference accounts for the estimated number of undocumented individuals entering per year, the cumulative results per year account for the total number (Bean et al., 2002). Additionally, I will rely on qualitative data to further analyze the explanations for a variation in legal statuses among immigrants from CA, supporting the results of the quantitative data. The isolation of the explanatory variable in this research allows its specific testing but logically results in the de-contextualization of the multi-factorial participation in the MS-13 and M-18. Legal status of residency for mara participation is prioritized over alternative variables for reasons discussed in section 5.3.
The research proceeds by collecting data on the indicators for economic exclusion defined in section 4.2. In order to do so, case studies on the specific immigrant groups from Guatemala and Nicaragua will be employed. This offers the ability to investigate the hard-to-define concept of economic exclusion in relation to mara participation in very specific contexts. Additionally, statistical evidence on units of analysis as well as the undocumented population in the U.S. will be extracted through the textual analysis of empirical evidence.
The final part of the research is concerned with the testing of the causal mechanism in order to test the second hypothesis (H2: Economic marginalization
18
leads to the increased probability of gang participation). Accordingly, I will utilize process tracing through the textual analysis of the empirical data. The analytical framework provided by the existing theories on gang participation (Bordua, 1961; Miguel Cruz, 2010; Miller, 1958; Vigil, 2003, 1988) will allow to analyze how economic exclusion could have increased the vulnerability for immigrant groups that were more likely to reside in the U.S. undocumented.
The utilization of counterfactuals will provide further evidence on the plausibility of the hypothesized mechanism and help confirm or reject the second hypothesis. The use of mixed methods seems appropriate to increase both validity and reliability of that data by building an argument on causal inference to test causal mechanisms. I am aware of the limitations inherent to qualitative small-N analysis such as lack of external validity in terms of generalization as well as omitted variable bias causing a spurious relationship among my variables and reducing causal inference. As I am primarily concerned with the particular mara-phenomenon, context-specific results are appropriate but have to be assessed critically.
4.4.
Case selection
Before collecting data and testing my hypotheses, I will elaborate on the selection of the cases and units of analysis employed in this paper.
In relation to the dependent and independent variables, data on the respective immigrant groups from Belize, Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua and Panama in the U.S. will be collected. A group-level analysis is considered helpful to understand if a covariance between immigrants’ participation in the maras and their respective probability to reside in the U.S. without documentation exists.
As stated in chapters 2 and 3, the formation of the maras is embedded in the context of the transnational migration pattern between the U.S. and CA. Immigrants from all countries in CA migrated to the U.S. The proliferation, fueled by migration, only affected the Northern triangle. In order to understand the proposed variation in the participation and its relevance for the further evolution
19
of the maras, all of the identified immigrant groups have to be incorporated in the research. Hence, this case selection seems appropriate.
The case studies will be dealing with immigrant groups from Guatemala and Nicaragua. This is partly based on the availability of data for these cases. Further, I am hoping to maximize the benefits of a qualitative analysis through the selection of these two countries for the following reasons.
Nicaraguan and Guatemalan immigrants share similar socioeconomic pre-dispositions for gang participation such as a history of violence, a national street gang culture as well as political end economic instabilities in their respective countries of origin. Whereas these factors could account for alternative explanatory variables for the varying participation in the U.S. maras, they do not equally hold for the selected cases and will be neglected in this research. Further do both cases seem appropriate, as they have experienced major migration flows to the United States unlike Belize, Costa Rica and Panama. The similar scale of immigration results in increased availability and comparability of the data.
The cases are expected show a variation in both the dependent and the independent variable.
Rogers and Baird (2015) and Rocha (2007) give theoretical evidence that Nicaraguan immigrants in the U.S. were more likely to be legally admitted and naturalized than other immigrants from CA and that Nicaraguans were not participating in the maras in the same extent as other immigrant groups.
The literature provides evidence that immigrants from Guatemala on the other hand were participating in the U.S. maras (Arana, 2005). The research will establish if migrants from Guatemala were less likely to obtain a legal residency permit. The case is expected to resemble the dynamics experienced by other immigrants from the Northern triangle after the initial formation of the maras. A case study investigation will examine if and how economic exclusion resulting from undocumented residency impacted the likelihood of joining the maras. Additionally, it will allow the in-depth study of potential alternative explanations. The case studies are further aiming to decrease the likelihood of a spurious
20
correlation among the variables (based on the case selection) by testing the underlying causal mechanisms.
The limitations to the selection of the cases inherently include that the case studies are case specific, resulting in limited generalizability. Nevertheless do they provide explanatory value for the developed theory and future research.
As the cases were selected based on the expected outcome, I am expecting a selection bias based on the expected variations in the variables. This will be incorporated in the analysis. Alternative explanatory variables and causal mechanisms will have to be incorporated and discussed to decrease this bias. Testing the first hypothesis for all seven CA cases therefore is important to see if it is valid apart from the expected correlation. The case study of Nicaraguan immigrants in the U.S. is still considered appropriate for providing information on mechanisms underlying the legal status of immigrants in relation to gang formation.
4.5.
Data collection and sources
The previous chapters have described how the research in this paper will be operationalized and have elaborated on the employed methodology as well as the cases. This section will discuss the limitations of the indicators of measurement for the respective variables as well as the quality of the available data and sources. The dependent variable is defined as the participation in the U.S. maras between 1980 and 2000. Participation is measured nominally as significant participation and no significant participation. The indicators and measurement scale try to increase the validity of the existing data for the proposed theory to the highest possible degree. The existing data is considered to be relatively robust as gang membership was not illegal in the U.S. Members of the maras were easily identified through their appearance (see figure 1) (Bruneau et al., 2011). Considering the reliability of the data, it could still be possible that despite lacking evidence, individuals from all Central American countries have participated in the maras 1980 and 2000. The selection of the cases for the case studies is further based on the reliability of data for Nicaraguan immigrants’ participation.
21
Sources for data on the participation in the maras is the primary anthropological research conducted about members of the U.S. maras (Gutiérrez, 2009) and the empirical evidence provided by the literature. I will mostly rely on data provided by Rodgers and Baird (2015b), Valdez (2011) and Vigil (1988, 2003).
The data is often serendipitous in its nature and lacks a substantive amount of reliable original research due to “intrinsic methodological difficulties involved in carrying out detailed primary research on gangs” (Rodgers, 2007). The difficult access to valid data on violence, crime and gang activity and the arguably often flawed body of existing information, with official statistics particularly inconsistent “deficient data collection, and the discretionary if not political use of crime data” (Wolf, 2012, p.68) makes data collection challenging.
A lack of precision in regard to the variations of the phenomena observed (the distinctions between the ethnicities of gang members for example) decreases its validity for this paper (Miguel Cruz, 2010; Oettler, 2011; Rodgers and Baird, 2015b; Rodgers and Hazen, 2015). The limitations in regard to the reliability of the data are acknowledged; the triangulation of differing sources and general skepticism seem appropriate when dealing with it.
The independent variable of this paper is the legal residency status of immigrants from CA in the U.S. between 1986 and 2000. The estimated scale and composition of the undocumented population in the U.S. is one indicator of measurement for this paper. Further indicators will be the statistics on naturalizations and deportations per year.
The U.S. Census Bureau provides the only official estimate of the undocumented population in the U.S. (1996, 2000, 2002). The INS publishes data on legal immigration into the U.S. per year, starting with the necessary differentiation
among countries of origins in 1986.22
There are severe limitations to the data on the undocumented population in particular, which decreases its reliability and validity. Major caveats include the biased reliability due to measurement errors (based on inconclusive data
22 The United Nations (2015), the Migration Policy Institute (MPI) and the PEW Hispanic Research
Center build their quantitative analyses on the existing data. This allows for comparability among the sources as the structural bias is reproduced.
22
collection per country of origin) and sampling errors (such as the underrepresentation of undocumented and marginalized individuals). As a result, the literature argues that the actual scale of undocumented individuals is higher than the estimates (Bean et al., 2002; MPI, n.d.; Passel, 2006). The errors are expected to be systemic, which does not reduce the validity of the data.
The calculations of naturalized and deported aliens will rely on the data provided by the INS in order to achieve validity by incorporating the specific systemic collection and measurement errors of the source. This allows for descriptive inference on the proportions of CA immigrants within the particular, isolated data. While this paper tries to identify which immigrant groups were more likely to stay in the U.S. undocumented, it is acknowledged that this by no means indicate that all immigrants from a particular country of origin were undocumented immigrants. It merely entails a proportional probability in comparison to other immigrant groups.
To test the degree to which economic exclusion holds as a causal mechanism, data on Guatemalan and Nicaraguan immigrants in the U.S., provided by the little existing research (InSight Crime, 2016; Meyer, 2016; Núñez, 1996; Rocha, 2013; Rodgers, 2013), will be analyzed. Indicators of measurement have been addressed in Chapter 4.2.
The limitations of these indictors and data include the following. The data collection for this paper relies on a small amount of existing sources, which is partly owed to the relevant time period (1980 and 2000). Statistical research increasingly picked up after 2005 and developed more specific to the respective differences on the group level.
A study by Gustavo López, published by the PEW Hispanic Research Center in 2015, marks the first attempt to show distinctive characteristics of the Guatemalan and Nicaraguan foreign-born population in the U.S. The empirical data is based on a survey in 2013 and a longitudinal comparison of statistics provided by the U.S. Census Bureau. The results are expected to face limited reliability (due to sampling and resulting measurement errors illustrated in the paragraph above). The data is
23
relatively robust in terms of cross-sectional inference on the averaged economic profile of each cohort based on the expected systemic errors.
The validity of this data for the time period relevant to this paper is confined but considered robust enough to provide some insight for the time prior to 2000. The sample includes undocumented and documented U.S. and foreign-born immigrants in the U.S., some of which having resided in the U.S. since 1990. Considering the immigration reforms and trends after 1996, the scale and demographic profiles of immigrant groups are expected to have changed but the general statistical profiles are expected to have remained relatively stable between 1990 and 2013 (Bean et al., 2002; López, 2015a; Passel, 2005).
I therefore argue that the data provided by the PEW study does not allow for direct inference on immigrant groups from Guatemala and Nicaragua in the U.S. before 2000 but, especially based on its incorporation of the U.S. Census Bureau data from 2000, allows for some general claims. I expect the sample groups of Nicaraguan and Guatemalan immigrants in Lopez research (2015) to be influenced by distinctive characteristics and dynamics.
The Nicaraguan immigrant population is more likely to have built up since the 1990s and to include individuals, which were already residing in the U.S. prior to 2000. I expect that this had a positive effect on the low levels of economic exclusion as a bigger proportion of the sample has been in the U.S. legally, increasing the chances of social and economic integration. The levels of economic exclusion could have been slightly higher in the 1990s, as economic integration was less advanced.
The sample of Guatemalan immigrants is more likely to include more first-generation undocumented immigrants (considering the increased numbers of illegal immigration from Guatemala after 2000) and re-immigrated individuals. I expect this to have affected the economic integration negatively. Eventually, economic exclusion was a little lower between 1990 and 2000, as the undocumented share of the Guatemalan immigrant population was smaller.
Data on alternative key variables and competing causal mechanisms will be collected synchronously and analyzed in the discussion part of this paper.
24
4.6.
Background of the cases
The following chapter will give an overview of the migration pattern between CA and the U.S. in order to increase the contextual interpretation of the research results.
Central America became increasingly relevant for the hemispheric migration patterns in the 1980s, when a decade of armed conflicts in the region led to an exodus towards “El Norte” (see Appendix, Figure 2). Most emigrants moved towards the U.S., Canada and, to a much lesser degree, to the more stable countries within the region, such as Belize and Costa Rica (Mahler and Ugrina, 2006). These transnational migration flows became subject to a increasing amount of scholarship, when numbers of emigrants from El Salvador, Guatemala and Nicaragua to the U.S. multiplied in the 1980s, followed by an influx in emigrants from Honduras in the 1990s (Mahler and Ugrina, 2006; Schulkin, 2012).
Between 1980 and 2010, the immigrant population from CA in general in the U.S. had increased from 345.655 to an estimated 2.020.368 people. Approximately 70% (800.095) of the foreign-born individuals from CA residing in the U.S. in 1990 were born in the Northern triangle, another 14,9 % were Nicaraguan and another 15,1% stated Belize, Costa Rica or Panama as their countries of birth. By 2000, the share of the Central American immigrant population from the Northern triangle had increased to 78,2% (1.580.853), whereas only 10,9% of the immigrant population were Nicaraguan (see Table 1) (Gibson and Jung, 2006; INS, 1997).
25 Table 1. Total immigrant Population from Central America from 1970- 2000, per county of origin23 Immigrant population from Central America, per year Countries of origin 1970 1980 1990 2000 Belize 8.860 14.436 29.957 42.133 Costa Rica 16.691 29.639 43.530 71.870 El Salvador 15.717 94.447 465.433 817.336 Guatemala 17.356 63.073 225.739 480.665 Honduras 19.118 39.154 108.923 282.852 Nicaragua 16.128 44.166 168.659 220.335 Panama 20.046 60.740 85.737 105.177 total 113.917 345.655 1.127.978 2.020.368 Source: Data from U.S. Census Bureau 2000
Overall, immigrants from CA faced substantial challenges in the U.S. On average,
they were significantly less educated but more likely to get employed, compared to
the overall foreign- and U.S.-born populations.24 In 1990, Central American-born
residents of the U.S. were more likely to speak English poorly and to be single males than other Hispanic groups (Gibson and Jung, 2006; Lesser and Batalova,
2017; MPI Data Hub, 2013).
Geographically, Central American immigrants were heavily concentrated in California, Texas and Florida, followed by New Orleans and Washington, DC. (García, 2006; Gzesh, 2006). The literature suggests that destinations of migration within the U.S. varied by countries of origin, conclusive and reliable data on the dispersion is not available.
23 The term "immigrants" (or "foreign born") refers to “people residing in the United States who
were not U.S. citizens at birth. This population includes naturalized citizens, lawful permanent residents (LPRs), certain legal nonimmigrants (e.g., persons on student or work visas), those admitted under refugee or asylum status, and persons illegally residing in the United States” (U.S. Census Bureau).
26
Central American immigrants in the U.S. were less likely to be naturalized citizens
than the overall foreign-born population, possibly due to the large number of TPS
beneficiaries and unauthorized aliens ineligible for U.S. citizenship (Lesser and Batalova, 2017). Nonetheless did about 500,000 Central American immigrants become lawful permanent residents (LPRs) and U.S. citizens after the enforcement
of the Immigration Reform and Control Act (IRCA) in 1986.
Research on intraregional migration and the immigrant population in the U.S. has demonstrated that migration patterns vary distinctively (Gibson and Jung, 2006;
Mahler and Ugrina, 2006; Meyer, 2016; OAS, 2015).
The scale of immigration from Belize, Costa Rica and Panama in the U.S. was
significantly smaller (see table 1).25 Emigration from Belize, Costa Rica and Panama
was dominated by the middle class (Gibson and Jung, 2006; Schulkin, 2012); a significant proportion of visa applications were work- and family-related compared to other sending countries from CA (Casas-Zamora, 2016; MPI Data Hub, 2013; Muggah, 2014).
Emigrating Costa Ricans, Panamanians and Belizeans were higher educated and
had better access to financial resources than other Central American migrants.
Empirical data shows that large numbers of people from El Salvador, Guatemala and Honduras moved to the U.S. in the 1970s and 1980s due to civil wars, political violence and economic hardship (Meyer, 2016). Immigrants from Guatemala and El Salvador predominantly applied for asylum in the U.S., as most did not qualify for other admission categories (Schulkin, 2012). Nicaraguan immigrants on average were low-skilled, seeking asylum for political and economic reasons (Núñez, 1996; Schulkin, 2012).
25 This could be conditioned by less push-factors in the sending countries including an advanced
socioeconomic development. Emigrants were less likely to be displaced for political and economic reasons and more likely to have obtained legal status of residence before leaving their countries of origins, considering a prevailing rationale for voluntary emigration (Migration Policy Institute (MPI), n.d.). Due to the less urgent need to migrate less individuals are expected to have migrated illegally. These mechanisms will not be tested in this study.
27
5. Research Results
Chapter 5 of this thesis will first outline the expectation in regard to the research results and state the findings. It will proceed by analyzing these in the context of the hypotheses and the developed theory. The last section of this chapter discusses alternative explanations for the observed outcomes and discusses their implications for future research.
5.1.
Expected findings
Based on the theory, this paper aims to test the following hypotheses: H1: Immigrants from the Northern triangle were more likely to reside in the U.S. without legal authorization than other immigrants from CA. H2: Economic marginalization leads to the increased probability of gang participation.Building on the literature review and the theoretical framework, I expect to find evidence demonstrating that immigrants from the Northern triangle were more likely to reside in the U.S. without a legal permit than other immigrants from CA. If the hypothesis were supported, it had to be tested whether economic exclusion resulting from undocumented residency holds as a causal mechanism for the cases of Guatemalan and Nicaraguan immigrants. Data collection on the mechanism linking the dependent and independent variables could empirically sustain the measured observations and reduce the possibility that the correlation is spurious. According to the theory, I am expecting to find evidence that Guatemalan immigrants in the U.S., on average, were facing higher levels of economic exclusion than Nicaraguan immigrants, on average. Indications could include lower income levels, lower employment rates and lower education levels for Guatemalan immigrants, possibly due to the higher numbers of unauthorized immigrants. In the following analysis, I am expecting to find evidence for the impact of economic exclusion on the increased voluntary mara participation and coercive recruitment. The first hypothesis will be falsified if there is no empirical evidence supporting the claim immigrants from the Northern triangle were more likely to stay in the
28
U.S. without documentation. The second hypothesis will be falsified if the research results shows that undocumented immigrants in general and Guatemalan immigrants, on average, in particular were not experiencing higher levels of economic exclusion. The proposed causal mechanism would be rejected if I found that undocumented immigrants, even if they were economically more excluded, did not join the maras for that reason.
Competing independent variables for mara participation had to be tested to identify the causal relationship(s) affecting the outcome. These could include the geographical location of migrant cohorts. Alternative mechanisms include social exclusion resulting from undocumented residency.
5.1.1. Participation in the maras
The M-18, initially founded by Mexican immigrants, started to accept Hispanic immigrants indiscriminately in the 1980s. The brutal rivalry with the emerging MS-13, founded by Salvadorian immigrants, led to the increased recruiting of members and catalyzed the growing scale of both gangs. Due to the local absence of law-enforcement and federal institutions, many low-income Hispanic neighborhoods were soon controlled by one of the two gangs (Vigil, 2014). By 1990, the M-18 and the MS-13 were consisting mainly of Guatemalans, Hondurans, Mexicans and Salvadorians (Ribando Seelke, 2016; InSight Crime, 2017; Wolf, 2012).
The data suggests that Nicaraguans have joined the maras “in very small numbers” (Gutiérrez, 2009, p.68), as Nicaraguans “did not participate in the local gang scene to the same extent” (Rodgers and Baird, 2015b, p.482) as other Central American immigrants.
Whereas the literature suggests that Central American immigrants were accepted into the maras equally (Rodgers and Baird, 2015b), the existing empirics show that participation has not reached significant levels for all seven observed cases. I do not find any evidence on membership of immigrants from Belize, Costa Rica and Panama.
