Migration Aspirations and Adolescents’ Ideal age at Union Formation in Western Mexico

Abstract

Migration systems shape social life, including the timing and sequencing of key demographic behaviors such as marriage, childbearing, and household formation. Existing research has linked migration and marriage in Mexico through various mechanisms but provides less guidance on whether aspirations for migration and marriage are closely linked. Given that union formation is itself distinct within migration contexts, this article focuses on adolescents’ plans for marriage and the extent to which migration aspirations shape the desired timing of their own union formation by examining how four distinct measures of migration aspirations are related to adolescents’ ideal ages at marriage in rural Jalisco, Mexico. Drawing from data on adolescents (n=1,403 adolescents) from the Family Migration and Early Life Outcomes Project (collected in 2017–2018), it uses ordinary least squares regression to analyze how various types of adolescent migration aspirations – including permanent migration, temporary labor migration, leaving the community at any point in time, and expected migration location – are associated with adolescents’ ideal age at marriage. As the article shows, all migration aspirations are associated with higher ideal ages at marriage in unconditional models. However, these associations are not always robust to the inclusion of other factors, including adolescent aspirations in other life domains, particularly education. Results highlight the ongoing transition from a “culture of migration” to a “culture of education” in Mexico. Given that Mexican migration has changed dramatically in recent years, the findings presented here provide a window for understanding how these changes in migration are reflected in adolescent goals and likely subsequent behavior.

INTRODUCTION

Migration systems shape social life, including the timing, sequencing, and structuring of key demographic behaviors such as marriage, childbearing, and household formation (Kandel and Massey 2002; Parrado 2004; Parrado and Zenteno 2002; Lindstrom et al. 2021; White and Potter 2013). In sending communities with strong migration traditions, migration itself becomes expected for young people and is often a normative part of the transition to adulthood (Ali 2007; Horváth 2008; Kandel and Massey 2002; Zenteno et al. 2013). In the case of marriage or union formation, the market of potential and desirable partners is drastically different within Mexican sending communities compared to similar, non-migrant communities, at least for the short term (Parrado and Zenteno 2002; Parrado 2004; Riosmena 2009). Sizeable portions of many Mexican sending communities’ local populations, particularly men, may be gone for extended time periods, shifting caretaking and (at least temporary) financial responsibilities onto those who stay (Arias 2013). Further, Mexican men who migrate often postpone marriage, but the financial independence they gain through migration can allow them to transition to their first union shortly after return (Massey and Espinosa 1997; Parrado 2004). Social norms regarding union formation and dissolution also shift in response to migration’s prevalence in a community (Frank and Wildsmith 2005). The extant literature on migration and marriage provides less guidance, however, on whether aspirations for migration and marriage are closely linked. Given that union formation is itself distinct within migration contexts (Lindstrom et al. 2021; Massey and Espinosa 1997; Parrado 2004), this article focuses on adolescents’ plans for marriage and the extent to which migration aspirations shape the desired timing of their own union formations.

Adolescent aspirations for the future are important to study because adolescents are in the transitional space between child and adult. While not yet adults, their aspirations and decisions have important implications for future opportunities and outcomes (Eccles et al. 2004; Schoon 2001; Creighton 2013). As adolescents approach adulthood, they have increased agency to take definitive steps to shape their own futures (Arnett 2000). Their decisions – whether borne of their own desires, shaped by constraints, or shaped by a combination of both – can have far-reaching effects on their life trajectories (Elder 1998). Adolescents are simultaneously making decisions about different life domains, including union formation, childbearing, education, work, and migration (Alcaraz et al. 2022; Elder 1998). Their marital aspirations provide key insights into how adolescents see their futures, whether/when marriage is a part of those futures, and to what extent adolescents consider marriage to be important (Allendorf and Thornton 2015; Jalil and Fischer 2020; Shahrabadi et al. 2017; Willoughby 2010). Migration aspirations, in turn, reflect both individual goals for migration and migration’s salience in individual communities; they are also associated with individuals’ subsequent migration patterns (Creighton 2013; De Jong 2000).

To better understand the aspirational dimension of the association between migration and marriage within a Mexican context, this article examines how four distinct measures of migration aspirations – permanent migration, temporary labor migration, leaving the community at any point in time, and expected migration location – are related to adolescents’ ideal ages for union formation. The study discussed here is set in rural areas of the central western Mexican state of Jalisco, which has historically had high levels of out-migration, although these rates have decreased in recent years, as in the rest of Mexico (Masferrer and Roberts 2012; Pew Research Center 2015). Drawing from data on adolescents (n=1,403 adolescents) collected in 2017 to 2018 from the Family Migration and Early Life Outcomes Project (FAMELO), I use ordinary least squares (OLS) regression to examine how various types of adolescent migration are associated with marital aspirations, specifically adolescents’ ideal age at marriage/first union. I also examine variation in ideal age at marriage between boys and girls as a means of understanding potential gendered differences in the association between migration and marital aspirations. I find that adolescents’ migration aspirations were positively associated with their desired age at marriage, although some of this association was attenuated when household resources and adolescent aspirations in other domains were taken into consideration.

I divide the article into four sections. First, I review the theory and empirical research on the association between migration and union formation, adolescent aspirations for migration and union formation, and the migration and union formation context in Mexico. Next, I describe the data sources and method for my analysis of the association between migration aspirations and adolescents’ ideal age at union formation. I, then, review both descriptive results and findings from the multivariate OLS regression models. Finally, I discuss the importance and implications of my findings for improving scholarly understanding of the relationship between migration and union formation.

THE ASSOCIATION BETWEEN MIGRATION AND UNION FORMATOIN

The sequencing and timing of marriage and other transitions to adulthood are distinct within migration contexts (Parrado 2004; White and Potter 2013). A “culture of migration,” where migration becomes a normative part of life, emerges in many sending communities (Kandel and Massey 2002), as has been documented in Mexico (Kandel and Massey 2002), as well as in other countries including Ethiopia, Romania, and Ukraine (e.g., Horváth 2008; Ali 2007; Mol et al. 2018; Schewel 2021). In these locales, migration is not an unusual decision, and adolescents in sending communities expect to migrate one day (Ali 2007; Horváth 2008; Kandel and Massey 2002). In Mexico, migration, particularly to the United States, has been an alternative and culturally accepted route to upward social mobility in lieu of higher education (Bentaouet Kattan and Székely 2015). In rural western Mexico, for example, young men use migration to obtain financial independence prior to forming unions (Parrado 2004). Across Mexico, marriage and consensual unions (where partners cohabit and form families) are both considered appropriate partnering behaviors (Castro Martín 2002; Heaton and Forste 2007). Marriage is an extremely common institution embedded within broader Mexican society, the prominent Catholic tradition, and Mexico’s history with colonialism (Echarri Cánovas and Pérez Amador 2007), although consensual unions have become increasingly common since in the 1970s (Esteve et al. 2012). This article uses the terms ‘marriage’ and ‘union formation’ interchangeably, except where explicitly specified in the cited studies.

Existing research has linked migration and marriage through various mechanisms. First, migration has been shown to shape marriage markets in sending communities. Migrants are physically separated from their origin marriage market, and individuals who stay behind also contend with an altered marriage market (Adserà and Ferrer 2016; Carlson 1985). That is, the pool of marriageable adults is altered, at least temporarily, in high-migration contexts and this, in turn, may shape both the timing and possibility of union formation. Second, migration may be used as a strategy to accumulate the resources necessary to establish a financially independent household, as has been documented in Mexican migration to the United States (Massey and Espinosa 1997; Parrado 2004). In these cases, because migrants leave their home communities to meet these financial goals, they may be older when they get married upon returning. Unmarried Mexican men, for example, tend to stay in the United States longer than their single peers (Li 2016), and trips to the United States may delay marriage for these men. However, men who migrate may also transition to their first union shortly after they return, partly as a result of their improved financial situation (Parrado 2004). By contrast, adolescent men who form unions or become parents are less likely to migrate (Zenteno et al. 2013). Migrants may also present as more desirable partners to those in their home communities (Choi and Mare 2012).

Migration and various measures of union formation and stability are closely linked, particularly in Mexico (Frank and Wildsmith 2005; Parrado 2004; White and Potter 2013). In Latin America, men who migrate internationally are less likely to form unions at all and more likely to experience union dissolution. Women who marry international migrants are more likely to be older when they marry, compared to their peers who marry non-migrant individuals (Lindstrom et al. 2021). Migration, thus, shapes the timing of marriage not only for those who leave but also for those who stay. Migration is also associated with strained relationship quality and higher likelihoods of union dissolution (Frank and Wildsmith 2005; Boyle et al. 2008). Marital instability can be an outcome of migration and can itself lead to migration (Frank and Wildsmith 2005).

Much of what we know about the relationship between migration and marriage in Mexico has centered on male migrants’ experiences, in part due to extensive data on male migration to the United States (e.g., Princeton’s Mexican Migration Project study, https://mmp.opr.princeton.edu/). However, women’s experiences are also central to this discussion. When Mexican women have greater control over household resources and more power in the household broadly, their male partners are less likely to migrate (Nobles and McKelvey 2015). Mexican women’s primary motivation for migration is to join their partners or other family members in the United States (Cerrutti and Massey 2001; Creighton and Riosmena 2013). Married Mexican women are most likely to migrate within the first year after their first birth, although the likelihood of migrating to the United States decreases sharply after this period (Lindstrom and Saucedo 2002). Choosing to leave one’s home community in search of alternate economic opportunities is a migration decision, but so is choosing to stay – a decision made by many Mexican women (Mata-Codesal 2018; Nobles and McKelvey 2015). Research on Mexican women’s migration to the United States is itself growing (e.g., Arias 2013; Lindstrom et al. 2021; Nobles and McKelvey 2015; Parrado and Flippen 2005), but less is known about young women’s intentions to migrate and how their decisions are shaped by living within sending communities.

ADOLESCENT ASPIRATIONS FOR MIGRATION AND MARRIAGE

Moving beyond migration and marriage behavior, this article focuses on adolescent aspirations for both migration and marriage. Aspirations capture the most optimistic measure of what people might want. Expectations, by contrast, measure what people think will actually happen (Morgan 2007). Studying aspirations allows researchers to better capture what adolescents desire, while expectations may reflect perceived constraints. Aspirations are also different from intentions, which combine goals for the future with individuals’ understanding of the types of actions they will take to make these aspirations achievable (Morgan 2007; Hitlin and Kirkpatrick Johnson 2015). This article uses a broad definition of migration ideals and focuses on what adolescents would like for their future, using a mix of aspirations, expectations, and intentions.

Adolescent aspirations in different life domains are strong, if imperfect, predictors of their future outcomes (Eccles et al. 2004; Schoon 2001; Sewell et al. 1969). Aspirations may reflect young people’s own understandings of the types of goals that are not only feasible but also desirable (Baker 2017; Frye 2012; Silva and Snellman 2018). Adolescents’ aspirations are also indicative of broader cultural understandings of the appropriate timing and sequencing of key transitions to adulthood (Frye 2012). Young people must make decisions about different areas of their life around the same time, and the decisions they make in one domain influence their outcomes or opportunities in other domains (Arnett 2000; Clark et al. 2009).

Adolescent Migration Aspirations

I focus on adolescent migration aspirations, defined here as adolescents’ desires or expectations to migrate. Just as there are various forms of and motivations for migration, so, too, are there different types of migration aspirations. Migration can occur domestically, where individuals move within the same country, or internationally, where they move to a different country. Reasons for migration can include family reunification, schooling, economic opportunities, and persecution (Garip 2016; Massey et al. 1998; Portes and Rumbaut 2014). There is considerable variation on the duration, timing, and frequency of Mexican internal and international migration (Garip 2016; Donato 2010; Masferrer and Roberts 2012). While some migration is temporary, other migration can be permanent. Temporary migration may also be cyclical; that is, migrants may engage in repeated trips between their origin communities and migrant destinations. The life-course stage during which individuals migrate often varies according to the type of migration in which they engage, as does the frequency of migration (Donato 2010). Opportunities for and the consequences of migration may vary according to documentation status, a point of particular relevance in the Mexican case (Garip 2016; Gonzales 2011). Although migration to the United States has historically been a widely accepted and accessible route to upward mobility in Mexico, undocumented migration is often associated with a host of negative outcomes, including exploitation at work, poor health, mental distress, and family instability (Arbona et al. 2010; Dreby 2015; Gonzales 2011; Quesada et al. 2014). A full discussion of the factors associated with and outcomes resulting from different types of migration is beyond this article’s scope,¹ but individuals with aspirations to migrate are likely considering the possibility and potential outcomes of different types of migration.

Migration intentions and aspirations are informed by various factors, including gender norms, family norms about migration, community exposure to migration, financial need, and social capital (Boccagni 2017; Creighton 2013; De Jong 2000; Jalil and Fischer 2020). Individuals who aspire to migrate are more likely to actually do so than those who do not have concrete intentions to migrate, controlling for the structural conditions that make realizing migration aspirations possible (Carling and Schewel 2018). For example, Mexican men and women who have aspirations to migrate to the United States are very likely to subsequently migrate to the United States. Similar patterns exist for individuals who aspire to migrate domestically within Mexico (Creighton 2013). Self-selection likely plays into the migration process, as individuals with the material and social resources necessary to migrate may be more inclined to have aspirations to migrate in the first place. Still, the translation from aspirations and intentions to actual migration underscores aspirations’ utility for understanding how individuals think about how migration may interact with other decisions, especially the timing of marriage.

Adolescent Marital Aspirations

Adolescent marital attitudes encompass multiple dimensions, such as values and specific aspirations, including whether adolescents see marriage as an important part of their future families, the importance of marriage versus other life choices as forms of meaning-making, and the timing and circumstances necessary for marriage (Clark et al. 2009; Paat and Hope 2015; Pike et al. 2018). Marital aspirations tell us a great deal about individuals’ ideology, conceptions about the future, and the influence of their social contexts (Sassler 2010; Willoughby 2010). Understanding adolescent goals for marriage highlights existing gender inequality, social norms and expectations around marriage, as well as both perceived and real conflict or complementarity between marriage and other goals (Alcaraz et al. 2022; Bayer 1969; Zahra 2020). This article focuses on the desired timing of marriage, which lends insight into how adolescents temporally prioritize union formation and to what extent they see union formation as in conflict with other goals.

Decisions about the timing of union formation may have particularly strong effects on young adults’ ability and desire to work, get an education, or migrate to another location (Zenteno et al. 2013). Further, adolescents’ aspirations in other domains reflect the perceived tradeoffs and consequences of getting married versus other choices (Ajzen and Klobas 2013). Marital aspirations and attitudes also vary over time (Willoughby 2010), consistent with research in emerging adulthood that characterizes this age as a time of dynamic attitudes, beliefs, and behaviors (Arnett 2000). Research in the United States documents that as adolescents get older, they increasingly have pro-marriage attitudes as they approach their own transitions to adulthood (Willoughby 2010). Additional research on marital attitudes and ideals is necessary in high-migration contexts, where union formation patterns are themselves distinct.

In this article, I focus on intentions for union formation as a precursor to subsequent family formation. Much like other demographic behaviors, intentions to marry and actual marriage behavior are influenced by both individual and societal constraints and norms (Ajzen and Klobas 2013; Hirsch 2003; Nobles and McKelvey 2015). Around the world, boys and girls face different pressures to marry, and their social and economic outcomes are closely tied to these decisions. Within both developed and developing countries, gender egalitarianism is on the rise, but traditional roles continue to prevail; boys and girls both face pressure to follow traditional pathways to adulthood and meaning-making by marrying and having children (England et al. 2020; Frye 2017; Nobles and McKelvey 2015). In Mexico, young men face pressure to economically provide for their households, whereas young women’s limited opportunities may leave them dependent on their future spouses (Boehm 2008; Broughton 2008; Hirsch 2003). These pressures are not uniformly applied and are particularly complex in contexts undergoing rapid social change, including Mexico, where economic and educational opportunities for women are increasing even as gender roles remain salient (Zenteno et al. 2013).

The Relationship between Migration and Marriage Aspirations

Despite robust research on the relationship between marriage and migration, migration aspirations, and marital aspirations, less is known about the relationship between migration aspirations and marital aspirations. Understanding how adolescents think about their future marriages and to what extent plans for migration shape adolescent goals will shed light on how pervasive migration’s effects are on family formation in historical sending communities in western Mexico.

Given the long history and size of migration flows from Mexico to the United States and more recently to domestic destinations (Garip 2016; Pérez Campuzano and Santos Cerquera 2013), Mexico is uniquely suited for a study of the relationship between migration and marital aspirations. A strong association between migration aspirations and the desired timing of marriage shows the salience of the connection between migration and marriage, even at early stages of the life course. A weak or non-existent association, by contrast, would reveal the importance of other factors for understanding adolescents’ ideal age at marriage. As Mexico has undergone substantial and rapid social change in recent decades, including declines in international migration (Chort and de la Rupelle 2016; Masferrer and Roberts 2012; Pew Research Center 2015) and increases in educational expansion (Ducoing Watty 2018; Monroy and Trines 2019; Valentine et al. 2016), it is an ideal context for understanding how migration and marriage aspirations may or may not be linked.

Education – both enrollment and attainment – may be in competition with both marriage and migration. As in the United States (Goldstein and Kenney 2001; Manning et al. 2014), in Mexico, highly educated women are more likely to delay marriage and childbearing and to have smaller families (Kroeger et al. 2015; White and Potter 2013). Enrollment in education may be seen as temporally in conflict with family formation, particularly in contexts where women take on significant childrearing and household responsibilities. Self-selection may also play a role, where individuals who are more educated may be less inclined to prioritize early union formation or large families (Kroeger et al. 2015; White and Potter 2013). Further, research in Mexico often portrays education and migration as in conflict (Kandel and Massey 2002; McKenzie and Rapoport 2011; Meza González and Pederzini Villarreal 2018). While individuals may forego higher education in favor of migration to be upwardly mobile, migration is also commonly used as a strategy for accessing educational opportunities (Luschei 2012). While education and international migration have historically been at odds in Mexico, domestic migration and education often co-occur (Bernard et al. 2014; Valentine et al. 2016). Because education is closely tied to the timing of both marriage and migration in Mexico, it is possible that aspirations for education may mediate the association between education and ideal age at marriage.

THE CASE OF MEXICO

This article draws from data collected in rural Jalisco, Mexico, between October 2017 and June 2018. The study period came on the heels of steep declines in migration and increased educational expansion in Mexico (Ducoing Watty 2018; Pew Research Center 2012; 2015). I provide key information about education, migration, family formation, and gender inequality that helps contextualize the analysis. Educational attainment rates in Mexico remain relatively low, even in the face of large educational expansion initiatives that accelerated in the 1990s (Ducoing Watty 2018; OECD 2017). Since the 1990s, the Mexican government has focused on increasing access to education through building additional schools, enacting compulsory education laws, and providing cash or in-kind assistance to address issues of early drop-out and low educational attainment (Bentaouet Kattan and Székely 2015; Ducoing Watty 2018). Primary school enrollment became compulsory in 2009 and enrollment rates are at near 100% (Gobierno de México 2022; Monroy and Trines 2019). Secondary school enrollment across Mexico is high, but less than half of young adults complete upper secondary school (OECD 2017).

Access to the fruits of educational expansion in Mexico has been uneven, with urban settings benefiting more than rural areas (Balbuena Cruz et al. 2015) and underfunded schools common in rural Mexico (Zúñiga and Hamann 2009). Reliable access to public transportation can help families access educational opportunities that are not available in their home communities, but rural communities often lack adequate forms of transportation, including public transit, to connect them to other localities (Balbuena Cruz et al. 2015). Thus, while there have been important efforts to increase educational opportunities in Mexico, access is still stratified by both socioeconomic and rural/urban statuses.

Migration from rural Mexico to the United States and other areas has historically been primarily a male endeavor (Garip 2016). Young Mexican men are often expected to one day migrate (Docquier et al. 2009; Kandel and Massey 2002; Zenteno et al. 2013). Research in Mexico shows that, as expected, migration rates for men are highest in historically high-migration areas (Creighton and Riosmena 2013; Zenteno et al. 2013). For women, by contrast, migration rates are actually lower in these concentrated sending areas where traditional gender roles are salient (Zenteno et al. 2013). Declines in migration from Mexico to the United States began around 2007 and were accelerated by stagnating wages and reduced employment opportunities in the U.S. as a result of the Great Recession (Villarreal 2014). Migration from Mexico to the United States remained at a net negative, where more people left the U.S. for Mexico than vice versa, for about a decade from 2005 to 2015 (Pew Research Center 2015; Romo 2016). Prior to the COVID-19 pandemic, Mexican migration to the United States returned to a net, though slight, positive; it is likely that the ongoing pandemic and its ramifications will affect migration behavior (Gonzalez-Barrera 2021). However, less is known about whether these shifts in migration are reflected in how young people in Mexico think about future migration.

Other trends in family formation and gender equality measures in Mexico represent broader societal changes, including shifting orientations toward increased gender egalitarianism, education, and women’s autonomy (United Nations Development Programme 2020; Kroeger et al. 2015; Instituto Nacional de Estadística y Geografía (INEGI) 2018). Recent generations in Mexico are experiencing their first union in their mid-to-late-twenties (Kroeger et al. 2015). The age at marriage has also been rising in recent years, with the average age at marriage being 33 for men and 30 for women (Instituto Nacional de Estadística y Geografía (INEGI) 2020).² The most recent total fertility rate (TFR) is 2.18 average births per woman, representing a significant and continuing decline since the 1960s, when the TFR was around 6.75 births per woman in Mexico (World Bank 2018). The average desired family size for Mexican women is 2.1 children (Instituto Nacional de Estadística y Geografía (INEGI) 2018). Mexico has also experienced increased gender equality over the last few decades, as noted by a decrease in their gender inequality index (GII), a composite measure created by the United Nations Human Development Program that takes into consideration gender differences in levels of inequality across reproductive health, empowerment, and the labor market. GII is measured on a scale from 0 to 1, where “0” indicates perfect equality and “1” indicates perfect inequality. In 2019, Mexico’s GII measure was 0.322 and has been declining each year since the 1995 GII measure of 0.485 (United Nations Development Programme 2020).

CURRENT STUDY

Drawing upon existing research on the relationship between union formation and migration, migration aspirations, and marital aspirations, this article focuses on the relationship between adolescent migration aspirations and their ideal age at marriage in rural Jalisco, Mexico. More specifically, it focuses on how distinct measures of adolescents’ aspirations for migration are associated with the desired timing of their future marriage, controlling for individual- and household-level sociodemographic and attitude variables, as well as adolescent aspirations in other domains. The extant literature on migration and union formation, reviewed above, does not fully address the aspirational component of the association between various union formation and migration outcomes. Given that an expectation to migrate becomes normative in contexts with high levels of out-migration (Ali 2007; Kandel and Massey 2002), we might expect that adolescents adjust their plans in other life domains to account for migration. Still, steep declines in international migration in both Mexico and rural Jalisco, the difficulty of the trip, and other domestic opportunities for upward social mobility may decrease both the importance and possibility of migration for young people. In turn, migration may play less of a role in shaping the desired timing of marriage if migration is no longer expected or possible.

I hypothesize that adolescents’ migration aspirations will be positively associated with their ideal age at marriage (H1). I analyze migration aspirations through four measures designed to capture different types of migration: domestic, international, labor (and temporary), and permanent migration. Because migration of any kind may alter the timing of other key transitions to adulthood, I hypothesize that all measures of migration aspirations will be positively associated with the ideal age at marriage.

Some of this positive association may be due to household resources or adolescent aspirations in other domains that are related to both migration and marriage. For example, in Mexico, household resources have historically been linked with higher likelihoods of male international migration and the timing of that migration (Kaestner and Malamud 2014). Individual dispositions about when to have children, how much education to pursue, and views on gender inequality have also been linked to the timing of union formation (Allendorf and Thornton 2015; Hayford and Morgan 2008). By including measures of household resources and adolescent aspirations in other domains, I account for variables that may drive differences in the desired timing of marriage or migration aspirations. Still, I hypothesize that at least some of the association between marriage and migration aspirations is due to a sequential conflict between migration and marrying earlier. I expect that the positive association between migration aspirations and the ideal age at marriage will persist net of household- and individual-level controls, including adolescent aspirations (H2).

Mexican men who migrate often get married later in order to become more financially stable prior to forming families (Massey and Espinosa 1997; Parrado 2004). Further, men who marry early are less likely to migrate (Zenteno et al. 2013). Because migration has historically been a male-dominated endeavor in Mexico and closely tied to the timing of marriage for men, I expect that in models estimated with an interaction term between gender and migration aspirations, the association between migration aspirations and desired age at marriage will be larger for boys (H2a).

Lastly, migrating for education is relatively common within Mexico (Ducoing Watty 2018; Jalil and Fischer 2020; Valentine et al. 2016), and higher education levels are associated with delayed union formation (Kroeger et al. 2015; White and Potter 2013). I hypothesize that adolescent educational aspirations will account for some of the positive association between migration aspirations and the desired age at marriage (H3).

DATA AND MEASURES

Data

I use data from the Family Migration Context and Early Life Outcomes (FAMELO) project,³ a multi-disciplinary study of health and socioemotional outcomes for children in high-migration contexts. This project collects data in three field sites: Gaza province, Mozambique; Chitwan Valley, Nepal; and Jalisco province, Mexico (P01 HD080659, Glick, PI). Data collection took place between October 2017 and June 2018. I limit my analyses to the data from Jalisco, Mexico, which is uniquely suited to study various aspects of migration, given the well-established tradition of international migration to the United States and its proximity to the state capitol of Guadalajara, a primary destination of internal migrants (Pérez Campuzano and Santos Cerquera 2008; 2013). The existence of a robust empirical literature on marriage timing and migration in Mexico also calls for a focus on this context (e.g., Lindstrom et al. 2021; Parrado 2004; White and Potter 2013).

FAMELO is a two-stage household-based survey. Communities were selected from municipalities with high levels of migration. Within sampled communities, households were eligible if they had at least one child ages 5 to17 and a primary caregiver in the household. Within eligible households, one or two children were randomly selected for study participation. The number of children interviewed per household was randomly selected, and the selected child(ren) and the primary caregiver were both interviewed. Primary caregivers (usually the child’s mother) answered questions about themselves, other household members (both those currently living in and away from the household), and the child being interviewed. Children answered questions about themselves and their household. In the Mexico site, 2,268 households were sampled, including 3,249 children and 2,268 adult caregivers. Adolescents ages 11 to 17 are the unit of analysis within this article. I limit the analysis to this age group because questions about aspirations were not asked for children younger than 11 years old. I also use measures of household characteristics and parent information from the adult data. A total of 1,567 adolescents ages 11 to 17 were interviewed.

Measures

Dependent variable.

Adolescents’ ideal age at union formation is measured using the question, “At what age would you like to get married or live with a partner?” Ideal age at union formation is a continuous measure with a range of 13 to 52 years old (median: 25.7, standard deviation: 3.6 years).

Migration aspirations (independent variables).

I account for migration aspirations through four different variables. First, permanent migration aspirations are measured by the question, “Where would you most like to live when you grow up?” This question has five categories: here in this community, Guadalajara, somewhere else in Jalisco, somewhere else in Mexico, and in another country. Temporary, work-related migration aspirations are measured through a binary indicator that asked whether adolescents wanted to work in another country before getting married (1=yes). I use a binary measure of general migration aspirations, constructed from the question, “Do you think you will ever live somewhere other than this community?” This measure captures mobility in a broad sense – whether adolescents ever envisioned themselves leaving their home community (1=yes). Lastly, I constructed a variable that captured where adolescents thought they would live if they ever left the community. Adolescents who indicated that they planned on ever living elsewhere were asked, “Where do you think you will live if you leave this community?,” with five possible answers ranging from “another community in this municipality” to “in another country.” I created a new measure from this variable and added an additional category of “does not expect to leave this community.”

Control variables

I control for individual- and household-level attitude and sociodemographic variables. First, I control for adolescent aspirations in the domains of marriage, family, and education, as these aspirations may also be connected to adolescents’ desired timing of marriage. As adolescents’ views on marriage may determine when they would like to get married, I also control for how important adolescents believe marriage is, measured in the categories: not at all important, somewhat important, neither important nor unimportant, important, and very important. Desired family size is a continuous measure created from adolescents’ responses to the question, “If it were up to you to decide, how many children would you like to have in your life?” Adolescents who indicated that they did not know (n=124), that it was up to God or their spouse (n=17), or refused to answer (n=1) were coded as missing.

Educational aspirations are measured through adolescents’ responses to the question, “If you could get as much schooling as you want, how far in school would you like to go?” Adolescents were given 22 options ranging from “no school at all” to “doctorate.” According to the variable’s distribution, I recoded educational aspirations with the following three categories: less than a post-secondary degree, post-secondary degree, and post-graduate education. Adolescents’ desired education for their future partner is measured identically to adolescents’ educational aspirations.

Gender is a binary measure where female is coded as “1” and male is coded as “0.” Age is a continuous measure ranging from ages 11 to 17. I measure adolescents’ views on gender inegalitarianism with responses to the statement, “It is best if the husband makes decisions for the household and the wife follows his decisions.” Adolescents indicated on a scale of 1–5 whether they “strongly disagree” to “strongly agree” with the statement; a higher value corresponds with more gender inegalitarian views.

Household socioeconomic status is measured, using a principle components analysis (PCA) of household assets and dwelling construction materials (Filmer and Pritchett 2001). This method is commonly used in the Demographic and Health Surveys (DHS) program and in research on low- and middle-income countries (see Corsi et al. 2012), where gross income or other measures from the formal economy do not fully capture all available household resources. PCA identifies a set of indicators with the most shared variance and constructs a weighted average of them. These indicators capture long-term consumption and resources, rather than short-term changes in formal income (Filmer and Pritchett 2001). I coded all responses to these questions into binary indicators. In the final construction of the household wealth index used in these analyses, I retained variables with eigenvector absolute values of 0.2 or higher (13 total indicators). The variable used here is a score predicted using the first component of the final PCA.

Parent education is measured through the primary caregiver’s educational attainment, where adults were given the same 22 options as in the original measure of adolescent educational aspirations. I consolidated the variable into three categories: primary school or less, any lower-secondary school, and upper-secondary school or more. Total number of siblings is a continuous measure. History of migration in the household is measured, using a continuous variable that measures the number of household members (including those currently present and those currently away) who had ever migrated for a period of at least 6 months. Lastly, I control for household religiosity with a binary variable coded as “1” if the primary caregiver attended religious services weekly or more.

Missing Data.

All adolescents ages 11 to 17 (n=1,567) were asked about their ideal age at marriage. Thirty-two adolescents noted that they did not want to get married; these observations are coded as missing. Additionally, 130 adolescents answered that they did not know their ideal age at marriage, and two adolescents refused to answer. These responses were coded as missing, bringing the final sample size to 1,403 adolescents. I, then, accounted for missing information on independent and control variables through multiple imputation, using the ice chained equation command in Stata and running 25 imputations (Royston and White 2011). Most independent and control variables were missing less than 6% of observations, with only expected migration location missing 8.1% of observations.

METHOD

Descriptive statistics (Table 1) are given for all non-missing observations for each corresponding variable. I, then, use OLS regressions to predict adolescents’ ideal age at marriage, as this measure is continuous and has a normal distribution. To account for within-household variation due to the fact that up to two children were interviewed per household, I cluster all analyses by household. I run eight separate models predicting ideal age at marriage (Table 2), using different measures of migration aspirations. For each measure of migration aspirations, I first estimate an unconditional model where I examine the bivariate association between migration aspiration and ideal age at marriage. Then, I estimate a conditional model that includes all control variables. Ideal age at marriage is predicted, using the following migration aspirations: permanent migration (Models 1 and 2); working abroad before marriage (Models 3 and 4); expecting to live somewhere else at any point (Models 5 and 6); and expected migration location (Models 7 and 8).

Table 1.

Descriptive Statistics (N = 1,403 adolescents)

	Full Sample		Male		Female
	Mean/Prop. (SD)	n	Mean/Prop. (SD)	n	Mean/Prop. (SD)	n
Ideal age at marriage (Range: 13–52)	25.7 (3.4)	1403	26.2 (3.5)	726	25.1 (3.1)	677
Permanent migration aspirations (Where would you like to live when you grow up?)		1359		701		658
Here in this municipality	.56		.57		.56
Guadalajara	.06		.06		.06
Somewhere else in Jalisco	.08		.07		.09
Somewhere else in Mexico	.09		.08		.10
In another country	.21		.22		.19
Wants to work in another country before marriage	.67	1370	.67	709	.67	661
Expectation that you'll live somewhere other than this community	.77	1324	.77	684	.76	640
Migration location (Where do you think you will live if you leave this community?)		1289		669		620
Does not expect to leave	.24		.23		.24
Another community in municipality	.02		.02		.02
Guadalajara	.17		.17		.17
Somewhere else in Jalisco	.17		.16		.18
Somewhere else in Mexico	.15		.14		.16
In another country	.25		.28		.22
Female	.48	1403	-	-	-	-
Age (Range: 11–17)	13.9 (1.9)	1403	13.9 (1.9)	726	13.8 (1.9)	677
Educational aspirations		1383		710		673
Less than post-secondary degree	.26		.34		.18
Post-secondary graduation	.67		.59		.74
Post-graduate education or higher	.07		.07		.07
Desired education of future partner/spouse		1320		672		648
Less than post-secondary degree	.26		.36		.16
Post-secondary graduation	.69		.59		.80
Post-graduate education or higher	.04		.05		.04
Importance of marriage		1386		713		673
Not at all important	.04		.03		.05
Somewhat important	.18		.16		.20
Neither important or unimportant	.18		.15		.20
Important	.45		.48		.41
Very important	.16		.18		.14
Desired family size (Range: 0–12)	2.6 (1.0)	1329	2.7 (1.0)	674	2.5 (1.0)	655
Husbands should make decisions	2.3 (1.1)	1394	2.5 (1.1)	719	2.0 (1.0)	675
Primary caregiver education		1384		712		672
Primary school or less	.36		.36		.36
Any lower-secondary school	.48		.48		.49
Upper-secondary school or more	.16		.16		.15
Household wealth index (range: −10 to 10)	0.1 (1.9)	1403	0.1 (1.9)	726	0.1 (2.0)	677
Number of household members who have ever migrated	0.6 (1.0)	1384	0.6 (1.0)	712	0.5 (0.9)	672
Total number of siblings (Range: 0–9)	2.1 (1.4)	1383	2.1 (1.4)	711	2.1 (1.4)	672
Weekly religious attendance	.65	1348	.65	689	.65	659

Table 2.

OLS regressions predicting the ideal age at marriage by individual migration aspirations measures

	Permanent migration		Work abroad before marriage		Expect to live elsewhere		Migration location
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7	Model 8
Where would you like to live as an adult? (vs. here in this community)
Guadalajara	0.97**	0.48
	(0.32)	(0.30)
Somewhere else in Jalisco	1.10**	0.68*
	(0.34)	(0.33)
Somewhere else in Mexico	0.51	0.12
	(0.33)	(0.32)
In another country	0.70**	0.20
	(0.24)	(0.24)
Wants to work abroad before marriage			0.62**	0.20
			(0.20)	(0.19)
Expect to ever live somewhere other than this community					0.95***	0.40+
					(0.23)	(0.23)
Where do you think you will live if you leave this community? (vs. does not expect to leave)
Other community in municipality							1.26+	0.82
							(0.68)	(0.65)
Guadalajara							1.36***	0.67*
							(0.30)	(0.30)
Somewhere else in Jalisco							0.87**	0.33
							(0.30)	(0.30)
Somewhere else in Mexico							0.28	−0.15
							(0.31)	(0.31)
In another country							1.14***	0.56*
							(0.28)	(0.28)
Gender and Educational Aspirations
Female (vs. Male)		−1.46***		−1.46***		−1.44***		−1.42***
		(0.18)		(0.18)		(0.18)		(0.18)
Child educational aspirations (vs. Less than post-secondary degree)
Post-secondary degree		0.87***		0.93***		0.87***		0.89***
		(0.24)		(0.24)		(0.24)		(0.24)
Post-graduate education		2.34***		2.42***		2.36***		2.35***
		(0.44)		(0.44)		(0.44)		(0.45)
Controls Included?	N	Y	N	Y	N	Y	N	Y
n	1,403	1,403	1,403	1,403	1,403	1,403	1,403	1,403
R-squared	0.01	0.13	0.01	0.12	0.01	0.12	0.02	0.13
Adjusted R-squared	0.01	0.11	0.01	0.11	0.01	0.11	0.02	0.12
Degrees of Freedom	1,156	1,156	1,156	1,156	1,156	1,156	1,156	1,156
F	5.23***	8.34***	9.92**	9.17***	16.89***	9.23***	5.74***	7.81***

Notes: All estimates derived from imputed data. Unstandardized coefficients presented, with robust standard errors in parentheses.

^***p<0.001

^**p<0.01

^*p<0.05

⁺p<0.1.

Control variables: gender, educational aspirations, age, husbands should make household decisions, importance of marriage, household SES, number of siblings, number of household members who have ever migrated, child educational aspirations, desired education of partner, parent education, and religiosity.

RESULTS

Descriptive Results

All descriptive statistics, both for the full sample and for boys and girls individually, are presented in Table 1. The average ideal age at marriage (range: 13–52) was about 26 years old, with a standard deviation of about 3.5 years. On average, girls in the sample had an ideal age at marriage 1.1 years lower than boys. The sample is about evenly split between boys and girls (48% female). The average age of adolescents in the full sample was about 14 years old.

When asked where adolescents would like to live when they were adults (a measure of permanent migration aspirations), most adolescents wanted to remain in the same community/municipality in adulthood (56%), followed by wanting to live in another country (21%), somewhere else in Mexico (9%), somewhere else in Jalisco (8%), and Guadalajara, Jalisco’s capital city (6%). About 67% of respondents wanted to work in another country before getting married. About 77% of adolescents expected that at some point in their lives, they would live somewhere other than their home community. While about 24% of adolescents did not expect to ever leave their home community, of those who expected to leave, the modal category for expected migration location was in another country (25%). Other adolescents expected to migrate (at least temporarily) to another community in the same municipality (2%), Guadalajara (17%), somewhere else in Jalisco (17%), or somewhere else in Mexico (15%). Overall, adolescent migration aspirations were relatively high. Worth highlighting is the fact that girls and boys had almost identical migration aspirations, despite the fact that migration in Mexico has typically been male dominated.

Most children aspired to obtain a bachelor’s degree (67%) and wanted the same for their future partner (69%). On average, adolescents wanted 2.6 children in their future families. Most adolescents believed that marriage was important or very important (61%). However, more girls than boys considered marriage to be either not at all important or somewhat important or were neutral about marriage (45% versus 34%). Boys were more likely than girls to agree that husbands should have more decision-making power in their families, with an average value of 2.5 for boys (on a scale of 1–5, where higher values indicate more inegalitarian views) compared to 2.0 for girls.

The modal category for primary caregiver education was any lower secondary school education (48%). The average number of household members who had ever migrated for at least six months was 0.6 migrants. Adolescents in the sample had an average of 2.1 siblings. Over half of adolescents in the sample (65%) had a primary caregiver who attended religious services at least weekly. Additional information can be found in Table 1.

Ideal Age at Marriage and Migration Aspirations

I now consider the association between migration aspirations and adolescents’ ideal age at marriage. Models 1 and 2 examined the association between permanent migration aspirations and ideal age at marriage. Models 3 and 4 included adolescents’ desire to work in another country before getting married. Models 5 and 6 used a measure of migration aspirations that captures the expectation that adolescents would leave their community at any point. Models 7 and 8 examined expected migration location. Results are presented in Table 2 in the form of coefficients and standard errors. For ease of interpretation, I only included coefficients for migration aspirations and two important control variables: gender and educational aspirations. Tables with full results for all variables in the model can be found in Tables A1 through A4 in the Online Appendix. I discuss the main results before examining the role of gender and educational aspirations.

In the first unconditional model (Model 1), permanent migration aspirations were significantly associated with adolescents’ ideal age at marriage, providing support for H1. Desiring to live in a place other than adolescents’ home community was associated with a higher ideal age at marriage for all locations (p<.01), except somewhere else in Mexico. When control variables are included, the size and significance of multiple comparisons were attenuated or disappeared altogether, with one exception. Adolescents who wanted to live somewhere else in Jalisco had an ideal age at marriage 0.68 years (about 8 months) higher than those who wanted to remain in their home community (p<.1), although this association was only marginally significant. Thus, H2 is not supported in the case of permanent migration aspirations because the association between migration aspirations and desired timing of union formation was not robust to the inclusion of control variables.

In Model 3 with no controls, adolescents who wanted to work abroad before marriage had an ideal age at marriage 0.62 years (about seven and a half months) higher than those who did not (p<.01). However, the size of this effect decreased, and the statistical significance disappeared, when control variables were included in Model 4. Thus, H2 is not supported. That is, desires to migrate temporarily for work were positively associated with the ideal age at marriage, but other variables better accounted for the variance in the ideal age at marriage.

Similar results exist for adolescents who ever expected to live somewhere other than their home community. In Model 5 without any control variables, adolescents who expected to live elsewhere at any point had an ideal age at marriage 0.95 years higher than those who did not expect to leave (p<.001). When control variables are added in Model 6, this migration aspiration was only marginally associated with the ideal age at marriage (p<.1). Thus, H2 is not supported for migration aspirations measured as adolescents’ expectation to live elsewhere at any point.

In Model 7 with no controls, adolescents who expected to live in Guadalajara, somewhere else in Jalisco, or in another country all had higher ideal ages at marriage than their peers who did not expect to leave the community (p<.01). In this unconditional model, H1 is supported. Adolescents who expected to live in Guadalajara if they left their community wanted to get married on average 1.36 years later than those who did not expect to leave (p<.001). Adolescents who expected to live somewhere else in Jalisco or in another country had ideal ages at marriage 0.87 and 1.14 years higher than those who expected to stay, respectively (p<.01). Once control variables are added to Model 8, some of these associations persisted, although they were more modest in size, lending partial support to H2. That is, adolescents who expected to live in Guadalajara had an ideal age at marriage 0.67 years higher than those who expected to remain in their home community (p<.05). Similarly, adolescents who planned to live in another country had an ideal age at marriage 0.56 years higher than those who did not expect to leave (p<.05). As a reminder, this measure of migration aspirations captures expected migration location if adolescents ever expected to leave for any reason or amount of time.

Gender and Educational Aspirations

Based on the results above, it is clear that various types of migration aspirations are related to when adolescents would like to get married. However, the associations between migration aspirations and adolescents’ ideal age at marriage disappeared or were attenuated when control variables were included in the conditional models. In Table 2, I highlight two variables that are key to understanding these associations: adolescent gender and educational aspirations.

Across all conditional models (Models 2, 4, 6, and 8), girls wanted to get married, on average, about 1.5 years earlier than boys (p<.001). Of particular interest is the robust and large association between child educational aspirations and ideal age at marriage. Across all conditional models, adolescents who wanted a post-secondary degree had a desired age at marriage almost one year higher than those who aspired to less than a post-secondary degree (p<.001), supporting H3. Especially striking is the effect for adolescents who aspired to a post-graduate education, who wanted to get married, on average, about two and a half years later than adolescents who aspired to less than a post-secondary degree (p<.001). Education appears to be closely linked with adolescents’ desired age at marriage. Recent research underscores how education and migration are increasingly linked in Mexico (Valentine et al. 2016), as I discuss below. Taken together, gender and educational aspirations appeared to mediate the relationship between migration and marriage aspirations.

In additional analyses (available upon request), I included an interaction between adolescent gender and migration aspirations in all four, full conditional models (Models 2, 4, 6, and 8). These interactions did not improve the models’ fit; interaction terms were also not statistically significant, meaning that migration aspirations’ effect on ideal age at marriage did not significantly differ for boys and girls. Thus, H2a is not supported. All other associations remained largely similar to those presented in the full models. A summary of all results appears in Table 3.

Table 3.

Summary of Results for All Migration Aspirations

Hypothesis	Permanent migration	Work abroad before marriage	Expect to live elsewhere	Migration location
H1: Migration aspirations and ideal age at marriage are positively associated.	Supported (partially robust to controls)	Supported (not robust to controls)	Supported (not robust to controls)	Supported (partially robust to controls)
H2: This relationship persists even when accounting for household- and individual-level controls, including adolescent aspirations.	Partially supported	Not supported	Not supported	Partially supported
H2a: This relationship is stronger for boys than for girls	Not supported	Not supported	Not supported	Not supported
H3: Educational aspirations account for some of the positive association between migration aspirations and desired age at marriage.	Supported	Supported	Supported	Supported
Other independent variables significantly associated with ideal age at marriage?	Female (−); Age (+); Husbands should make decisions (−); Educational aspirations (+)	Female (−); Age (+); Husbands should make decisions (−); Educational aspirations (+)	Female (−); Age (+); Husbands should make decisions (−); Educational aspirations (+)	Female (−); Age (+); Husbands should make decisions (−); Educational aspirations (+)

Notes: Addition and subtraction signs after listed variables denotes the direction of the association between the respective variable and ideal age at marriage.

DISCUSSION AND CONCLUSION

This article has examined the association between migration aspirations and adolescents’ ideal age at marriage in rural Mexico to better understand the aspirational dimension of the relationship between migration and marriage. Overall, I find that among young people in Jalisco, Mexico, migration aspirations were associated with higher ideal ages at marriage for adolescents in unconditional models. That is, adolescents may view migration as disruptive to the timing of other transitions to adult roles and, thus, delay the desired timing of marriage. This association holds for all measures of migration aspirations, including permanent migration, temporary labor migration, and general measures of migration. However, when individual- and household-level sociodemographic variables and adolescent aspirations in other domains were accounted for, many of the associations between migration aspirations and ideal age at marriage were attenuated or no longer statistically significant. Below, I discuss potential reasons for the variation in robust associations between different measures of migration and desired marriage timing in the conditional models.

Adolescents who planned to live somewhere else in the Jalisco province (but not Guadalajara) when they were adults had a higher ideal age at marriage, although this effect was only marginally significant. The significance of moving to another comparable location in Jalisco specifically may be explained by adolescents’ desires for education and to remain in or return to their hometowns. For example, most universities in Jalisco are located in the state capital, Guadalajara. Adolescents who wanted to get a higher education may have wanted to move to Guadalajara, but these effects may be better picked up by educational aspirations in the model, particularly because education and marriage timing are often strongly and positively associated (Choi and Mare 2012; Kroeger et al. 2015). Further, it is possible that boys and girls may have wanted to migrate to other locations temporarily but ultimately planned to return to their home communities and presumably wanted to get married when they returned. Temporary moves for things like education, work, and life experience are unfortunately not captured in the data used in this article.

Although a significant proportion of adolescents wanted to work in another country prior to getting married (66%), this measure of temporary, international labor migration was not significantly associated with desired age at marriage in the conditional model (Model 4). Contrary to prior research (e.g., Parrado 2004; Zenteno et al. 2013), adolescents may not have planned to leverage temporary international migration as a means to financial stability prior to union formation. Future research should continue to examine the relationship between migration and partnering behavior (Lindstrom et al. 2021), as it is currently unclear whether the timing of this association has shifted or whether the association has changed altogether. This relationship may be further complicated by the differing reasons for and opportunities resulting from internal and international migration, as well as temporary, cyclical, or permanent migration.

Adolescents who expected to live elsewhere at any point had a higher desired age at marriage than those who did not expect to leave, but this association was only marginally significant (Model 6). Expected migration location remained positively associated with ideal age at marriage, even after introducing controls (Model 8). Adolescents who expected to live in Guadalajara or another country if they left their community had a desired age at marriage about 0.6 years higher than those who did not expect to leave. Although this measure of expected migration location is similarly phrased to the measure of permanent migration aspirations, it actually points to different considerations that adolescents make about migration. When adolescents are asked about where they want to live as adults, they are likely thinking about a settled future (Swidler 1986) and where they see themselves setting up a permanent residence. Where adolescents see themselves living as adults is substantively different from asking adolescents if they ever see themselves leaving the community and, if so, where they envision relocating. Whereas asking where adolescents would like to live as adults more directly captures permanent migration aspirations (or the lack thereof), the measure I use for expected migration location gave adolescents space to acknowledge either permanent or temporary migration. While a temporary relocation may increase the desired age at marriage, particularly if this migration occurs around the normative time of marriage, expected adult location may seem farther away to adolescents and, thus, may not be directly associated with the ideal age at marriage when other variables are taken into consideration.

When we consider the role of other adolescent aspirations, a new story emerges. Particularly notable is the large and positive association between educational goals and ideal age at marriage across all conditional models, regardless of how migration aspirations are measured. Adolescents with more ambitious educational goals wanted to get married later than those with lower educational goals. In fact, adolescents who wanted a post-graduate education wanted to get married nearly two and a half years later than adolescents who wanted less than a post-secondary degree. This association may be evidence that adolescents see enrollment in higher education and getting married as incompatible or at least somewhat in conflict.

Education’s mediating effects on the association between migration aspirations and desired timing of marriage may also result from the fact that education and migration are increasingly linked in Mexico. Recent research has highlighted how, in the context of declining opportunities and benefits from migration(particularly to the United States), individuals in Mexican sending communities focus on education as a means of being upwardly mobile (Valentine et al. 2016). That is, there appears to be an ongoing transition in the Mexican context from a “culture of migration” (Kandel and Massey 2002) to a “culture of education” (Valentine et al. 2016). These two goals are not necessarily in conflict; in fact, young people in Mexico are often required to leave their rural villages to access higher education (Ducoing Watty 2018; Valentine et al. 2016). Migration in and out of Mexico has historically been male dominated, but in the face of rapid educational expansion, education may now take the place of migration in helping accomplish the financial goals of both male and female partners prior to marriage, although more research is needed in this area. This growing focus on education as a means for upward mobility in Mexico may also help explain the lack of gender differences in the association between migration aspirations and desired timing of marriage. Existing research on this more contemporary focus on education in place of or in conjunction with migration in Mexico is somewhat limited, particularly because much of the data used in this research were collected prior to sharp declines in Mexican migration to the United States (Pew Research Center 2012; 2015). Examining the complementary relationship between education and migration is beyond this article’s scope, but future research should continue to examine how young people view family formation, education, and other transitions to adulthood in light of increasingly rare opportunities for migration from Mexico to the United States.

This article is not without limitations. First, although it uses rich data from Mexico, the data are representative of only the current setting and encompass a single point in time. Because the data are cross-sectional, they do not lend themselves to analysis of change over time. Rapid social change in migration flows, access to education, and gender equality reforms serve as the backdrop to the patterns examined here (Gonzalez-Barrera 2021; Ducoing Watty 2018; Nobles and McKelvey 2015; United Nations Development Programme 2020). These changes likely work together to shape both the current behavior and aspirations of young people and their ideas about what is possible in the future. Future research should continue to examine marital aspirations in diverse high-migration contexts through both qualitative and mixed-methods approaches. Second, while a strength of this article is the use of multiple migration aspirations, the variables used in this analysis are not necessarily mutually exclusive. Additional data on migration aspirations may allow researchers to better parse out exactly which types of migration may be associated with the desired timing of marriage. Lastly, data on adolescent relationships are not available for the full sample of adolescents. It is possible that adolescents in relationships may view leaving their home communities and marriage differently than those who are not in romantic relationships.

This article yields important insights into how migration aspirations are related to adolescents’ desired marriage timing. Whereas the prior literature largely focused on the relationship between marriage and migration behavior (e.g., Lindstrom et al. 2021; Riosmena 2009; Parrado 2004; White and Potter 2013), this analysis departs from this stream of research by focusing on the aspirational component of this relationship. This article is also one of the first to examine migration and marriage in the shifting migration context in Mexico, and additional work remains to be done in this area. Migration aspirations appear to be positively associated with adolescents’ desired age at marriage, mirroring similar trends in actual migration and age at first union. However, these associations are not always robust to the inclusion of other factors, including adolescent aspirations in other life domains, particularly education. Given that Mexican migration, both domestic and international, has changed dramatically in recent years, the findings presented here provide a window for understanding how these changes in migration are reflected in adolescent goals and likely subsequent behavior.