Foiled Aspirations: The Influence of Unauthorized Status on the Educational Expectations of Latino Immigrant Youth

Abstract

Latino immigrant adolescents have the highest high school dropout rates of any race-ethnic or nativity group in the United States. One potential reason for high dropout rates among Latino immigrant youth is that many are unauthorized entrants. These unauthorized Latino immigrant youth have few opportunities to attend college, and, as they become aware of barriers to their educational progress and employment, they may lower their educational expectations. Using data from the Latino Adolescent Migration, Health, and Adaptation Project (N=275), we examine the association of unauthorized entry into the U.S. with the educational expectations of Latino immigrant youth. We find that adolescents entering the U.S. without authorization have lower educational expectations than those who enter with authorization. These differences in their expectations persist after controlling for differences in their pre-migration, migration, and post-migration experiences. Policies and programs that reduce barriers to higher education and labor market opportunities can potentially help to foster higher educational expectations among unauthorized immigrant youth and may promote their high school completion.

INTRODUCTION

Latino youth have the highest high school dropout rates in the United States. Recently available data indicate that 21% of Latinos ages 16–24, 8% of non-Latino blacks, 6% of non-Latino Asians, and 5% of non-Latino whites had dropped out of high school (Aud et al. 2010). Among Latino youth, the foreign-born have significantly higher dropout rates than the U.S.-born (34% vs. 11%; Aud et al. 2010). One potential reason for high dropout rates among Latino immigrant youth is that many are unauthorized entrants. Approximately, 1.4 million foreign-born Latino children live in the United States and national estimates suggest that 80–85% of those who immigrated to the U.S. in the past 10 years entered without authorization (Motel 2012; Passel and Cohn 2008).

These unauthorized Latino immigrant youth have few opportunities to attend college, and, as they become aware of barriers to their educational progress and employment, they may lower their educational expectations (Flores and Chapas 2009; Abrego and Gonzales 2010; Suarez-Orozco et al. 2011). Unauthorized entrants cannot apply for federally-qualified student loans to attend college and cannot enroll in public universities or community colleges as in-state residents (Abrego and Gonzales 2010; Androff et al. 2011; Flores and Chapas 2009). Thus, in most states, they must pay out-of-state tuition and face significant financial barriers to college entry even after meeting the academic qualifications for admission. Rates of high school completion and educational expectations among immigrant children, especially unauthorized entrants, may be further depressed by their limited economic gains from high school completion. Even with a high school degree, unauthorized youth often cannot obtain jobs in the formal economy. Instead they obtain jobs in the informal or underground economy where they earn lower wages and can be easily victimized by discriminatory practices (Gonzales 2011; Massey 1987; Rivera-Batiz 1999). Overall, limited labor market opportunities and educational policies creating financial and legal barriers for unauthorized immigrants to enter college can dampen their educational expectations and decrease their motivation to graduate from high school (Abrego 2006; Suarez-Orozco et al. 2011).

In this study, we examine the influence of unauthorized entry into the U.S. on the educational expectations of Latino immigrant youth (ages 12–18) in North Carolina. We describe differences in, adolescents’ educational aspirations, adolescents’ educational expectations, and parents’ educational expectations for children with and without authorization to enter the U.S. We then evaluate how pre-migration, migration, and post-migration experiences differ for children who enter the US with and without authorization. Finally, we evaluate the association between authorization status and educational expectations after controlling for key aspects of their pre-migration, migration, and post-migration experiences. We find that the lowered expectations of immigrant youth without authorization cannot be accounted for by differences in their countries of origin, class origins or socio-economic backgrounds, or school experiences in the U.S. For children without legal status in the U.S., educational and labor market opportunities are blocked and they lower their educational expectations accordingly.

With data on the authorization status of Latino immigrant youth, our analysis makes a unique contribution to the literature on educational attainment and the segmented assimilation of Latino immigrant youth in the US southeast. First, few studies have evaluated how educational aspirations or expectations vary among immigrant children or their parents by authorization status. Studies suggest that substantial heterogeneity exists within Latino populations not only by immigrant generation but also by country of origin, English language usage, and family socioeconomic status (SES). These variations influence both educational expectations and performance (Bohon et al. 2006; Feliciano 2006; Pong and Landale 2012). Yet authorization status is typically an unobserved aspect of this heterogeneity. It is not measured in most national datasets. Given that 1 out of every 20 Latino children in the U.S. is unauthorized (Passell 2011), this critical aspect of social stratification in the U.S. needs to be further explored and unpacked.

Second, with one recent exception (Greenman and Hall 2013), studies evaluating the influence of authorization status on education have focused on immigrant populations living in California or other states in the U.S. southwest with a history of immigration from Mexico and Central America (Abrego 2006; Bean et al. 2011; Gonzales 2011; Menjivar 2008). This study focuses on a population that is not well represented in national datasets or in previous research with unauthorized immigrants, immigrant youth in the U.S. Southeast. Between 1990 and 2000, North Carolina had the fastest growing Latino population and fastest growing immigrant population in the United States (Guzman 2001; MPI 2013). As such, it has been the focus of much research on emerging immigrant states. Immigrant children with and without legal status who live in these emerging immigrant states may face additional challenges adjusting to life in the U.S. than those who live in historical receiving states. Co-ethnic communities which can help new immigrants transition are small; schools and other health and human service providers have little experience working with immigrant populations; and, at least among emerging immigrant states in the South, immigrant parents face substantial legal restrictions on access to health and human services such as the Supplemental Nutrition Assistance Program and Medicaid (Ko and Perreira 2010; Perreira et al. 2012; Valencia and Johnson 2006).

Third, this study contributes to the small existing body of literature on legal status and the educational progress of Latino immigrant youth by examining whether the de-motivation of unauthorized youth begins early in high school with a sample of Latino student who are mostly age 16 or younger. Greenman and Hall (2013) find no significant association between legal status and high school dropout rates among Latino immigrant adults (ages 18–24). Nevertheless, among immigrant adults (ages 18–24) who graduate from high school, those with legal status in the U.S. have higher odds of enrolling in college than those without legal status. Similarly, Bean et al. (2011) find a significant educational premium for second-generation adults whose parents have attained some combination of legal status relative to those whose parents have remained unauthorized since entry. Finally, in their ethnographic work, Abrego (2006), Gonzales (2011) find evidence that undocumented youth begin to give up on their educational dreams and aspirations during high school as they transition into adulthood. By understanding what factors influence the formation of immigrant youth’s educational expectations and differences in the experiences of unauthorized and authorized youth, we can gain insight into the processes that shape their long-term educational attainment and economic well-being.

BACKGROUND

Drawing on a revised form of Sluzki’s stages of migration framework (Zuniga 2002; Ko and Perreira 2010) as well as segmented assimilation theory (Portes and Rumbaut 2001; Portes and Zhou 1993), we consider how unauthorized status and the migration process influences youths’ educational expectations. According to Sluzki, the migration process can be divided into three distinct stages. Experiences in each stage influences the next and work together to affect the health and development of youth (Zuniga 2002). In the pre-migration stage, families make the decision to migrate. Their motivations for migration reflect both their socioeconomic backgrounds and opportunities in their countries of origin as well as the potential risks and rewards of migrating. In the migration stage, the physical act of migrating occurs. These migration experiences will vary by age and the immigration pathway (e.g., refugee, temporary work visa, family visa) available to individuals (Rumbaut 2004). For some immigrants, especially those unable to enter the U.S. legally, the migration stage can involve considerable risk and exposure to trauma (Perreira and Ornelas 2013). In the post-migration stage, immigrants settle in the U.S. and begin to adapt to a new culture and language, to new family dynamics, to new school environments, and to new neighborhood or community environments.

In considering the educational adaptation of children of immigrants, segmented assimilation theory emphasizes country of origin and the roles of (1) pre-migration parental human capital or parent’s class origins (i.e. income and education), (2) post-migration family structure and English language acquisition, and (3) post-migration social environments or contexts of reception (Portes and Rumbaut 2001; Portes and Rivas 2011). In considering the context of reception, three aspects are paramount–governmental inclusion/exclusion, social acceptance/rejection, and co-ethnic community (Portes and Rumbaut 2001; Portes and Rivas 2011). These factors can work together to promote either the upward socio-economic assimilation of immigrant youth or their downward assimilation.

Assimilation can happen along many dimensions (e.g., language, adoption of cultural norms, health). In this study, we focus on educational assimilation and educational aspirations and expectations as an initial indicator of educational assimilation. Low educational aspirations/expectations can portend dropping out of high school and close the door to upward socio-economic mobility (Perreira, Harris, Lee 2006; Portes and Rivas 2011). High educational aspirations/expectations can foreshadow high school completion, college entry, and the potential for upward socio-economic mobility (Perreira, Harris, Lee 2006; Portes and Rivas 2011).

According to segmented assimilation theory, downward assimilation would be expected for children from low socio-economic backgrounds whose parents have low-levels of education; children who are unauthorized entrants experiencing a high level of governmental exclusion; children who experience racism and discrimination in their schools or neighborhoods after settlement; and children who live in communities where there are few co-ethnic resources. In other words, Latino immigrant children growing up in emerging immigrant communities in the Southeastern U.S. will be at high risk of downward assimilation. Moreover, those without legal status will be at higher risk of downward assimilation than those with legal status.

Educational Aspiration and Expectations

In research on educational attainment, two questions on educational motivations are frequently ascertained – educational aspirations and educational expectations. Questions on aspirations seek to identify how much education youth want to achieve in an ideal world without constraints. Questions on expectations, on the other hand, attempt to measure how much education youth will realistically achieve given the personal and financial barriers they face to continuing their educations. Because they reflect more concrete goals, youth typically report lower expectations than aspirations (Bohon et al. 2006; Boxer et al. 2011; Goldenberg et al. 2001). For this reason, educational expectations are also more likely to vary by authorization status and are therefore the focus of this analysis.

Previous research has found significant differences in both educational aspirations and expectations by race-ethnicity as well as by nativity (Bohon et al. 2006; Glick and White 2004; Kao and Tienda 1995, 1998). Data from the National Educational Longitudinal Study (NELS) show that, in comparison to Asians (Boys: 72%, Girls: 79%), Blacks (Boys: 48%, Girls: 56%), or Whites (Boys: 58%, Girls: 65%), fewer Hispanic youth aspired to college or more (Boys: 45%; Girls: 47%) in tenth grade (Kao and Tienda 1998). However, among Hispanics and Asians, first- and second- generation youth report significantly higher college aspirations than third+ generation youth (Kao and Tienda 1995). First-generation youth are foreign-born youth with foreign-born parents; second-generation youth are U.S.-born youth with foreign-born parents; and third+ generation youth are U.S.-born youth with U.S.-born parents. Immigrant parents also report higher educational aspirations and expectations for their children than native-born parents of every race-ethnicity (Goldenberg et al. 2001; Raleigh and Kao 2013). Moreover, they retain this optimism toward the educational advancement of their children over time, even as they potentially become more aware of the legal, linguistic, and financial challenges faced by their children (Turney and Kao 2009). Few studies, however, have evaluated how adolescent’s expectations, adolescent’s aspirations, or parent’s expectations vary by children’s legal status.

Pre-Migration Experiences

Research has consistently found that adolescents’ early life experiences shape their future choices and outcomes (Eccles 1994). However, most studies of immigrant educational progress focus entirely on the post-migration experiences of children (Crosnoe and Lopez Turley 2011). The only pre-migration aspects of a child’s experience typically considered are children’s countries of origin and their parents’ class origins or socio-economic backgrounds. Nevertheless, these two factors do explain much of the variation in economic outcomes among the first- and second generation children of immigrants. Children from Mexico typically have lower educational expectations, lower academic achievement, and higher high school dropout rates than immigrant children from other Latin American, Asian or African origins (Crosnoe and Lopez Turley 2011; Feliciano 2006; Perreira, Lee, and Harris 2006). In addition, immigrant children whose parents have low levels of education typically have lower educational expectations, lower academic achievement, and higher high school dropout rates than those with more highly educated parents (Crosnoe and Turley 2011; Feliciano 2006; Perreira, Lee, and Harris 2006; Pong and Landale 2012). Given their strong associations with educational outcomes, this study considers children’s country of origin and their parent’s socioeconomic backgrounds. We expect children from Mexican origins, children from higher poverty backgrounds, and children whose parents have completed less than high school to have lower educational expectations.

We also recognize that other family circumstances prior to migration, previous experiences visiting the U.S, and motivations for migration can potentially influence children’s ease of adapting to life in the U.S. and their educational progress. Separation from one’s parents (or caregivers) has been associated with detrimental effects on immigrant adolescents’ educational success (Suarez-Orozco and Suarez-Orozco 2001); previous tourist visits to the US have been positively associated with immigrants’ aspirations to attend a 4-year college (Kandel and Kao 2001); and children who are brought to the U.S. by their parents explicitly to improve their educational opportunities may be a selective group whose parents maintain a high level of involvement in their school activities and strongly encourage them to excel (Fernandez-Kelly and Portes 2008; Hagelskapmp, Suarez-Orozco, and Hughes 2010). Thus, we expect that children who experienced no separation from their parents prior to migrating, who had previously visited the U.S., and who moved to the U.S. primarily for educational opportunities will have higher educational expectations.

Migration Experiences

Immigrants vary in the degree of stress and the type of stressors they experience throughout the migration process (Garcia Coll and Magnuson 1997; Kuperminc et al. 2009). For children of immigrants, the degree and type of stress experienced throughout the migration process depends mainly upon their legal status and their age at migration (Suarez-Orozco and Suarez-Orozco 2001; Rumbaut 2004). Youth who enter the U.S. without authorization are more likely to have been exposed to trauma both prior to or during their travels to the U.S (Perreira and Ornelas 2013). The types of trauma experienced by unauthorized immigrant youth include witnessing murder, assault, rape, and extreme food deprivation or dehydration -- experiences that are widely recognized as traumas by mental health experts (Kessler et al., 1995; Copeland et al., 2007). Childhood trauma and victimization have been previously associated with impaired school performance and lower educational attainment (Breslau et al. 2008, Levels, Dronkers, and Kraaykamp 2008; Macmillian 2001). From a developmental perspective, children who arrive during early (ages 0–5) or middle (ages 6–12) childhood may have an easier time adapting – learning new languages, navigating new social systems, and forming new friendships (Garcia Coll and Magnuson 1997; Kuperminc et al 2009). In comparison, children who arrive at older ages may migrate primarily to work and have lower educational expectations (Chiswick and DeBurman 2004; Rumbaut 2004). In this analysis, we measure exposure to trauma prior to and during the migration process and age at migration. We hypothesize that immigrant children arriving to the U.S. at older ages and immigrant children who have experienced trauma will have lower educational expectations.

Post-Migration Experiences

Upon settlement in the U.S., the experiences of immigrant youth are shaped by federal and state policies which determine their access to resources, influence their patterns of settlement, and affect their exposure to stress and discrimination in the U.S. (Androff et al. 2011; Dreby 2012). These policies range from restrictions on their access to public assistance benefits to local enforcement of immigration laws to state restrictions on access to driver’s licenses for unauthorized immigrants (Perreira et al. 2012; Jones-Correa 2012). As argued by segmented assimilation theory, these policies create reception contexts which handicap the integration of unauthorized adult immigrants and their children (Dreby 2012; Portes and Rumbaut 2001). Because this study focuses on one state, we do not have sufficient variation in the sample to identify the effects of these types of policies. Therefore, we focus on students' reports of experiences in their neighborhood and school environments after migration.

In their neighborhoods and schools, immigrant children become exposed to U.S. cultural norms (Suarez-Orozco and Suarez-Orozco 2001; Portes and Rumbaut 2001). Positive social interactions within these environments can promote optimism and achievement, whereas negative social interactions can demoralize youth and lead to lowered educational expectations (Guyll et al. 2010; Perreira, Fuligni, and Potochnick 2010; Stone and Han 2005). Additionally, the presence of a co-ethnic community able to provide supportive resources that help immigrants adapt and “learn the ropes” can facilitate children’s educational progress and promote higher expectations (Portes and Rumbaut 2001; Perreira, Lee, and Harris 2006). Previous research demonstrates that perceptions of discrimination or unequal opportunity in the U.S., experiences of discrimination, neighborhood disorder, and safety problems in schools reduce educational expectations and attainment for all children, but especially minority and immigrant youth (Crosnoe 2005; DeGarmo and Martinez 2006; Milam et al. 2010, Valenzuela 1999). Evidence that social support from teachers and peers or the presence of a co-ethnic community can counter these stressors and improve educational expectations or motivations is mixed (DeGarmo and Martinez 2006; Kuperminc et al. 2009; Perreira, Lee, and Harris 2006; Perreira, Fuligni, and Potochnick 2010).

For this analysis, we consider the effects of neighborhood disorder (i.e. lack of safety), discrimination, and school disorder (i.e. lack of safety). We hypothesize that they will be negatively associated with educational expectations. We further consider the effects of teacher support and school satisfaction with the hypothesize that they will be positively associated with educational expectations. Lastly, there is little variation in the size of co-ethnic communities in our data. However, we evaluate the effects of living in an urban versus rural environment. Several urban communities in North Carolina have established community-based organizations which support the adaptation of immigrants. In addition, schools in urban North Carolina typically have more resources for immigrant students (NCDPI 2009). Thus, we expect that children living in urban communities will have higher educational expectations.

Lastly, we consider the family environment after migration. As the primary context of socialization and child development, family has always taken center stage in explaining the educational expectations of children, including immigrants (Feliciano 2006; Glick and White 2004; Kao and Tienda 1995, 1998). Families headed by two parents and those with higher incomes in the U.S. can provide more resources to their children and potentially set higher educational expectations for them (Glick and White 2004; Kao and Tienda 1995, 1998; Pong and Landale 2012). The psychosocial aspects of the family can also influence youth’s educational expectations (Portes and Rumaut 2001; Valenzuela and Dornbusch 1994). In particular, often referred to as familism, the loyalty, respect, and support that family members have for one another can promote resiliency in children and lead children to set high expectations for themselves out of a sense of duty and obligation to their parents and the sacrifices their parents have made (Gonzales et al. 2011; Crosnoe and Turley 2011). Lastly, the acculturation of parents and children can influence their capacities to navigate school environments and obtain resources that will allow them to achieve their aspirations (Glick and White 2004; Portes and Rumbaut 2001; Ko and Perreira 2010). We measure both family structure and familism with the expectation that children from two-parent families and families with higher levels of familism will have higher educational expectations. We consider three measures of acculturation (i.e. psychological acculturation, English language use, and years in the U.S.) and expect each to be associated with higher educational expectations.

DATA AND METHODS

Data

We used data from the Latino Adolescent Migration, Health, and Adaptation Project (LAMHA), a population-based study of 281 first-generation Latino immigrant youth (ages 12–18) enrolled in middle or high schools in North Carolina between 2004 and 2006. To ensure a representative sample, data were collected through a stratified random cluster design. A total of four urban and six rural school districts including 11 high schools and 14 middle schools participated. All foreign-born students from Latin America or the Caribbean were invited to participate in the study. However, only one youth per household could participate. During home visits, adolescents and their parents completed an interview-administered survey in their preferred language (English or Spanish). The response rate was 69%. Further details on the sampling design have been comprehensively described elsewhere (Chantala and Perreira 2007). The LAMHA study was approved by the Institutional Review Board of the University of North Carolina at Chapel Hill.

Sample

In 2004–05, when data collection for this study began, approximately 583,087 Latinos resided in North Carolina and 58% were foreign-born (U.S. Census 2005). The majority of these foreign-born Latinos had moved to North Carolina from Mexico (73%); had entered the U.S. without authorization (76%); and had lived in the U.S. for fewer than 5 years (42%; Kasarda and Johnson 2006; Passel and Cohen 2011). At this same time, approximately 98,366 Latino children lived in North Carolina and 21% were foreign-born non-citizens (U.S. Census 2005). Based on various estimates, between 15,000–19,000 of these Latino children had entered the U.S. without authorization (Passel and Cohen 2011).

Like the foreign-born Latino population more generally, the majority (73%) of Latino adolescent immigrants in the LAHMA sample came from Mexico (74%). Participants not from Mexico had come from other Central American, South American, or Caribbean countries. Seventy-four percent (N=213) of the LAMHA sample entered the U.S. without authorization and most (93%) had parents who were also unauthorized entrants. Adolescents had lived in the U.S. for fewer than five years (57%) and many (59%) lived in poverty as defined by 2006 Federal Poverty Levels. The majority of youth primarily utilized English in their everyday lives (71%); whereas the majority of their parents primarily utilized Spanish (55%). Youth in our sample did not live in communities with sizable co-ethnic populations. Only 8.6% of the population was Latino in urban communities whereas only 6.3% was Latino in rural communities. After deleting missing observations on the dependent variable (N=6), our study sample for this analysis of educational expectations consisted of 275 Latino immigrant adolescents. In logit analyses, our sample was reduced to 271 adolescents due to missing values on covariates.

Measures

We derived all measures utilized in this study from either adolescent or parent interviews. All measures included in the LAMHA surveys had been previously evaluated for use with Latino and Spanish-speaking populations. Several were adapted from the Children of Immigrant Longitudinal Survey (CILS; Portes and Rumbaut 2001). For all scales used in this analysis, we report internal reliability measures from the LAMHA sample. We also summarize significant correlations over .20 (available upon request) in the text below.

Educational aspirations and expectations

To measure aspirations, youth were asked, “What is the highest level of education you would like to achieve?” To measure expectations, youth and their parents were asked, “Realistically speaking, what is the highest level of education you think you [your child] will achieve?” The final response categories for each measure (i.e. youth’s aspirations, youth’s expectations, and parent’s expectations) included: (1) high school or less, (2) vocational/trade school, (3) a bachelor’s degree, and (4) graduate degree.

Unauthorized Status

Our primary variable of interest is whether the student moved to the U.S. without authorization (1=yes, 0=no). We asked parents what kind of documentation or visa their children had when they first entered the U.S. and whether they crossed the U.S. border with the assistance of a smuggler. These questions were asked at the end of the survey after reminding participants that the data would be de-identified.

Pre-Migration Experiences

Utilizing data from the parent interview, we identified whether adolescents had moved to the U.S. from Mexico (1=yes, 0=no), had completed a high school education (1=yes, 0=no), and had experienced living in extreme poverty prior to migration (1=yes, 0=no). Our measurement of extreme poverty followed guidelines from Mexico (CONEVAL, 2010) where poverty is indicated when a parent reports living in a home with dirt floors or no indoor plumbing. Poverty prior to migration and parents’ high school completion were highly correlated (r=−.71). Combining data from the parent and adolescent interview, we also identified whether the family had moved primarily for a better education or to attend a U.S. school (1=yes, 0=no), primarily for work (1=yes, 0=no), or primarily to reunify with family (1=yes, 0=no). Based on adolescents’ interviews, we measured separation from both parents (1=yes, 0=no), mother only (1=yes, 0=no), or father only (1=yes, 0=no) prior to migration; and previous visits to the U.S. prior to migration (1=yes, 0=no). Separation from a mother was correlated with separation from a father (r =.64) and with living in extreme poverty prior to migration (r=.31).

Migration Experiences

Based on adolescents’ interviews, we measured youths’ age at migration (<age 6=1; <age 13=1). We also identified whether youth had directly experienced trauma prior to moving to the U.S. or during their journey to the U.S. by asking adolescents about whether they had experienced specific types of traumatic events such as the murder of a family member, being a victim of a violent crime, or being a victim of natural disaster or war. Adolescents were also prompted through an open-ended question to share additional details and other traumatic experiences. Respondents discussed their experiences of domestic violence, assault, and severe illness or injury (e.g., dehydration and hunger) in response to these open-ended questions. In accordance with the DSM-IV definition of trauma, responses to all open-ended questions were coded as traumatic events when they involved “actual or threatened death or serious injury, or threat to the physical integrity of oneself or others” (American Psychiatric Association, 2000:427). Because it focuses on trauma directly experienced by respondents and associated with their migration experiences, our measure of trauma is more restrictive than measures focusing on lifetime trauma that has either been experienced directly or witnessed (Kessler et al., 1995; Copeland et al., 2007). For a detailed examination of the trauma experienced by youth who participate in the LAMHA study, see Perreira and Ornelas (2013). We do not have sufficient power to detect significant associations of any one particular type of trauma (e.g., witnessing murder vs. rape vs. dehydration) with educational aspirations. Therefore, we coded these experiences as any exposure(s) to trauma (1=yes, 0=no). Entering the U.S. at age 13 or older correlated with exposure to trauma (r = −.24).

Post-Migration Experiences

At the community level, we first identified whether adolescents lived in a primarily urban county (1=yes, 0=no) where over 50% of the population lived in an urbanized area. Second, we measured neighborhood disorder with an index based on an average of five items on whether parents indicated that racial tensions, delinquency, crimes, gangs, or drugs were not a problem (=0), a minor problem (=1), or a major problem (=2) in their neighborhood (Oropesa 2012). Neighborhoods where this index was greater than 1 were coded as somewhat unsafe (=1; α_LAMHA = .79). Third, we followed Portes and Rumbaut (2001) and measured adolescents’ perceptions of discrimination in the U.S. based on the average response to three statements such as racial discrimination limits economic opportunity in the U.S. (1 = Disagree a lot to 5=Agree a lot; α_LAMHA = .58). We also identified whether adolescents reported having personally experienced discrimination in the U.S. because of their race/ethnicity (1=yes, 0=no). Among these measures, only urban and unsafe neighborhood were correlated (r=.29).

Based on additional measures from the School Success Profile (Bowen et al. 2005), we evaluated several aspects of adolescents’ school experience including school satisfaction, school safety, teacher support, and general social support. The sum of seven true-false statements (e.g., I enjoy going to school), school satisfaction ranged from 0–7 with a high internal consistency as measured by the Kuder-Richardson 20 formula (KR-20) for count data (KR-20 _LAMHA = .68). The sum of eleven true-false statements (e.g., My teachers really care about me), teacher support ranged from 0–11 with a high internal consistency (KR-20 _LAMHA = .79). The sum of eight yes-no questions (e.g., Are there people you talk to at least weekly who listen to you without giving advice or judging you), social support ranged from 0–8 with a good internal consistency (KR-20_LAMHA =.59). Finally, school safety problems (also referred to as disorder) were assessed using the average of nine items indicating how much of a problem gangs, delinquency, vandalism, substance use, racial tensions and other factors were at their school (KR-20_LAMHA = .89). Schools with some problems on average were coded as somewhat unsafe (1=yes, 0=no). Among these measures, only school satisfaction and teacher support were sizeable and significantly correlated (r=.61).

Within the family, we measured family poverty after migration, family structure and relationships, and parent and child acculturation. Poverty after migration was defined as living at 100% of the Federal Poverty Level (FPL) in 2006 based on family income and family size reported by the parent. To account for the structure of family relationships, we identified whether the adolescent lived with two parents (1=yes, 0=no). We also measured familism using the average of responses to a 7-item scale of familism (α_LAMHA = .91; Gil and Vega 1996). Family poverty after migration was highly correlated with parents’ high school completion (r=−.25), living in a two-parent family (r=−.38), and poverty prior to migration (r=.38).

Two different scales, the 10-item Psychological Acculturation Scale (PAS; Tropp et al. 1999) and the 4-item language acculturation subscale of the Short Acculturation Scale for Hispanics (SASH; Marin et al. 1987), measured parent and youth acculturation. The SASH (α_youth = .77, α_parent = .77) concentrates more on language usage while the PAS (α_youth = .91, α_parent = .88) focuses on social contacts and cultural understanding. The correlation between these two scales (r_youth = .44, r_parent = .52) suggests they measure related though not identical aspects of acculturation. For both SASH and PAS, respondents answered questions using a five-point Likert scale. We then averaged responses to create a score from one to five with a higher score indicating more acculturation. Due to the highly skewed distribution of the SASH, we created two binary variables indicating whether the adolescent utilized primarily Spanish (1=yes, 0=no) and whether the parent utilized primarily Spanish (1=yes, 0=no). A third aspect of acculturation, years lived in the U.S., was measured with a variable for parents and children indicating fewer than 5 years lived in the U.S. (1 = yes, 0 = no). Adolescent’s years in the U.S. correlated significantly with both the SASH (r=.59) and the PAS (r=.26). Years in the U.S. and acculturation measures were not significantly correlated for parents.

Demographic Controls

We controlled for age and gender in our analyses.

Analytic Plan

Though our primary focus was on adolescent expectations, we began our analysis with an examination of the distributions of and inter-relationships between adolescent expectations, adolescent aspirations, and parent expectations (Table 1). Next, we examined differences in the distributions of these three variables and all covariates by legal status at entry into the U.S. (Table 2). This later analysis allowed us to identify potential confounding variables that could explain the observed association between legal status at entry and adolescent expectations. We then report the means for all covariates (Table 3, column 1) and the unadjusted associations between each covariate and our key dependent variable (Table 3, column 2), adolescent expectations. Unadjusted associations were estimated using ordered logits of the dependent variable with only one covariate in the model. After identifying those covariates with unadjusted associations significant at the .05 level and testing that the proportional odds assumption was not violated, we estimated ordered logits of the association between unauthorized entry into the U.S. and adolescent expectations adjusted for nine key covariates (Table 4). For comparison, this same model is reported for adolescent aspirations and parent expectations. Given the sample size, we aimed to develop parsimonious models and avoid the inclusion of highly correlated variables. Because parents’ high school completion is correlated with poverty both prior to (r = −.71) and after migration (r = −.53), and is also strongly associated with parents’ PAS score (r = .36), we included only parents’ high school completion as a covariate in our final adjusted models. To account for the sample design, we weighted all analyses and adjusted standard errors for the rural–urban stratification and school-level clustering of the data.

RESULTS

Latino immigrant youth in this sample have high educational aspirations and expectations (Table 1). Two-thirds (66%) aspired to finish at least a 4-year college degree and over half (51%) expected to complete a 4-year college or graduate degree. Only their parents’ expectations for them exceeded their own aspirations. Seventy-seven percent of parents expect their children to complete a 4-year college or graduate degree.

Table 1.

Mean Differences in the Educational Aspirations and Expectations Foreign-born Latino Youth, by Legal Status at US Entry (N=275)

	Total	Unauthorized	Authorized
	%/mean (se)	%/mean (se)	%/mean (se)
Educational Expectations
Complete high school or less (1)	39% (5.16)	46% (4.81)	21% (6.95) ***
Complete vocational school (2)	9% (3.31)	11% (4.20)	5% (3.11)
Complete bachelor’s degree (3)	19% (3.55)	18% (2.80)	23% (8.39)
Complete graduate degree (4)	32% (2.81)	25% (3.01)	52% (6.66) **
Rank (1–4)	2.44 (0.08)	2.23 (0.09)	3.05 (0.17) ***
Educational Aspirations
Complete high school or less (1)	30% (5.54)	33% (5.56)	20% (7.00) *
Complete vocational school (2)	4% (1.54)	5% (1.96)	3% (2.11)
Complete bachelor’s degree (3)	29% (3.02)	33% (3.69)	18% (7.05)
Complete graduate degree (4)	37% (4.45)	29% (5.37)	58% (6.91) **
Rank (1–4)	2.73 (0.14)	2.58 (0.15)	3.14 (0.18) **
Parental Expectations
Complete high school or less (1)	20% (2.82)	22% (3.65)	14% (7.71) ***
Complete vocational school (2)	4% (1.47)	4% (1.63)	3% (2.04)
Complete bachelor’s degree (3)	49% (2.77)	52% (2.78)	40% (6.03) †
Complete graduate degree (4)	28% (3.71)	22% (3.79)	44% (7.31) *
Rank (1–4)	2.85 (0.09)	2.75 (0.11)	3.21 (0.20)
Sample Size	275	213	62

^†p<.10,

^*p<.05,

^**p<.01,

^***p<.001

Note: The data are weighted and standard errors are adjusted for the sample design. Unweighted Ns indicate the actual sample size for each category.

For many adolescents, the realization that they have entered the U.S. without authorization can dampen their spirits and lower their educational aspirations and expectations. Because it requires adolescents to consider what is realistically possible, the effects of legal status at entry can be most acutely observed with respect to educational expectations. The majority of youth (57%) with unauthorized entry did not expect to complete college or graduate school, whereas the majority of youth with authorized entry (75%) did expect to complete college or graduate school (Table 1). By contrast, parents’ educational expectations for their children did not differ as significantly by their child’s authorization status (Table 1). They either did not realize the difficulty of overcoming barriers to college education for children with an unauthorized entry or remained optimistic about their children’s opportunities for educational advancement regardless of their children’s authorization status.

Differences in the educational expectations of youth by authorization status can potentially be explained by observable differences in their pre-migration, migration, and post-migration experiences (Table 2). Adolescents who entered the U.S. without authorization experienced significantly different pre-migration, migration, and post-migration experiences from those with authorization. Unauthorized entrants were more likely to have emigrated from Mexico, to have lived in high poverty, to have lived separately from their fathers, and to have moved to the U.S. prior to age 6. Authorized entrants were more likely to have high-school educated parents, had more frequently traveled to the U.S. prior to migration and had moved for educational opportunities in the U.S. Focusing on post-migration factors, we found no significant differences in the community experiences of unauthorized and authorized entrants. However, in comparison to authorized entrants, unauthorized entrants were more likely to have low levels of social support, to live in poverty in the U.S., to have low levels of acculturation, to primarily communicate in Spanish, and to have lived in the U.S. for fewer than 5 years.

Table 2.

Mean Differences in Covariates, by Legal Status at Entry (N=275)

	Unauthorized	Authorized
	%/mean (se)	%/mean (se)
Pre-migration Experiences
Immigrated from Mexico	84% (4.10)	41% (5.67) ***
Lived in extreme poverty	22% (0.03)	2% (2.25) ***
Parent(s) completed high school	25% (4.06)	65% (2.46) ***
Separated from parent 1+ years	65% (2.92)	34% (13.41) *
Separated from mom 1+ years	34% (2.07)	18% (10.02)
Separated from dad 1+ years	58% (3.73)	28% (10.95) **
Moved to US for Work	43% (4.22)	46% (9.02)
Moved to the U.S for Family	28% (4.49)	12% (4.59)
Moved to US for Education	5% (2.09)	21% (6.45) *
Previous travel to the US	9% (1.91)	33% (7.94) **
Migration Experiences
Less than age 6 at migration	19% (3.00)	9% (4.38) **
Less than age 13 at migration	83% (2.75)	75% (1.05)
Experienced any trauma	32% (3.58)	23% (5.00)
Post-Migration Community Experiences
Urban Community	77% (6.88)	87% (5.15) †
Neighborhood unsafe	29% (6.46)	28% (11.39)
Discrimination Index	2.78 (0.09)	2.58 (0.18)
Experienced discrimination	37% (5.01)	42% (5.19)
Post-Migration School Experiences
Social support score	5.5 (0.16)	6.0 (0.17) **
School Satisfation	5.8 (0.07)	5.9 (0.08)
School unsafe	26% (6.20)	40% (10.67)
Teacher Support	9.6 (0.14)	10.0 (0.38)
Post-Migration Family Experiences
Lives with two parents	69% (7.24)	86% (6.88)
Lives in poverty	66% (4.19)	40% (7.81) **
Familism	4.35 (0.08)	4.30 (0.12)
Parent’s Acculturation Score (PAS)	1.66 (0.06)	1.96 (0.14) *
Adol. Acculturation Score (PAS)	1.94 (0.10)	2.43 (0.08) **
Parent Uses Primarily Spanish	61% (2.92)	38% (7.75) **
Adol. Uses Primarily Spanish	32% (6.36)	22% (5.99) †
Parent lived in US < 5 yrs	14% (2.26)	59% (12.67) **
Adol. lived in US < 5yrs	52% (10.39)	69% (10.05) ***
Demographic Controls
Age	14.47 (0.44)	14.62 (0.70)
Female	53% (5.54)	62% (6.52)
Sample Size	213	62

^†p<.10,

^*p<.05,

^**p<.01,

^***p<.001

Note: The data are weighted and standard errors are adjusted for the sample design.

Unadjusted odds ratios calculated from ordered logit models (Table 3) show how strongly unauthorized entry shapes educational expectations. The odds of expecting to complete college versus high school or vocational school were 0.3 times lower for unauthorized entrants compared to authorized entrants. Stated differently, the odds of expecting to complete college versus high school or vocational school were 3.3 times higher (i.e. 1/.03) for authorized entrants compared to unauthorized entrants.

Several covariates associated with unauthorized entry also shaped educational expectations (Table 3). Educational expectations were significantly lower for youth who had moved from Mexico and lived in poverty prior to migration. Those whose parents had completed high school and who moved to the U.S. primarily for education had higher educational expectations. Post-migration experiences positively associated with educational expectations included social support, teacher support, and acculturation. Post-migration experiences negatively associated with educational expectations included school-level disorder (i.e. lack of school safety), and Spanish language use by either the parent or adolescent. Living in an urban community with more co-ethnic resources and experiencing discrimination were not significantly associated with educational expectations in our sample. Regardless of their legal status, immigrant youth live in similar neighborhoods throughout North Carolina where they have little access to co-ethnic community supports and where they encounter similar experiences of discrimination. Thus, there is less variation in these factors in our sample than in larger samples of multiple states and samples including both first, second, and third generation children of immigrants.

Table 3.

Sample Means and Unadjusted Ordered Logits for Educational Expectations (N=275

	Total	Unadusted
	%/mean (se)	OR [95% CI]
Legal Status at Entry
Unauthorized	74% (2.98)	0.30 (0.17–0.55) ***
Pre-migration Family Experiences
Immigrated from Mexico	73% (3.73)	0.35 (0.21–0.57) ***
Lived in extreme poverty	17% (2.69)	0.41 (0.22–0.78) **
Parent(s) completed high school	35% (4.43)	3.81 (1.81–7.99) ***
Separated from parent 1+ years	57% (4.17)	0.90 (0.65–1.26)
Separated from mom 1+ years	30% (2.46)	1.63 (0.64–4.15) †
Separated from dad 1+ years	51% (4.81)	0.66 (0.43–1.02)
Moved to US for Work	44% (4.22)	1.20 (0.68–2.14)
Moved to the U.S for Family	24% (4.49)	0.63 (0.34–1.15)
Moved to US for Education	9% (1.32)	3.96 (1.14–13.71) *
Previous travel to the US	15% (2.72)	1.70 (0.94–3.09) †
Migration Experiences
Less than age 6 at migration	16% (3.23)	0.97 (0.35–2.68)
Less than age 13 at migration	81% (4.36)	0.63 (0.37–1.07) †
Experienced any trauma	30% (2.37)	0.72 (0.45–1.16)
Post-Migration Community Experiences
Urban Community	79% (6.14)	1.53 (0.84–2.78)
Neighborhood unsafe	29% (7.26)	1.01 (0.52–1.98)
Discrimination Index	2.73 (0.11)	0.57 (0.30–1.06) †
Experienced discrimination	38% (4.50)	1.55 (0.69–3.50)
Post-Migration School Experiences
Social support score	5.6 (0.16)	1.31 (1.07–1.62) *
School Satisfation	5.8 (0.05)	1.35 (0.98–1.84) †
School unsafe	29% (3.48)	0.54 (0.35–0.81) **
Teacher Support	9.7 (0.16)	1.26 (1.03–1.55) *
Post-Migration Family Experiences
Lives with two parents	73% (4.33)	1.28 (0.88–1.88)
Lives in poverty	59% (4.59)	0.90 (0.54–1.51)
Familism	4.34 (0.07)	1.39 (0.99–1.95) †
Parent’s Acculturation Score (PAS)	1.74 (0.07)	3.57 (1.16–11.04) *
Adol. Acculturation Score (PAS)	2.07 (0.08)	1.98 (0.99–3.96) †
Parent Uses Primarily Spanish	55% (3.67)	0.17 (0.03–1.07) †
Adol. Uses Primarily Spanish	29% (5.70)	0.46 (0.30–0.70) ***
Parent lived in US < 5 yrs	25% (3.95)	1.15 (0.61–2.16)
Adol. lived in US < 5yrs	57% (10.10)	1.30 (0.83–2.04)
Demographic Controls
Age	14.51 (0.50)	0.97 (0.84–1.11)
Female	56% (3.99)	3.40 (1.39–8.29) **

^†p<.10,

^*p<.05,

^**p<.01,

^***p<.001

Note: The data are weighted and standard errors are adjusted for the sample design.

After adjusting for significant bivariate associations, adolescent expectations still differed significantly by whether the adolescent entered the U.S. with authorization (Table 4). Adjusted for key confounders, the odds of completing college versus completing high school or vocational school were 1.69 times higher (i.e. 1/.59) for authorized entrants compared to unauthorized entrants. These models also showed that, after adjusting for other covariates, school safety problems and adolescents’ Spanish language usage continued to predict lower educational expectations while parents’ high school completion predicted higher educational expectations. After adjusting for school safety problems, adolescent social support and teacher support were no longer associated with educational expectations.

Table 4.

Final Ordered Logits for Educational Expectations and Aspirations

	Adol. Expectations	Adol. Aspirations	Parent Expectations
	OR [95% CI]	OR [95% CI]	OR [95% CI]
Legal Status at Entry
Unauthorized	0.59 (0.35–0.99) *	0.54 (0.27–1.05)	0.62 (0.29–1.33)
Pre-Migration Family Experience
Immigrated from Mexico	0.57 (0.30–1.08) †	1.31 (0.69–2.49)	0.61 (0.34–1.09) †
Parent(s) completed HS	2.23 (1.13–4.39) *	2.67 (1.11–6.41) *	1.27 (0.75–2.18)
Moved to US for Education	2.21 (0.61–8.02)	2.15 (0.59–7.78)	1.57 (0.33–7.33)
Post-Migration School Experiences
Social support score	1.12 (0.92–1.37)	1.25 (0.95–1.66)	1.04 (0.91–1.18)
School unsafe	0.56 (0.38–0.83) **	0.56 (0.35–0.89) *	0.54 (0.16–1.75)
Teacher Support	1.13 (0.96–1.32)	1.02 (0.85–1.21)	0.93 (0.82–1.05)
Post-Migration Family Experiences
Adol. Uses Primarily Spanish	0.40 (0.23–0.70) **	0.49 (0.27–0.89) *	0.39 (0.15–1.05) †
Demographic Controls
Age	0.94 (0.80–1.10)	0.95 (0.80–1.13)	0.94 (0.71–1.25)
Female	2.51 (1.19–5.29) *	1.72 (1.07–2.77) *	1.92 (1.02–3.65) *
F	4.09	10.00	5.92
Sample Size	271	271	257

^†p<.10,

^*p<.05,

^**p<.01,

^***p<.001

Note: Data are weighted and standard errors are adjusted for the sample design. Cutpoints for ordered logits are not reported. The F-statistic shows all models are significant at p<.01. The Variance Inflation Factor (VIF) has a lower bound of one and quantifies the severity of multicollinearity in a least squares reqression analysis. Models in Table 3 were re-estimated using OLS. The average VIF for all models is 1.12. The highest VIF is 1.32 on authorized status in the parents’ expectations model. Thus, multicollinearity is not a concern for these models.

DISCUSSION

Segmented assimilation theory argues that the educational expectations of first and second generation children of immigrants depend, in large part, on three aspects of reception by the host society –government inclusion/exclusion, social acceptance/rejection, and co-ethnic community (Portes and Rumbaut 2001; Portes and Rivas 2011). Our analysis supports segmented assimilation theory by highlighting the critical role that governmental reception plays in shaping the incorporation of adolescent immigrants into U.S. society. It also confirms evidence from recent ethnographic studies describing the incorporation patterns of Latino undocumented youth and adults in California and Arizona (Abrego 2006; Abrego and Gonzales 2010; Gonzales 2011; Menjivar 2008).

A potentially life-long stressor, unauthorized status in the U.S. affects all aspects of adolescents’ lives, including their schooling experiences. Only 43% of immigrant Latino children who lacked legal status in the U.S. expected to complete college or graduate school, whereas 75% of immigrant Latino children with legal status in the U.S. expected to complete college or graduate school. Differences in their expectations by legal status could not be accounted for by their pre-migration country and class origins, migration experiences, or post-migration community, school and family experiences.

We did not find strong evidence of an association between social acceptance/rejection or co-ethnic community and educational expectations for immigrant Latino youth in our sample. Adolescent immigrants with unauthorized status migrate to the U.S. under significantly different circumstances than adolescents with authorization to enter the U.S. legally; but, as has been found in ethnographic studies, they share similar experiences after migration (Abrego 2006). For the most part, they lived in the same communities, attended the same schools, experienced similar levels of discrimination, and had similar perceptions regarding the pervasiveness of discrimination in the U.S. Authorized immigrants were slightly more likely to live in urban areas with established co-ethnic communities and community-based organizations to assist with their incorporation. But, their urban residence was not associated with higher educational expectations. After accounting for unauthorized status, the only other aspect of immigrants’ post-migration reception that significantly affected educational expectations was school safety. Youth who felt unsafe at school had lower educational expectations.

Segmented assimilation theory also argues that family circumstances both before and after migration will influence the educational expectations of first and second generation children of immigrants (Portes and Rumbuat 2001). Children’s class origins or SES background as reflected by living in poverty prior to migration, whether their parents completed high school, and living in poverty after migration correlated highly with one another. Among these measures, we found the strongest association between parents’ high school completion and adolescents’ educational expectations. Immigrant parents with less than a high school education have difficulty helping their children navigate the U.S. educational system and providing academic support to their children (Perreira et al. 2006; Abrego 2006). Additionally, we found that adolescent’s Spanish language use (but not parent’s) was associated with lower educational expectations. Without fluency in English, these adolescents can face social isolation in their schools and additional barriers to their educational advancement (Ko and Perreira 2010; Portes and Rumbaut 2001).

While this study cannot identify the role that public policies play in de-motivating immigrant youth, it provides evidence that immigrant youth who move to the U.S. without authorization have lower expectations for their educational attainment than immigrant youth who move to the U.S. with authorization. In the U.S., two policies have been proposed and, in one case, implemented which have the potential to raise the educational expectations and achievement of unauthorized immigrant youth: (1) the Development, Relief, and Education of Alien Minors (DREAM) Act, and (2) Deferred Action for Childhood Arrivals (DACA).

The federal DREAM Act is a bill designed to allow states to provide in-state resident tuition (IRT) and a pathway to citizenship for immigrant children who have been brought to the U.S. without authorization (Abrego and Gonzales 2010; Flores and Chapas 2009). Though the act has been introduced to Congress several times since 2001, it has not passed. Nevertheless, thirteen states allow unauthorized students, who meet residency and high school graduation requirements in their states, to attend a state university at in-state resident tuition rates (Potochnick 2014). Using Mexican foreign-born noncitizens as a proxy for undocumented youth, studies evaluating the educational consequences of in-state resident tuition (IRT) policies for unauthorized youth suggest that these policies raise their educational expectations and improve their educational attainment (Kaushal 2008; Potochnick 2014). Other states, such as North Carolina, have legislation pending that would not only prevent unauthorized students from receiving in-state tuition discounts but would completely ban admission to community colleges or universities for students who “do not have a lawful immigration status under federal law” (NILC 2013). These bans would extend to students who gain temporary lawful presence through the DACA program.

Implemented in 2012, the Deferred Action for Childhood Arrivals (DACA) program allows unauthorized immigrants age 30 or younger to gain temporary lawful presence and legally work in the U.S. Approximately 2.1 million unauthorized immigrants potentially qualify for DACA (McHugh 2014) and, as of February 2014, 610,194 have been approved (USCIS 2014). To qualify, youth must meet several requirements (for details see DHS 2012). Most important to promoting educational attainment, youth must be currently enrolled in school, have graduated from high school, or have obtained a general education development (GED) certificate. Because youth who gain temporary lawful presence under DACA may also qualify to legally work in the U.S., their job prospects and the economic benefits from high school completion can also increase. Thus, the DACA program could potentially increase the educational expectations of unauthorized youth living in the U.S. (McHugh, 2014).

The capacity of our study to distinguish between students with and without authorization status, to describe differences in their migration and settlement experiences, and to show how these differences together with their legal status at entry influence their educational expectations is a unique strength. Measures of immigrant legal status are rarely available in national datasets, and cannot be imputed from children’s foreign-born status or country of origin (Van Hook et al. 2015). In addition, national datasets do not contained detailed data about immigrant youth’s migration and settlement experiences, especially among those settling in the U.S. Southeast.

At the same time, we must acknowledge our study’s limitations. Most importantly, our study is limited to a cross-sectional sample of one state. Therefore, we cannot generalize to the experiences of Latino immigrant students in other states nor can we evaluate the effects of differences in education or other policies among states that may influence the school experiences of Latino immigrant students. Secondly, our study provides conservative estimates of the effects of unauthorized entry on educational expectations. The majority (71%) of youth in our study were ages 12–15 and enrolled in 7^th–10^th grades. Older, unauthorized students who may better understand how their legal status limits their educational opportunities may have already dropped out and may not have been included in our school-based sample. The effects of unauthorized status may also vary by age and gender, but we were unable to detect any significant interactions. Future studies of unauthorized immigrant youth should consider differential effects by age and gender. Third, our measure of familism may not fully capture its multi-dimensional nature (Calzada et al. 2013) and our measures of discrimination and social support have reliabilities just under .60, the minimum threshold recommended by Nunally and Bernstein (1994). As a result, significant relationships between these measures and educational aspirations or expectations may be more difficult to detect. Finally, our small sample size reduces the power of our estimations. Thus, we reported both unadjusted ordered logits showing the relationship between each possible covariate and educational expectations and adjusted ordered logits for the small subset of 10 covariates associated with adolescents’ expectations, adolescents’ aspirations, and parents’ expectations. We also estimated results using binary and exact logistic models (not shown). Though the coefficient estimates differed somewhat between models, all estimations resulted in similar conclusions. Despite the small sample, this analysis clearly demonstrates that unauthorized status, parents’ education, and school safety are among the most important factors associated with the educational expectations of Latino immigrant youth.

In the current political environment, many researchers understandably avoid collecting data on individuals’ authorization status. Yet without these data, we overlook a critical component of the opportunity structure which places over half of immigrant youth at risk of dropping out, falling into the shadows, and becoming part of a permanent underclass in the U.S. Policies and programs, like DACA and IRT, that help to bring unauthorized immigrant youth out of the shadows can potentially foster their higher educational expectations and their completion of high school. Moreover, as shown by Robles (2009), the greater educational attainment of these youth will not only benefit these youths and their families but also the U.S.economy.