Abstract
How do unauthorized immigrant parents promote family functioning to navigate challenging conditions and contexts in the United States? This article offers the first quantitative analyses contrasting the family organization and maternal knowledge of Mexican and Central-American immigrant mothers by legal status. Using Los Angeles Family and Neighborhood Survey data with a sample of mothers of school-aged children, the analyses investigate whether mothers’ documentation status, origin country/region, and access to social and instrumental support are associated with the frequency of family dinners, the consistency of family routines, and the knowledge of their child’s associations and friendships. Relative to their US-born and documented Mexican immigrant counterparts, undocumented Mexican immigrant mothers have as many or more frequent family dinners, more predictable family routines, and the same level of knowledge about whom their child is with when not at home. Whom mothers can rely on for emergency childcare and financial support also is linked with both family organization and levels of maternal knowledge about their child. More quantitative research is needed about how undocumented immigrant parents actively employ different family functioning strategies to promote strengths and resiliency in their lives in the midst of challenging contexts driven by lack of legal status.
Introduction
In recent decades, extensive public attention and immigration enforcement activities have focused on immigrants who are unlawfully present in the United States (e.g., Meissner et al. 2013). About one-quarter of all immigrants are estimated to reside in the United States without authorization, mostly from Mexico and Central America (Passel and Cohn 2017). 1 In 2016, households with undocumented immigrant parents included nearly six million children, most of whom were born in the United States, and more than a million US-born adult children (Passel and Cohn 2018). Undocumented or liminally legal immigrants encounter hostile contexts of reception that significantly limit their mobility and that cause long-lasting individual and institutional isolation, marginalization, and harm (Menjívar 2006; Hall, Greenman, and Farkas 2010; Bean et al. 2011; Yoshikawa 2011; Menjívar and Abrego 2012; Gonzales, Suárez-Orozco, and Dedios-Sanguineti 2013; Salas, Ayón, and Gurrola 2013; McConnell 2015). Consequently, there is scholarly consensus that lacking legal documentation in the United States is detrimental to the well-being of undocumented immigrants and the mixed-status families to which many of them belong (Brabeck and Xu 2010; Dreby 2015; Yoshikawa, Suárez-Orozco, and Gonzales 2017; Cardoso et al. 2018).
Given these realities, most US-based scholarship about undocumented immigrants concentrates on identifying the seemingly insurmountable barriers of lacking legal status and their deleterious spillover effects for children, families, and communities (e.g., Yoshikawa 2011; Dreby 2015; McConnell 2015; Brabeck and Sibley 2016; Landale et al. 2016). However, a growing body of research also documents the substantial strengths that undocumented immigrant parents, their children, and their families exhibit in spite of and in response to economic, political, and social vulnerabilities (e.g., Parra-Cardona et al. 2006; Raffaelli and Wiley 2013; Brabeck, Sibley, and Lykes 2016). Among other findings, this work identifies ways that undocumented Mexican immigrant parents living in pervasively hostile environments strive to protect their children by maintaining positive parent-child relationships and supporting family functioning (Ayón and Naddy 2013; Philbin and Ayón 2016). While these mostly qualitative studies have pioneered this growing body of research and provided rich descriptions and insights about undocumented immigrant parents and their families, complementary quantitative analyses are useful for examining whether and to what extent the experiences documented in these studies are generalizable.
This article quantitatively examines the family organization and knowledge of Mexican and Central-American-origin mothers by mothers’ nativity, origin country/region, and documentation status. Using Los Angeles Family and Neighborhood Survey data (L.A. FANS), we investigate the family routines and knowledge about their child’s associations and friendships among US-born, documented immigrants and undocumented immigrant mothers of children from elementary through high school (i.e., 6 years to 17 years old). 2 Including both Mexicans and Central Americans allows us to quantitatively explore if and how legal status and origin country/region factors conjointly shape differential experiences of family life within ostensibly similar legal statuses. Reflecting the paucity of quantitative research comparing Mexican and Central-American immigrant parents and families, ours are, to our best knowledge, the first quantitative analyses to contrast the family organization and maternal knowledge of Mexican and Central-American immigrant mothers by origin country/region and legal status, and, for Mexicans, by generation and legal status.
A long line of research has emphasized immigrant family strengths (e.g., Kao and Tienda 1995; Valdes 1996; Crosnoe 2006; Leidy et al. 2010; Jung, Fuller, & Galindo 2012; Ayón and Quiroz Villa 2013; Bermudez and Mancini 2013), complementing recent work demonstrating the assets and resources that undocumented immigrant parents employ to promote family resilience (e.g., Ayón and Naddy 2013; Ayón and Quiroz Villa 2013; Philbin and Ayón 2016). Drawing from this multidisciplinary scholarship and the perspectives of family strengths and family resiliency (McCubbin and McCubbin 1996; Walsh 2003; Defrain and Asay 2007), we conceptualize family routines and maternal knowledge as part of the seemingly invisible assets that families, including those with undocumented immigrant parents, use to maintain individual and family functioning despite experiencing significant vulnerabilities. The results presented here extend the literature on immigrant family settings for children across multiple age-related developmental stages (e.g., Hardway and Fuligni 2006; García Coll and Marks 2009) and on the intersection of legal status and family life at different stages of human development (e.g., Yoshikawa 2011; Brabeck and Sibley 2016; Bradley et al. 2016; Yoshikawa et al. 2017). More broadly, the study also expands understandings of the heterogeneous experiences of Mexican and Central-American immigrants in the United States, and how immigrant families engage in practices that promote family well-being and resiliency (e.g., Yoshikawa 2011; Menjívar and Abrego 2012; Abrego 2014; Philbin and Ayón 2016).
This article also examines whether social support available to Mexican and Central-American immigrant mothers of school-aged children is connected with family organization and maternal knowledge, expanding on research about how social support shapes immigrant mothers’ parenting practices (e.g., Jung et al. 2012). We examine multiple forms of support, including perceptions of financial and nonfinancial help available from immediate family, fictive kin, and more distant relations. Thus, our findings offer insights about the complex “protective processes” and “support structures” in the family setting, promising directions for further unpacking the immigrant paradox, or immigrants’ generally better-than-expected outcomes compared to their US-born counterparts in spite of socioeconomic and other disadvantages (Marks, Ejesi and García Coll 2014, 61).
In the next section, we bring together interdisciplinary scholarship about immigrant families to develop hypotheses about how documentation status, sending or origin country/region, and social support are associated with family routines and maternal knowledge of their child’s associations and friendships. Carrying out these analyses with diverse Latinx families is important, given the known benefits of family organization and maternal knowledge for children and families (e.g., Bámaca et al. 2005; Romero and Ruiz 2007; Cardoso and Thompson 2010; Roche and Ghazarian 2012) and the increasing diversity of Latin American immigrants in the United States in recent decades (O’Connor, Batalova, and Bolter 2019). Extending previous scholarship, our results quantitatively demonstrate the substantial strengths and positive practices that Mexican and Central-American immigrant families with undocumented immigrant mothers exhibit.
Research on Immigrant Wellbeing and Families
Scholars in education, psychology, sociology, human development and family studies, and other fields have investigated the roles of individual and family factors shaping immigrants’ lives in the United States, including—but not limited to—origin country and dominant language used in the household (e.g., Portes and Rumbaut 2001; Hardway and Fuligni 2006; Crosnoe 2012; Fuller and García Coll 2010; Crosnoe 2012). The family context, these works show, is a critical microsystem where children are socialized, accomplish developmental tasks, and acquire social competencies used in other arenas (e.g., Crosnoe 2006; Galindo and Fuller 2010; Jung et al. 2012).
A large literature has documented the immigrant paradox, the relatively advantageous outcomes of immigrants compared to natives (Marks et al. 2014). This concept first emerged from epidemiologists who observed more favorable than expected prenatal, birth, and other characteristics for children of Mexican immigrant and Mexican-American mothers in the southwestern United States (Markides and Coreil 1986). Since then, interdisciplinary scholarship has identified advantages for immigrant adults and their children in multiple domains (Markides and Eschbach 2005). For instance, studies show that children of more recently arrived immigrant parents have similar or even more favorable educational and developmental outcomes than children of US-born parents or more established immigrants, despite experiencing lower socioeconomic status and other overlapping vulnerabilties (Kao and Tienda 1995; Fuligni 2001; Portes and Rumbaut 2001; Crosnoe 2006; 2012; García Coll and Marks 2009; Jung et al. 2012; Guerrero et al. 2013). As Viruell-Fuentes, Miranda, and Abdulrahim (2012) contend, studies delving into the mechanisms of observed advantages in the immigrant paradox should account for the effects of structural factors, such as legal status, on shaping outcomes. Indeed, researchers recommend that next steps in further unpacking the immigrant paradox include simultaneous consideration of risk factors (i.e., poverty and relatively low education), protective factors (i.e., sources of support), and analyses explicitly considering legal status (Marks et al. 2014), aims of the current study.
Scholars characterize much early research on immigrant (and US-born) Mexican and other Latinx parenting as operating from a “deficit” framework that a priori approached these communities as being at risk (e.g., Fuller and García Coll 2010; Díaz, Denner, and Ortiz 2017). At present, there are significant overlapping stressors affecting Mexican and/or Central-American immigrants’ family functioning and stability, including low pay and persistent economic insecurity, employment-related challenges (i.e., a lack of autonomy and long hours), and difficulties accessing services and programs for which they and their children are eligible (Hamilton and Stoltz Chinchilla 2001; Yoshikawa 2011; Raffaelli and Wiley 2013). Reduced access to social networks is another challenge. In leaving extended family members behind in the origin country, many Mexican and Central-American immigrants report reduced levels of available social support (Leidy et al. 2010; Jung et al. 2012). Not having the same socioemotional and tangible support in the United States as they did in their origin country adds to the hardships that these immigrants experience in their parenting and family lives (Perreira, Chapman, and Stein 2006; Leidy et al. 2010; Dreby 2015).
However, a long line of work also has emphasized the strengths and assets that Latinx immigrant and native-born parents possess that promote and strengthen individual and family resilience (Tienda 1980; Fuller and García Coll 2010; Galindo and Fuller 2010; Jung et al. 2012; Suárez-Orozco, Yoshikawa, and Tseng 2015). Resilient people sustain “healthy functioning” despite long-term demands, enabling them to be more successful than might be expected (Schetter and Dolbier 2011, 637). Resilient families are able to successfully cope with transitions over the life cycle, despite stress and hardship (McCubbin and McCubbin 1996; Walsh 2003). As an example of practices supporting resilience, many Mexican families socialize their children with an emphasis on good behavior, respectful interactions with adults, and cooperation (i.e., bien educado, respeto, cariño) (Valdes 1996; Holloway and Fuller 1997). These practices foster Mexican and Latinx children’s strong positive social behaviors, development, and academic success, despite experiencing other disadvantages (Crosnoe 2006; Galindo and Fuller 2010; Jung et al. 2012).
Strong relationships and access to social support are part of the “pivotal processes” (Jung et al. 2012, 1,525) that foster individual- and family-level social and emotional well-being among Mexican and other Latinx immigrant and US-born family members (Bámaca et al. 2005; Benzies and Mychasiuk 2009; Cardoso and Thompson 2010; Leidy et al. 2010; Bermudez and Mancini 2013; Philbin and Ayón 2016). Being able to rely on extended relatives or friends, co-residing or not, can provide much-needed assistance with childcare, advice, and financial help that enables parents to maintain more stable routines and provides more nonparental sources of information about children (Romero and Ruiz 2007). Aside from blood relatives, fictive kin-type relationships also provide access to important social and financial resources (Nelson 2013).
Legality and Family Life
Lacking legal status is linked with well-known risk factors or stressors that immigrants typically experience, such as unstable family income, poverty, and housing problems (Brabeck and Xu 2010; Cardoso and Thompson 2010; Bean et al. 2011; Yoshikawa 2011; McConnell 2015). Additional long-term challenges for undocumented immigrants, others with impermanent statuses, and their often mixed-status families include encountering exploitation in multiple domains, uncertainty about the future, fears of deportation, and the pervasive effects of the laws on their lives (Brabeck and Xu 2010, Yoshikawa 2011; Menjívar and Abrego 2012, Perreira and Ornelas 2013; Dreby 2015; Cardoso et al. 2018). These issues can have insidious effects on family lives. For example, in increasingly hostile climates, undocumented Mexican and other Latinx immigrant parents experience significant discrimination that compromises their parenting, including lowering their knowledge about their children’s whereabouts and friendships (Ayón and García 2019). Undocumented immigrant parents trying to decrease their risk of deportation may limit their activities outside the home and experience feelings of disempowerment when it comes to exerting parental authority, further altering family dynamics (Cardoso et al. 2018). Undocumented immigrants also often have relatively fragile support systems (Yoshikawa 2011; Dreby 2015) and report less social support than their documented immigrant counterparts, such as less help with childcare and financial assistance (Brabeck et al. 2016). This limited social support might reduce their ability to maintain consistent family routines and other effective parenting strategies (e.g., Perreira et al. 2006; Leidy et al. 2010).
Despite these and other challenges, undocumented immigrant parents and their families display considerable individual- and family-level strengths that help them cope with overlapping layers of vulnerability (Parra-Cardona et al. 2006; Raffaelli and Wiley 2013; Brabeck et al. 2016; Philbin and Ayón 2016). For instance, undocumented Mexican immigrant mothers intentionally try to protect and support their children by building family unity, improving their communication and emotional connections with children, and drawing on their networks to foster individual and family resiliency (Ayón and Naddy 2013; Ayón and Quiroz Villa 2013; Philbin and Ayón 2016). Undocumented Mexican and Central-American immigrants are more likely to live with extended relatives (i.e., nieces/nephews) and nonrelatives than their documented or native-born co-ethnic peers (Hall, Musick, and Yi 2019). Thus, undocumented immigrants in these living arrangements may have access to people who, when willing and able to provide emotional and tangible assistance, could be beneficial for family functioning.
Descriptive results from quantitative analyses offer additional indications that undocumented immigrant parenting may not be as compromised as might be expected. Undocumented immigrant parents report similar levels of parent-child involvement, communication, parenting confidence and attachment (Brabeck and Sibley 2016), family routines (Noah and Landale 2018), parenting/family stress and strain (Brabeck et al. 2016; Noah and Landale 2018), and maternal depression (Landale et al. 2016) as their documented immigrant peers. Given the importance of these domains for advantageous child and family outcomes (Bámaca et al. 2005; Benzies and Mychasiuk 2009; Roche and Ghazarian 2012; Marsiglia et al. 2014), delving further into the family lives of immigrant mothers of diverse documentation statuses is valuable.
Origin Country/Region and Legal Status
Mexican and Central-American immigrants in the United States share some similar experiences, such as enduring long family separations from family members left behind, discrimination from others, and low-paid and exploitative working conditions (Perreira and Ornelas 2013; Raffaelli and Wiley 2013; Abrego 2014; Dreby 2015). They also differ in their experiences. Relative to Mexican migration, Central-American migration is more recent, beginning in earnest in the 1980s (Hamilton and Stoltz Chinchilla 2001). Other important variations between Mexican, Salvadoran, Guatemalan, and Honduran immigrants include diverse national-origin contexts, ethnic backgrounds, and different push factors out of their origin countries related to violence, political persecution, poverty, and disasters (Hamilton and Stoltz Chinchilla 2001; Menjívar 2006; Menjívar and Abrego 2012; Abrego 2014; Flores-Yeffal and Pren 2018).
Immigrants’ documentation status intersects with origin country/region, context of exit, and context of reception to similarly and/or differently shape immigrants’ experiences (Abrego 2014). For instance, US immigration and citizenship laws produce diverse legal-status categories of migrants and, consequently, varying contexts of reception that profoundly impact immigrants’ lives (Menjívar 2006; Abrego 2014). Some Salvadorans and Hondurans are eligible for Temporary Protected Status (TPS), a liminal category that allows for temporary legal residence and employment in the United States but not a path to US citizenship (Menjívar and Abrego 2012); others have been accepted as political refugees and can become US citizens (Hamilton and Stotlz Chinchilla 2001). 3 Mexicans are not eligible for TPS, but many were able to eventually become Lawful Permanent Residents and/or become naturalized citizens via the 1986 Immigration Reform and Control Act (Bean et al. 2011), as were some Guatemalans and Hondurans via the 1997 Nicaraguan Adjust and Central American Relief Act (Menjívar and Abrego 2012). Although immigrants in these different categories might be considered legal immigrants, they have heterogeneous options and experiences based on their specific documentation.
Thus, considering these factors, Mexican and Central-American immigrants should not be assumed to have identical family lives and parenting practices in the United States, even for those sharing similar umbrella legal statuses (documented or undocumented). Survey-based studies concentrating on issues of legal status among these groups, however, tend to aggregate Mexican and Central-American immigrants together, typically because of small sample sizes of Central Americans (e.g., McConnell 2015; Hall et al. 2019) or rely on analyses with a small number of Central-American respondents (e.g., Flores-Yeffal and Pren 2018). Nevertheless, when feasible, including Central-American immigrants allows for quantitative explorations about the similarities and dissimilarities between Mexican and Central-American immigrants’ vis-à-vis legal status in the United States.
Family Routines and Maternal Knowledge
The outcomes explored in this article, family routines and maternal knowledge about their child’s associations and friendships, are suggestive about family functioning and resilience more broadly. Consistent family routines, high levels of parent involvement, and close supervision of children promote individual- and family-level well-being and resilience for immigrants and nonimmigrants (Romero and Ruiz 2007; Benzies and Mychasiuk 2009; Cardoso and Thompson 2010; Leidy et al. 2010; Schetter and Dolbier 2011; Ayón and Quiroz Villa 2013; Perreira and Ornelas 2013; Raffaelli and Wiley 2013). For example, family routines, such as having regular times for meals and bedtime and eating dinners together, provide organization, structure, and predictability for children (Spagnola and Fiese 2007). Eating dinner together as a family is especially valuable as it involves sharing family stories and cultural values, thereby promoting children’s linguistic development and deepening emotional connections among family members (Spagnola and Fiese 2007). Such home-based social activities also encourage children’s acquisition of social competencies that can promote school success (Galindo and Fuller 2010).
Youth from Mexican immigrant backgrounds with higher levels of family routines, research has shown, are less likely to engage in risky or delinquent behaviors and have higher academic achievement than peers without these routines (Schetter and Dolbier 2011; Roche and Ghazarian 2012). Similarly, parents’ knowledge about their children promotes beneficial youth outcomes and family resilience in Mexican and other Latinx immigrant families (Cardoso and Thompson 2010). For instance, higher levels of parental knowledge about children’s whereabouts and peer associations are linked with higher self-esteem and academic performance, lower depressive symptoms, and less delinquency and aggression among Latinx and other children (Bámaca et al. 2005; Romero and Ruiz 2007; Spagnola and Fiese 2007; Roche and Ghazarian 2012). Practices that support individual and family well-being conjointly contribute to family functioning and resilience.
Data and Methods
The cross-sectional data examined in this article come from the first wave of the L.A.FANS collected in Los Angeles County from 2000 to 2002 (Peterson et al. 2004; Pebley and Sastry 2011). Although these data are now 20 years old, they have several unique strengths that make them ideal for testing our specific hypotheses. First, L.A.FANS is one of the few sources with extensive information about family life, children, and parents, and with sufficiently detailed respondent-provided information about their documentation status to categorize Mexican and Central-American immigrants as authorized or unauthorized. 4 Second, Los Angeles County is the most important US county for Mexican and Central-American immigrants (Zong and Batalova 2018; O’Connor et al. 2019), and for undocumented immigrants (Migration Policy Institute, n.d.). Independent analyses of L.A.FANS indicate that these data are representative of undocumented immigrants in Los Angeles (Bachmeier, Van Hook, and Bean 2014). Third, the early 2000s capture most contemporary Mexican and Central-American immigrants who are still living in the country. About 60 percent of Mexican immigrants and about half of Central-American immigrants in 2017 arrived to the United States before 2000 (Zong and Batalova 2018; O’Connor et al. 2019). As a recent National Academies of Sciences, Engineering, and Medicine (2015) review of datasets notes, “given the wealth of data contained in L.A.FANS…it offers a promising source for studying the consequences of unauthorized status in the nation’s largest single undocumented population” ( 420).
Respondents in approximately 40 randomly selected households completed the survey in each of 65 census tracts in Los Angeles (Peterson et al. 2004). People living in poor and very poor census tracts, co-residing in households with children, and identifying as Hispanic/Latino were oversampled. Respondents completed the personal interview in either English or Spanish. One adult from each household was randomly selected to fill out the adult questionnaire, and one child was randomly selected if children were present at the household. A primary caregiver, nearly always the mother, provided information about the selected child (Peterson et al. 2004). The files are merged so that each record includes information about the child respondent, mother, immediate family, and household. Starting with all mothers who were matched to a child (n=3,122), we first restricted our analytic sample to those who self-identified as Mexican/Mexican American (n=1,468) or as Central American (n=339). Then, we restricted our sample to mothers with school-aged children (ages 6 years to 17 years). We excluded children younger than 6 years, as children in early childhood require more flexibility in routines than older children attending school (Spagnola and Fiese 2007; Wildenger et al. 2007), and may have different family routines and maternal knowledge than older children. The resulting sample sizes of US-born Central-American mothers were too small to keep in the analyses and were excluded. The final analytic sample size is 1,133: 914 children with Mexican-origin mothers and 219 children with Central-American-origin mothers.
Hypotheses
We hypothesize that despite the substantial hardships associated with undocumented status in the United States, undocumented Mexican and Central-American mothers possess a variety of strengths that bolster their family organization and knowledge about their children. Therefore, we expect that they report the same or higher levels of family routines and equal maternal knowledge of their child’s associations and friendships as others from the same country/region (Hypothesis 1). Mexican and Central-American immigrants have diverse sociopolitical contexts and push factors in their origin countries, in addition to different options for legal status and the ability to become a citizen once in the United States (i.e., TPS, lawful permanent residence, and naturalized citizenship) (Hamilton and Stoltz Chinchilla 2001; Menjívar 2006; Bean et al. 2011; Abrego 2014). Given these differences, we hypothesize that Mexican and Central-American immigrant mothers sharing the same umbrella legal status (i.e., documented and undocumented) may have dissimilar family routines and maternal knowledge in the United States (Hypothesis 2). Finally, given the importance of social support for overall family socio-emotional functioning, family organization, and maternal knowledge (Romero and Ruiz 2007; Cardoso and Thompson 2010; Jung et al. 2012), we hypothesize that Mexican and Central-American mothers reporting more access to/support from immediate family or extended family/fictive kin are likely to have more family routines and maternal knowledge than those indicating less or more distant social support, regardless of legal status and origin country/region (Hypothesis 3).
Dependent Variables
Mothers provided data for our four dependent variables: (a) family dinners, (b) family routines, (c) maternal knowledge of whom their child is with when not at home (child’s associations), and (d) maternal knowledge of child’s friends and friends’ parents (child’s friendships). Given the importance of family dinners in youth development relative to other family routines (Spagnola and Fiese 2007), we operationalize family dinners as one dependent variable, indicating the days per week the family ate dinner together (ranges 0–7), which taps into the frequency of contact and interaction as they share a meal. In contrast, family routines focuses on the predictability and consistency of routines. This indicator is based on responses to four items about the number of days per week the mother, any present spouse/partner, and children ate breakfast, ate dinner, did chores, and went to bed at about the same time each day. Each item in this measure is indicated by how many days per week, ranging from 0 to 7, and the scale is created by averaging responses to the four items. Maternal knowledge of whom the child is with indicates how often the mother knew whom the child is with when not at home (ordinal variable ranging from 1=only rarely to 4=all the time). Maternal knowledge of child’s friends and parents is the mean score of two items about how many of the child’s friends and child’s friends’ parents that the mother knew (1=none, 4=all). These operationalizations are based on prior studies of family organization and parental knowledge (Kerr and Stattin 2000; Bámaca et al. 2005; Spagnola and Fiese 2007; Padilla-Walker, Harper, and Bean 2011).
Mothers’ Nativity, Legal Status, Origin Country/Region, and Source of Support
Information on mothers’ nativity, legal status, and origin country are used to create five dichotomous categories. US-born Mexicans are US-born mothers who self-identified as Mexican. Mexican-born or Central-American-born respondents who identified as US citizens or legal permanent residents, or who reported having asylum, refugee status, temporary protected status, or a valid visa, were classified as documented Mexican or Central-American immigrants, respectively. Although this approach retains the largest possible sample size of both groups, there are large and meaningful differences between being a US citizen, being a noncitizen but eventually eligible for citizenship, and having only temporary authorizations like TPS that can be rescinded at any time (Menjívar 2006; US Citizenship and Immigration Services 2020). Consistent with previous studies, undocumented Mexican or undocumented Central-American immigrants responded negatively to questions about naturalized citizenship and all other forms of legal residence (e.g., Bean et al. 2011; Brabeck and Sibley 2016; Landale et al. 2016; McConnell et al. 2020).
Other focal independent variables pertain to mothers’ social support, operationalized with multiple indicators (e.g., Cohen and Wills 1985; Jung et al. 2012). A dichotomous variable indicates the frequency of getting together with relatives or friends (1 equals at least bi-weekly contact, 0 equals less frequently). We operationalize availability of support in two ways: whom caregivers might call for emergency childcare in the middle of the night and whom they would call first if they needed to borrow money for their rent or mortgage. Responses include immediate/vertical family (i.e., caregiver’s mother/father, and mother-in-law or father-in-law), horizontal/fictitious family (i.e., caregiver’s sibling or neighbor), others (i.e., babysitter, employer, or co-worker), and no one to call. Immediate/vertical family is the reference category.
Control Variables
Prior research connects child, maternal, and family characteristics with family routines and maternal knowledge, and many are incorporated in the analyses as control variables. For example, child’s age and gender are linked with time spent in family activities, extracurricular activities, family routines, and parental knowledge (Kerr and Stattin 2000; Dubas and Gerras 2002; Roche and Ghazarian 2012). Maternal knowledge might differ with the child’s age due to differences in child disclosure of information or in the amount/kinds of information that parents solicit from children (Padilla-Walker et al. 2011). Child age is coded as a set of three binary categories roughly representing different schooling levels and developmental stages: early to middle childhood (6–9 years old, the reference category), early adolescence (10–15 years old), and middle adolescence (16–17 year olds). Child’s gender is coded as 1 if female.
Turning to maternal characteristics, single parents may have less time, energy, and resources to engage in consistent family routines (Wildenger et al. 2008), and to solicit information from children and parents with co-residing partners (Padilla-Walker et al. 2011). Employed mothers working more hours might share some of these challenges, relative to mothers without outside employment or those working fewer hours. Mothers with lower levels of education and income, and with more children in the household exhibit compromised family routines and/or less parental knowledge (Spagnola and Fiese 2007; Padilla-Walker et al. 2011; Bradley et al. 2016). Binary indicators represent whether the mother was a single parent (1 if single parent, 0 if not single parent), her education (1 if mother did not complete high school, 0 if completed high school), and whether the family was poor (i.e., under the 2000 federal poverty threshold for a family of a given size). A set of four binary categories represent whether a mother worked and weekly hours worked: 0 hours (reference), 1 to 39 hours, 40 hours, and 41 hours or more. A continuous indicator represents the total number of children in the household and mother’s age at the time of the survey, as both might be associated with routines and maternal knowledge. A binary variable for mother’s religion is Christian/Catholic (1=Christian, Protestant, and Catholic; 0= otherwise) and taps into the possibility that Christian-based religious beliefs might be relevant.
Two variables control for immigrant mothers’ integration in the host country. Duration in the United States is associated with changes among children, parents, and families (Portes and Rumbaut 2001), which could extend to family organization and maternal knowledge. Therefore, one is a continuous variable indicating how many years the respondent had spent in the United States. For US-born Mexicans, the number of years spent in the United States equaled zero. A variable for survey language indicates whether the survey was answered in English (coded as 1; 0 otherwise), offering indirect information about language proficiency or preference otherwise unavailable in these data.
Finally, mothers’ emotional and psychological well-being may be associated with family organization and maternal knowledge. For example, mothers who report experiencing difficulties with parenting demands (i.e., parenting strain) also could have less consistent family routines and less knowledge about their child. Operationalized following previous work (Noah and Landale 2018), parenting strain is an averaged scale based on the level of agreement with four statements, such as “Being a parent is harder than I thought it would be” and “I often feel tired, worn out, or exhausted from raising a family.” Respondents could offer five response options to these statements, ranging from “1=completely false” to “5=completely true” (higher values indicate higher parenting strain). This scale’s internal consistency was α=0.66, an alpha value that is adequate (Taber 2018), theoretically driven, and a significant covariate in the analyses.
Mothers who feel more efficacious about their abilities and life (i.e., self-efficacy) (Pearlin et al. 1981) may have more family organization and knowledge about their child’s associations and friendships. L.A.FANS asked seven questions related to the Pearlin self-efficacy scale (Pearlin et al. 1981), such as “no way I can solve problems that I have” and “I can do anything I set my mind to.” Respondents answered these prompts with four response options ranging from “strongly agree” to “strongly disagree.” Constructed following the literature (Pearlin et al. 1981), the questions are combined into an averaged scale with a higher value reflecting the mother’s higher self-mastery (values range from 1 to 4). Internal consistency for this scale was α=0.66, an “adequate” alpha value (Taber 2018). Depressed mothers often have lower engagement, less energy, and fewer positive interactions with others (Lovejoy et al. 2000), which could be connected with family routines and maternal knowledge about their child’s associations and friendships. Following Landale and colleagues (2016), mothers are coded as depressed=1 if their probability of depression was greater than 0.5 on the Composite International Diagnostic Interview Short-Form (CIDI-SF) depression inventory and 0 otherwise. Given that our analysis relies on cross-sectional data, it is not possible to determine causal order between these maternal characteristics, family organization, and maternal knowledge. Thus, the analyses identify whether these variables are associated but do not identify causal relationships.
Analytical Approach
Ordinary Least Squares (OLS) regressions examine the associations between family dinners, family routines, mothers’ knowledge of whom the child is with, mothers’ knowledge of the child’s friends and parents, and the independent variables. Two OLS regressions are estimated for each dependent variable: (a) incorporating only the nativity, origin country, and legal status variables where undocumented Mexican mothers are the reference group, and (b) adding indicators of support for mothers and the full set of control variables. 5 The results section presents the unstandardized OLS coefficients from these regressions. Identical analyses are estimated with a change in the reference group to documented Central-American immigrant mothers, allowing for comparisons with undocumented Central-American immigrants (Hypothesis 1) and documented Mexican immigrants (Hypothesis 2). Other than these nativity, legal status, and origin country/region indicators, results for all other included variables are the same. Therefore, these ancillary analyses are discussed in the results section but not presented in a separate table (results available upon request).
There are no missing values for maternal nativity, one of our focal independent variables. However, to generate the correct parameter estimates and standard errors, multivariate normal regression multiple imputation methods were used to account for other missing data, creating 25 imputed data sets to follow best practices in analyzing data with missing values (Johnson and Young 2011). In addition, appropriate survey data analysis procedures and weights were used to account for the complex survey sample design such as stratification and clustering using “mi estimate: svy” commands in Stata. Lastly, a check of variance inflation factors (VIF) for all multivariate analyses indicates no multicollinearity issues. All analyses were conducted using Stata 14.
Results
Table 1 includes the descriptive statistics and effect sizes of all variables by mothers’ nativity, origin country/region, and legal status. The profiles suggest family dinners between about four to more than six times a week and relatively high levels of predictable family routines (values ranging between 4.6 to 5.2; maximum is 7) among Mexican and Central-American respondents. Mothers knew whom their child was with when not at home between most and all the time (values range 3.7 to 4, maximum of 4), and knew between some and most of their child’s friends and friends’ parents (values between 2.3 and 2.7), consistent with previous work (e.g., Romero and Ruiz 2007). Analyses of statistical significance presented in Table 1 as superscripts reveal that among Mexican mothers, undocumented immigrants had the highest frequency of family dinners, more consistent family routines, and similar maternal knowledge about their child’s associations but the lowest level of knowledge about their child’s friendships. Undocumented Central-American mothers had more family dinners than documented Central-American immigrants but similarly consistent family routines and equal levels of maternal knowledge.
Weighted Descriptives for the Final Sample.
Source: Los Angeles Family and Neighborhood Study Wave 1; U.S. Census 2000.
Note: Multiple imputation (m=25); weighted.
a Significantly different from undocumented Mexican immigrant mothers, p<.05 (one-tailed test).
b Significantly different from documented Central American immigrant mothers, p<.05 (one tailed test).
Reported effect sizes are Cohen’s d values (reference is undocumented Mexican immigrant mothers).
Turning to availability of social support, between 42 percent and 74 percent of the sample reported biweekly contact with relatives or friends. Undocumented Mexican immigrant mothers had less biweekly family contact than their US-born counterparts and were the most likely to ask horizontal/fictitious family for emergency money or childcare rather than immediate family, compared to other Mexican respondents (Table 1). That undocumented Mexican immigrant mothers reported less frequent family contact and relied on more distant connections for help than Mexican Americans aligns with the relatively low levels of support that undocumented immigrant parents have previously reported (Brabeck et al. 2016). Nevertheless, more than 90 percent of all groups other than undocumented Central-American immigrants had someone to ask for emergency childcare at night, and nearly all had someone from whom to borrow money to pay housing costs. US-born Mexican mothers were most likely of all groups to have immediate/vertical family support for both emergency childcare at night and emergency money, while undocumented Central-Americans were the least likely. Although there is variation across other characteristics, sampled mothers tended to be in their mid to late 30s, reside with a partner, have 2-3 children in the household, have relatively low levels of parenting strain (maximum is 5), and were unlikely to be depressed (Table 1). Over 90 percent of mothers in the sample identified as Christian/Catholic, in line with patterns for people in Mexico and Latin America observed in other survey data (Pew Research Center 2013). Immigrant mothers were more likely to have less than a high school education, to take the Spanish rather than the English version of the survey, and less likely to work (and work fewer hours) relative to US-born Mexican mothers. Tests of statistical significance indicate differences between undocumented Mexican and Central-American immigrants in some of the independent variables, such as who would help with childcare and money, Pearlin self-efficacy, parenting strain, and child’s age.
Family Dinners
The first two columns of Table 2 present the OLS analyses regressing the weekly frequency of family members eating dinner together on mothers’ origin country/region, nativity, and legal status indicators, and a second model that incorporates the complete set of independent variables. Compared to undocumented Mexican immigrant mothers, all other Mexican mothers had fewer family dinners (Model 1). Once other variables are controlled, undocumented Mexican immigrant and US-born Mexican mothers reported similar numbers of family dinners per week, while documented Mexican immigrant mothers had fewer weekly family dinners (unstandardized OLS coefficients in Model 2 are 0.17 and -0.82, Table 2). Also consistent with our hypotheses of better-than-expected family organization for undocumented immigrants, ancillary analyses indicate that undocumented Central-American immigrants also had significantly more family dinners per week than their documented Central-American immigrant counterparts, net of all variables. The results further support our second hypothesis about differences between immigrants sharing the same umbrella legal status. Undocumented Central-American immigrants had more family dinners than undocumented Mexican immigrants (OLS coefficient of 0.84, Model 2); ancillary analyses indicate that documented Central-American immigrants also had more family dinners than their documented Mexican counterparts.
Ordinary Least Squares Regression of Family Routines and Maternal Knowledge: Unstandardized Coefficients and Standard Errors.
Source: Los Angeles Family and Neighborhood Study Wave 1; U.S. Census 2000.
Note: Multiple imputation (m=25); weighted. Standard errors in parentheses.
*p < .05 **p < .01 ***p < .001.
As expected, some forms of social support are positively associated with family organization. For example, mothers who had more frequent family contact also had more frequent family dinners than those with less contact (0.56, Model 2). On the other hand, whether it was immediate family or other networks who could help with emergency childcare and financial requests was not associated with the frequency of family dinners. Turning to other variables, child age and mother’s age were related to the frequency of family dinners. The hours the mother worked also matter, with mothers working 41 hours or more having the fewest family dinners per week compared to unemployed mothers (-1.29). Controlling for hours worked and other variables, single mothers actually had more family dinners than those with a partner, as did those identifying as Christian or Catholic. The latter result may reflect a religious motivation or the uniqueness of the small proportion of the sample who identified as other religions/nonreligious (Table 1). Mothers with more years of US residence had more family dinners. Mothers with higher self-efficacy scores reported fewer family dinners, as did those who were depressed or who reported more parenting strain (OLS coefficients of -0.34, -0.30, and -0.15, respectively). The remaining variables were unassociated with the weekly frequency of family dinners.
Family Routines
As expected, undocumented Mexican immigrants had significantly higher levels of consistent family routines than both US-born and documented immigrant Mexican mothers (Model 3, Table 2), including once the full set of covariates were controlled (OLS coefficients of -0.59 and -0.37, Model 4). Ancillary analyses indicate similar results: undocumented Central-American immigrant mothers had more family routines (at the .10 level of significance) than their documented Central-American peers, net of other variables. Contrary to our second hypothesis that Central American and Mexican immigrant mothers sharing the same legal status might have dissimilar levels of family organization, undocumented immigrants from Mexico and Central America had the same levels of these family routines (Model 4), as is true for ancillary contrasts between documented Mexican and Central-American immigrant mothers.
Social support was positively associated with more family routines. For instance, mothers who had more frequent family contact or who would ask an immediate family member for emergency childcare over a horizontal/fictitious family member had more consistent family routines (0.35, -0.10, respectively, Model 4). However, compared to those who would ask immediate family members for emergency money, those who had no one to ask, had more family routines (0.28, Model 4). These mothers represent a tiny proportion of the sample (Table 1). Turning to control variables, mothers with older children had less consistent family routines than those with 6–9 year olds, as did older women, those in poor families, and mothers working 40 hours or more per week relative to their counterparts with other characteristics. Christian/Catholic mothers reported more family routines than non-Christian/Catholics. Mothers who had higher self-efficacy scores or who were depressed reported fewer family routines than those who had lower self-efficacy scores or who were not depressed (-0.41 and -0.16, respectively).
Maternal Knowledge of Whom Child is With
Undocumented Mexican immigrant mothers had similar levels of knowledge about whom their child was with when not at home as documented Mexican immigrants but less knowledge than US-born Mexicans (Model 6, Table 2). Ancillary analyses comparing Central-American immigrants to one another indicate that undocumented immigrants had more knowledge at the .10 level than documented immigrants. Thus, the results provide partial support of our first hypothesis. Moreover, the results align with the second hypothesis regarding expected variation between those sharing the same umbrella documentation status. Undocumented Central-American immigrants had more knowledge of their child’s associations than did their undocumented Mexican counterparts (OLS coefficient of 0.17, Model 6), with the same pattern also appearing in ancillary analyses comparing documented Central American and Mexican immigrant mothers (at the .10 level of significance).
Social support had a mixed relationship with maternal knowledge of whom their child was with when not at home. Having more frequent contact with relatives or friends was not linked with more knowledge; however, access to more immediate networks who could provide financial assistance was connected with more knowledge. For example, compared to mothers who would first ask immediate relatives for emergency money, those who would ask horizontal/fictitious family or who had no one to ask for money had less maternal knowledge of whom their child was with (-0.20, -0.09, respectively, Model 6). In contrast, mothers who would ask horizontal/fictitious family for emergency childcare had more information about whom their child was with than those who would ask immediate/vertical relatives (0.14). These mothers may have had a larger extended network available to help with children and indirect, additional sources of information about their child's social lives (e.g., Romero and Ruiz 2007). In terms of other covariates, compared to mothers of 6–9 year-old children and mothers of girls, those with older children or with sons had less knowledge. Older women, single parents, working mothers, and those with more children in the household had less knowledge, while Christian/Catholic mothers, those with more years in the United States, those having a higher Pearlin Self Efficacy scale, those who were not depressed, and those reporting less parenting strain reported more knowledge about whom their child was with compared to mothers with different characteristics.
Maternal Knowledge of Child’s Friends and Friends’ Parents
In the baseline analyses, undocumented Mexican immigrant mothers reported knowing fewer of their child’s friends and friends’ parents than other Mexican mothers (Model 7, Table 2). However, controlling for the full set of characteristics, US-born Mexicans and undocumented Mexican immigrants had similar levels of knowledge about their child’s friendships, while documented Mexican immigrant mothers reported more knowledge (OLS coefficient of 0.15, Model 8). Ancillary analyses suggest that undocumented and documented Central-American immigrant mothers had the same knowledge of their child’s friendships. Regarding expected variations between Mexicans and Central Americans sharing the same documentation status, undocumented immigrants from both countries/regions had the same level of this knowledge (Model 8), as did documented immigrants in ancillary results. Thus, these results offer some support for the first hypothesis about similar or more advantageous maternal knowledge for undocumented Mexican immigrants relative to other Mexicans. However, contrary to the second hypothesis, the results indicate no residual differences in knowledge about their child’s friendships between Mexican and Central Americans sharing the same legal status.
As expected, social support was linked with maternal knowledge of child’s friendships. Mothers with biweekly family contact had more knowledge than those without this contact (0.05, Model 8). Compared to those who would ask immediate family for emergency childcare at night or for money, respondents who would ask “others” (babysitter, co-worker, and pastor) had less knowledge of their child’s networks (-0.54 and -0.32, respectively). These patterns suggest that biweekly family contact and being able to rely on immediate relatives are positively associated with more knowledge. However, this relationship may vary somewhat by the type of networks and kind of support. For example, those who had no one to ask for emergency childcare had more knowledge of children’s networks than those who would ask immediate relatives (0.29). Turning to the control variables, compared to mothers of 6–9 year olds, mothers of 16–17 year olds had less knowledge of child’s friendships, as did mothers with less formal education, single mothers, those in poor families, and those with more children in the household. Mothers who identified as Christian/Catholic were more knowledgeable about their child’s friends and those friends’ parents than were mothers of other religions/no religion. Finally, depressed mothers had less maternal knowledge, as did those with higher Pearlin self-efficacy values.
Discussion and Conclusion
Previous work has outlined more advantageous than expected characteristics and specific strengths present in Mexican immigrant families (Kao and Tienda 1995; Valdes 1998 Crosnoe 2006; Leidy et al. 2010; Jung et. al. 2012), with recent qualitative studies showing how undocumented Mexican immigrant parents actively work to help their families survive hostile climates (Ayón and Naddy 2013; Philbin and Ayón 2016). Broadly situated within the immigrant paradox, resilience, and family strengths perspectives, this article tests three hypotheses in a multivariate framework. First, we examined whether undocumented Mexican and Central-American immigrant mothers might have better-than-expected levels of family organization and maternal knowledge relative to documented immigrants and, in the case of Mexicans, compared to their documented immigrant and US-born counterparts. Next, we investigated potential heterogeneity in the family lives of Mexican and Central-American immigrants sharing the same documentation status. Finally, the analyses examined the connections of access to instrumental and emotional support with family organization and maternal knowledge.
The results supported our first hypothesis that undocumented immigrants report the same or higher levels of family routines and maternal knowledge as others from the same country/region for three out of four outcomes, controlling for other variables. For instance, undocumented Mexican immigrant mothers in these data had more frequent family dinners together, more predictable family routines, and equal knowledge of their child’s friendships as US-born Mexicans but less knowledge about whom their child was with. Undocumented Mexican immigrant mothers also had more frequent family dinners, more family routines, and equal knowledge of whom their child was with as their documented Mexican immigrant counterparts. Patterns were similar for undocumented Central-American immigrants compared to their documented Central-American immigrant peers.
Moreover, the results partially supported our second hypothesis that Mexican and Central-American immigrant mothers sharing the same umbrella legal status (i.e., documented and undocumented) have dissimilar family routines and maternal knowledge in the United States, observed for two out of four outcomes. Mexican and Central-American immigrant mothers sharing the same documentation status had different frequencies of family dinners and disparate levels of knowledge of whom the child was with but similar consistencies in family routines and similar knowledge of child’s friends and friends’ parents, net of other variables. In addition, the results consistently supported the third hypothesis that Mexican and Central-American mothers reporting more social support are likely to have more family routines and maternal knowledge than those indicating less social support, net of legal status and origin country/region. More frequent family contact was positively connected with more family dinners, more consistent family routines, and mothers knowing more of their child’s friends. Whom mothers could ask for help also matters. For example, mothers who relied on more distant networks for assistance typically had less knowledge of their child’s associations and friendships than those who could ask an immediate relative.
To be sure, undocumented Mexican and Central-American immigrant mothers and their families in these data experienced significant vulnerabilities in multiple domains such as being poor (Table 1), a common situation for many undocumented immigrants and their families (Yoshikawa 2011; Gonzales et al. 2013; Salas et al. 2013; Cardoso et al. 2018). At the same time, the lack of other differences by nativity and legal status is notable. For example, undocumented Mexican immigrant mothers reported similar rates of depression as other Mexicans—an encouraging sign, given that the analyses consistently indicated that depressed mothers reported less organization and knowledge about their child. Although we are unable to specify causal directions, such results suggest that undocumented Mexican immigrant mothers’ mental health may be an important force helping maintain family functioning in difficult circumstances. Being Christian/Catholic was independently associated with all four outcomes, suggesting that it could be fruitful to further explore the relationship of religious identity and family life among immigrant and US-born Latinx people. Another relationship to explore in future work is between mothers’ feelings of self-efficacy, their family organization, and their knowledge about their child’s associations and friendships. US-born Mexican mothers appeared to be disadvantaged in one way, with implications for the domains studied here: maternal employment. About two-thirds reported working 40 hours or more per week (Table 1)—a characteristic associated with fewer family dinners, family routines, and less knowledge of child’s associations (Table 2).
Taken together, our findings quantitatively demonstrate that undocumented Mexican and Central-American immigrants possess specific maternal and family-level strengths in domains known to promote positive youth outcomes and family well-being (Bámaca et al. 2005; Defrain and Asay 2007; Romero and Ruiz 2007; Benzies and Mychasiuk 2009; Cardoso and Thompson 2010; Leidy et al. 2010). For instance, children of undocumented Mexican immigrant mothers have more family dinners and more consistent family routines than children of US-born and documented Mexican immigrant mothers, before and after controlling for a wide range of other variables. The results also suggest that—at least in the early 2000s—Mexican and Central-American immigrants sharing the same documentation status were similar in some aspects of family organization and maternal knowledge but different in others, consistent with known variations between these groups (Hamilton and Stoltz Chinchilla 2001; Menjívar 2006; Abrego 2014; O’Connor et al. 2019). Consequently, the results complement and extend existing knowledge about what parenting strategies families with undocumented immigrant parents use to facilitate individual and family resilience, and how those strategies meaningfully vary by the mother’s origin country/region (Parra-Cardona et al. 2006; Raffaelli and Wiley 2013; Brabeck et al. 2016; Philbin and Ayón 2016).
The results also support Marks and colleagues’ (2014) recommendation that the immigrant paradox literature further investigate the role of legal status, family settings, and support structures as potential mechanisms for variation in the strength of the paradox across populations. Clearly, researchers should fully investigate how lacking legal authorization or having a liminally legal status in the United States might inhibit the better-than-expected outcomes described as part of the immigrant paradox. More scholarship examining the strengths, behaviors, and social networks that immigrants of varied legal statuses employ and draw on to survive hostile circumstances also could suggest new theoretical and empirical insights about the paradox. As households headed by undocumented immigrant adults are more likely to include children than those headed by US-born or documented immigrants (Passel and Cohn 2018), the importance and impacts of such efforts are further amplified.
Limitations of this article include the use of one wave of cross-sectional data to examine associations between variables, thereby eliminating the ability to make causal claims. Another is that the L.A.FANS does not have data about familism or acculturative stress that could be related to the outcomes documented here (Bámaca et al. 2005; Leidy et al. 2010; Marsiglia et al. 2014). In addition, although these preliminary results suggest that Mexican and Central-American immigrants are similar to each other in some ways, relatively small sample sizes in the data led to combining distinct statuses within a documented umbrella category (US citizens, Legal Permanent Resident, TPS, and holding a valid visa). L.A.FANS data also do not collect information about Central-American respondents' specific country of origin or ancestry. Thus, this article offers limited information about the intersectionality of legality, illegality and the often liminal, nonbinary experiences and diverse categories that US immigration laws create. More scholarship is needed about Salvadoran, Guatemalan, Honduran, and other Central-American immigrants of diverse documentation statuses, especially as these groups continue to increase in size in the United States (O’Connor et al. 2019). Although Los Angeles today is similar in racial diversity as at the time these data were collected (Clark et al. 2015), the national policy context for unauthorized migration has changed since that time (Pierce, Bolter, and Selee 2017). More recent, nationally representative, quantitative data sets with sufficient numbers of diverse Latin American immigrant groups that collect the detailed legal status of parents and include multi-dimensional indicators of social support, family functioning, and resilience are needed to disentangle these relationships further. Finally, bringing greater analytical attention to the often-invisible ways that immigrant parents, especially those who are undocumented or who experience liminal legal statuses, foster individual and family well-being in the ever-more challenging sociopolitical contexts is more urgent than ever.
Footnotes
Acknowledgments
A previous version of the paper was presented at the 2019 Annual Meeting of the American Sociological Association. We are grateful to Jordan Conwell, the Editor, and the anonymous reviewers for their valuable suggestions on earlier versions of the manuscript. The findings and conclusions of this article are those of the authors and do not represent the views of any funding agency.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Data collection for the Los Angeles Family and Neighborhood Survey (L.A.FANS) was funded primarily by grants HD35944 and HD49865 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, grant AG22005 from the National Institute on Aging, and grant ES13907 from the National Institute of Environmental Health Science
