Abstract
Recent scholarship proposes a “two-step” approach for better understanding mechanisms underlying the migration process, suggesting we study migration aspirations separately from migration behavior and that the one does not always translate directly into the other. Research on aspirations, however, concentrates on the Global South, despite growing migration flows originating in the Global North. Here, we fill this gap, drawing on a nationally representative online survey we commissioned in 2014 in the United States. Bivariate analysis shows that fully one-third of Americans surveyed reveal some aspiration to live abroad, a plurality of those primarily for the purpose of exploration. Multivariate analysis suggests that certain elements of cultural and social capital, including the networks Americans have with prior and current US citizen migrants, structure these aspirations, in tandem with strength of national attachment. Further, both cultural and economic aspects of class, alongside race and national attachment, shape where American aspirants envision going and why. While the existing literature addresses the motivations and profile of American migrants already living abroad, ours is the first study to examine Americans’ aspirations prospectively from the point of origin, thereby connecting the literature on Global North migration flows to that on migration aspirations.
Introduction
Recent scholarship proposes a “two-step” approach to understanding the migration process, calling for separate analysis of, on one hand, individuals’ evaluations of migration as a potential course of action and, on the other, translation of such evaluations into either mobility or immobility (Carling and Schewel 2018). Renewed interest in migration aspirations revives a strand of research from the 1980s and 1990s that once looked deeply into the micro-dynamics of migrants’ pre-migration decision-making (De Jong et al. 1985; Sandu and De Jong 1996; De Jong 2000). It also takes inspiration from recent developments, among them the continued tightening of borders in Europe and North America, as well as the renewed social scientific interest in emotions, temporalities, and ongoing migration “projects” (Carling and Collins 2018).
We contribute to reemerging research on migration aspirations with a case study of US-born US citizens. We commissioned a unique, nationally representative online survey conducted in the United States in 2014 and, using that data, provide a quantitative picture of 877 Americans’ aspirations to live abroad, including their preferences for where they would like to go, for how long, and why. In line with the two-step aspirations approach, we measure their aspirations prior to potential migration and from the point of origin. Acknowledging that migration aspirations are “socially embedded and culturally informed” (Bal and Willems 2014, 249; see also Salazar 2011; Benson and Osbaldiston 2016; Koikkalainen and Kyle 2016), we also model the influence of key cultural, financial, and social capital variables on each dimension.
Why study Americans’ migration aspirations? Because the number of Americans living abroad is growing, yet this group remains understudied, even as the utility of studying migration aspirations has been demonstrated in other Global North contexts (Van Dalen and Henkens 2013; Scheibelhofer 2018; Williams et al. 2018). In 2016, the US Department of State estimated that approximately nine million US citizens lived abroad, equivalent to 3 percent of the total US citizen population and a substantial increase over its prior estimates of 7.6 million in 2013 and 6.3 million in 2012 (Klekowski von Koppenfels 2014, 32). Mexico and Canada are currently the top two destinations of US citizens living abroad, followed by the United Kingdom, Germany, Australia, and Israel (C. Smith 2014; OECD 2015, 256; Pew Research Center 2016). In Mexico alone, US citizens are now the largest immigrant population, at an estimated 850,000 and growing (Schafran and Monkkonen 2011).
Still, estimates of how many US citizens live abroad vary (see also C. Smith 2010; Fors Marsh 2016), not only because different countries employ different methods of counting Americans with varying visa and citizenship statuses (Klekowski von Koppenfels and Costanzo 2013) but also because not all Americans self-identify, or are seen as, migrants in the first place (Croucher 2009b; Klekowski von Koppenfels 2014; Croucher 2015). This makes generating a representative sample from an unknown population base tricky at best, impossible at worst. It also means that our knowledge of Americans’ motivations for migration is typically measured retrospectively, after migration (e.g., Trundle 2009; Croucher 2012; Klekowski von Koppenfels 2014), and typically focuses on migrants who are most visible or living in areas of high concentration. In contrast, emerging scholarship on migration aspirations collected at the point of origin centers primarily on flows originating in the Global South (Scheibelhofer 2018), offering little insight on how migration aspirations arise among Americans or other residents of economically advanced countries in the Global North.
The 2007 to 2013 World Gallup Polls are one exception (OECD 2015). Using their microdata, Docquier, Peri, and Ruyssen (2014) find few differences in how a host of economic, social, cultural, and policy factors shape migration aspirations among all potential migrants in 138 countries worldwide. However, their study was conducted at the aggregate level, leaving space for finer-grained analysis. How many individuals in societies of the Global North aspire to move and live abroad? If so, where, for how long, and why? What does such potential movement mean to them? To begin addressing these questions, we start by situating our US case study within the literature on migration aspirations. After introducing our primary data and methods, we present results from bivariate models showing how many Americans reveal aspirations to migrate and, if so, to where, for how long, and why. Next, we construct a range of multivariate models to compare the relative influences of cultural, financial, and social capital on each dimension. We conclude by discussing our main findings and their implications, the limitations of our study, and ideas for future research.
Migration Aspirations in Social Context
According to Scheibelhofer (2018, 2), aspirations are “hopes, plans, ambitions or goals that can be clearly formulated or kept rather vague” and that simultaneously address both the present and the future. Scheibelhofer, studying Austrians, argues that aspirations are not purely situational but instead “rather enduring” in their influence over an individual’s life course. Indeed, prospection — the representation of possible futures — can function psychologically as “mental simulation” to shape human action, as individuals engage, in different ways, in “episodic future-thinking” (Seligman et al. 2013, cited in Koikkalainen and Kyle 2016). Set within a broader capabilities framework, Appadurai (2004) argues that aspiration is a future-oriented cultural capacity that is strongly classed — more open to and commonly enjoyed by more affluent individuals. Other scholars argue even more forcefully that imagination, as well as the ability to act upon it, is shaped by not only historical context but also wider cultural repertoires and power dynamics (A. Smith 2006; Benson 2012).
Migration aspiration, specifically, refers to the basic conviction that leaving a particular place would be better than staying (Carling and Collins 2018, 7). “Inherently elusive” (Carling and Schewel 2018, 4), migration aspiration can exist along several continua — including one moving from strong aspiration to leave to strong aspiration to stay, the (in)voluntariness of potential migration, and the degree to which plans have (not) been solidified (Carling 2002; Bivand Erdal and Oeppen 2018; Carling and Collins 2018; Carling and Schewel 2018). As with other types of aspirations, migration aspiration varies across time and space (Van Mol et al. 2018) and likely reflects class and other stratification in people’s capabilities to imagine migration as one facet of their broader goals and outcomes (de Haas 2014).
By contrast, migration behavior, Carling and Schewel (2018) contend, is a separate stage and object of analysis, although it may sometimes be correlated with migration aspiration. Certainly, individuals’ migration intentions have been shown to be moderate to strong predictors of actual migration patterns in a variety of national contexts, including the Netherlands and Mexico (see De Jong et al. 1985; De Jong 2000; Creighton 2013; Van Dalen and Henkens 2013; Czaika and Vothknecht 2014; Docquier, Peri, and Ruyssen 2014; Theiss-Morse and Wals 2014; Wals and Moreno n.d.). Still, aspirations, intentions, and even self-reported likelihoods of migrating do not always translate seamlessly into actual migration behavior (Gardner et al. 1985; de Groot et al. 2011). An individual may have ideas or preference to migrate but lack the ability or resources (whether information, social networks, or money) to do so, resulting in what Carling (2002) calls “involuntary immobility.” Other people can migrate without ever having developed an aspiration to do so — whether enticed by an unanticipated job offer or coerced by natural or political disaster (Kokkalainen and Kyle 2016). Finally, individuals may adjust their migration aspirations over time.
For all these reasons, Koikkalainen and Kyle (2016) argue that aspirations should ideally be measured prior to mobility taking place and should include people who do not necessarily wish to migrate as well as aspirants who never come to move. In practice, however, most studies “sample on the dependent variable” twice — first, by analyzing only individuals who have already migrated, and second, by measuring their aspirations retrospectively. In response, a wave of studies on migration aspirations have emerged at various points of origin, including the aforementioned Gallup World Polls and EUIMAGINE, a survey of migration aspirations among people living in four different community contexts of Morocco, Senegal, Turkey, and Ukraine. Together, these surveys provide a useful global overview of migration aspiration patterns; they reveal that migration aspirations vary widely within (and not just between) sending country contexts, even after controlling for individuals’ socioeconomic status, and that migration aspirations exceed actual migration rates. Still, as Carling and Schewel (2018) point out, point-of-origin studies often fail to conceptualize and operationalize migration aspirations uniformly, which contributes to ongoing conceptual vagueness. At the same time, scholars largely agree that migration aspirations, like migration behavior, emerge from interactions between individual characteristics and the wider macro-level emigration environment (i.e., the broader social, economic, political, and legal context) (Carling and Collins 2018; Carling and Schewel 2018). Docquier, Peri, and Ruyssen’s (2014) cross-country analysis, for example, evaluates how various economic, policy, linguistic, cultural, and social network factors shape aggregate migration aspirations, as well as the transition of those aspirations into subsequent mobility.
Finally, research shows that migration aspirations can be reflections of individuals’ other “life” aspirations and identity projects, including who they want to be at a future point (Carling and Schewel 2018), which they often connect to images and ideas about other places, or where they want to be (Benson 2011, 2012; Timmerman, Hemmerechts, and de Clerck 2014; Benson and O’Reilly 2016). Prior migrants can be a key source of such imagery, complementing those that circulate through domestic institutions, material goods, and media (Glick Schiller and Salazar 2013; Collins 2018). Indeed, migrants remit not just money but also non-economic ideas about culture, gender, race, and politics to non-migrants within “transnational social fields” (Levitt 2001; Roth 2012; Joseph 2015). These ideas have been shown to exert normative influence over non-migrants’ imaginings of themselves in different places, which in turn shape new aspirations to move (Gardner 1993; Van Mol et al. 2018).
Migration Aspirations in the Global North
While scholarship on migration aspirations is expanding quickly, it focuses primarily on people in or from economically deprived world regions where migration is often envisioned as a pathway to a better and more economically secure future and for whom the experience of migration and adaptation often involves various adversities (Scheibelhofer 2018, 2). The Gallup World Polls are a notable exception (Docquier, Peri, and Ruyssen 2014), as is Scheibelhofer’s (2018) study of migration aspirations among Austrians in New York City. Members of the latter group, Scheibelhofer found, did not make one single “decision” to migrate but rather went through different stages in a larger “process” of arriving at a decision to leave and then settling on their destination. Further, aspirations for self-determination and self-realization were prominent among them: they saw New York City as a place of innovation in comparison to Austria and preferred the United States for that reason.
Although Scheibelhofer’s (2018) study is limited by the “mobility bias” Koikkalainen and Kyle (2016) identify, as it measures migration aspirations retrospectively, her findings still resonate with those in a separate literature on migration from the Global North, which focuses on flows of what are often dubbed “lifestyle migrants” moving from Northern to Southern Europe or from North to Central and South America (King, Warnes, and Williams 2000; O’Reilly 2000; Croucher 2009b, 2015, 2018; Benson and O’Reilly 2009; Oliver and O’Reilly 2010; Benson 2011; Benson and Osbaldiston 2016; Benson and O’Reilly 2016). Benson and O’Reilly (2009) define “lifestyle migrants” as “relatively affluent individuals, moving either part-time or full-time, permanently or temporarily, to places which, for various reasons, signify for the migrants something loosely defined as quality of life” (p. 621). Various studies in this tradition find that among lifestyle migrants, motivations for self-actualization (e.g., leaving something behind, starting anew, and achieving a new self or set of goals) are salient compared to the fulfilment of economic need. They also find that movement is often made possible by such migrants’ economic and noneconomic privilege, including high levels of cultural, financial, or social capital (Benson 2012; Croucher 2012, 2015, 2018), and that “lifestyle” migrants often construct and attach noneconomic criteria, like “paradise” or the “rural idyll,” to images they hold of specific geographic destinations (Benson 2011; Osbaldiston 2011; Benson 2012; Hayes 2015a; Viteri 2015; Kordel and Pohle 2018).
Nevertheless, Benson and Osbaldiston (2016) caution that as a label, lifestyle migration is often “adopted uncritically” and “without much thought to the theoretical implications it implies” (p. 409). Similarly, as far back as 2002, King cautioned against “false characterisations” of migrants as only “poor, uprooted, marginal and desperate” individuals coming from the Global South (p. 89) and noted that space in theorizations of migration should be left to include those migrants who were not marginal or desperate, as well as those from the Global North. 1 In other words, migrants from the Global North are not necessarily categorically distinct in motivation (i.e., noneconomic, consumptive) or material condition (i.e., affluent) from those from the Global South. Whether migrants originate in the Global South or the Global North, their motivations are complex and multi-causal (Castles 2010). This has led Benson and O’Reilly (2016) to argue more recently that lifestyle migration is best viewed not as a single migrant type but as an inductive “lens” for understanding how consumption-based and noneconomic motivations for migration intersect with economic ones among all migrants in the present era of globalization (see also Croucher 2015). Even so, investigations of how migrants from the Global North develop their initial aspirations to live abroad remain limited, especially by retrospective measurement and a lack of comparisons to non-migrant populations still at home. If we are to better model the first “aspiration” step of Carling and Schewel’s (2018) two-step model, we must move back to Global North migrants’ points of origin to correct this mobility bias.
Data and Expectations
In this article, we measure migration aspirations from the point of origin in one Global North sending country — the United States — prior to migration behavior potentially taking place. Our data come from an original internet survey we commissioned of 1,015 adults ages 18 and older living in the United States that was conducted by GfK Custom Research North America using its web-enabled KnowledgePanel® on July 11–13, 2014. 2 Subtracting all panelists who were first-generation immigrants, non–US citizens, or both, we arrived at a final sample of 877 US-born citizens for the present analysis.
Dependent Measures
We use four survey questions to calculate dependent variables for our multivariate analysis.
Aspiration
First, in line with the literature on migration aspirations, we developed aspiration as a compound variable that measures various aspects of aspiration but does not restrict aspiration to permanent moves, 3 as the 2007 to 2013 Gallup World Polls did 4 (see Table 1). We found that one-third of survey panelists (N = 288, 33.1%) revealed some aspiration to live abroad, including those who did not “realistically think it will happen” (N = 130, 14.9%), those who would consider it (N = 111, 12.8%), and those with stronger desires or plans (N = 47, 5.4%). On the other hand, a clear majority of survey panelists (N = 509, 58.4%) reported no aspiration to live abroad. In between, just under one-tenth (N = 74, 8.4%) reported no aspiration but indicated they would consider it if a special opportunity were to arise.
Americans’ Aspirations to Live Abroad.
a Six respondents were dropped because of missing data on aspiration.
Geographic preference of aspiration
Second, our survey asked panelists who revealed some aspiration to live abroad to note the country or world region in which they would be most interested in living (geographic preference). 5 As Table 2 shows, over half of panelists (52.1%) who answered this question revealed an aspiration to live abroad in Western Europe, Australia, or New Zealand. Behind them, a fifth of aspirants (19.5%) indicated Latin America or the Caribbean; together with those who indicated Mexico (3.8%), they comprise nearly one-quarter. Just under one-tenth of aspirants preferred Canada and Asia each. Given small sample size, we constructed dummy variables to see what factors predict these aspirants’ interest in living abroad in each of the following three geographic categories: Canada (versus elsewhere); Western Europe, Australia, or New Zealand (versus elsewhere); and Latin America or the Caribbean (including Mexico) (versus elsewhere). 6
Geographic and Temporal Dimensions of Americans’ Aspirations to Move Abroad (Only among Panelists Who Reveal Aspirations).
Timing preference of aspiration
Third, our survey asked panelists who revealed some aspiration to live abroad how long, ideally, they would like to do so (timing preference). 7 Table 2 shows that while approximately half of aspirants (49.8%) indicated that they would ideally like to live abroad for less than one year, almost one-third of aspirants (30.2%) envisioned somewhat longer-term moves between one and five years, and another fifth (19.8%) envisioned moving for longer than five years. Therefore, we constructed a dummy variable to see what factors predict these aspirants’ interest in living abroad for one year or longer (versus less than one year).
Motivations for aspiration
Finally, our survey offered panelists who revealed some aspiration to live abroad six choices of possible motivating factors (motivations) — (1) to work, (2) to study, (3) to join a partner, (4) to retire, (5) to explore, or (6) to leave a bad situation in the United States — and asked them to select up to three of the most likely, in order of strength. Based on the extant literature on American migration, we expected to uncover a range of motivating rationales (Schafran and Monkkonen 2011; Klekowski von Koppenfels 2014), including both “pull” factors such as exploration or job and study opportunities and “push” factors such as economic risk minimization or dissatisfaction with their personal or general situation at home (although see Klekowski von Koppenfels 2014; Vance and McNulty 2014; Morgan et al. 2016). 8
As Table 3 shows, to explore is by far the most frequently and highly ranked motivation among aspirants; well over three-quarters (N = 87.4%) ranked it as one of their top three motivations, and almost half (44.0%) ranked it as their leading motivation. To retire, to leave, and to work were the next most frequently ranked motivations, by approximately half of aspirants each (50.8%, 49.0%, and 48.3%, respectively), though we note that to retire was listed most commonly as a second motivation and to leave as a third motivation. To study (33.0%) and to join a partner (19.4%) were ranked least often, but still by roughly one-fourth to one-third of all aspirants. For ease, we condensed these responses into dummy variables for each motivation, coded 1 if a panelist selected a given motivation with a top 1 to 3 ranking and 0 otherwise. 9
Motivations behind Americans’ Aspirations to Move Abroad (Only among Panelists Who Reveal Aspirations).
Independent Measures
Controls
We control for age, gender, and US region of residence, all of which have been indicated in the literature as potentially shaping the development of aspirations to migrate (e.g., De Jong 2000; Kley 2011; Coulter 2013; Viteri 2015). We also control for race, political ideology, and strength of national identification as American. Not only is race the key “dispersive prism” through which Americans consider their identities, opportunities, and senses of belonging within the nation (Masuoka and Junn 2013), there is also some indication that American voters who feel a weakened sense of belonging to their nation (Hardwick 2010), especially during moments immediately following a lost presidential election (Alter 2012; Motyl 2014; Alter 2016), may vocalize greater aspiration to leave. Indeed, stronger national identity has been found to reduce migrants’ emigration intentions in other sending-country contexts like Mexico and Iceland (Theiss-Morse and Wals 2014). Because fully 86.0 percent of our sample reported a “very strong” American identity and another 12.0 percent a “somewhat strong” American identity — compared to just 2.0 percent who identified “not very strongly” as American or “not at all” — for parsimony we constructed a dummy variable for very strong American identification (versus anything less).
Cultural capital predictors
Social class, in both its economic and noneconomic dimensions, figures heavily in international migration flows from the Global North (Scott 2006; Benson and O’Reilly 2009, 2016; Croucher 2015; Miles 2015; Viteri 2015; Benson and O’Reilly 2016). Benson (2012) most explicitly calls for scholars to pay greater attention to the “cultural drivers” of migration aspiration and behavior, for which research on migration aspirations has found some support (Canache et al. 2013; Van Dalen and Henkens 2013; Docquier, Peri, and Ruyssen 2014). Here, our cultural capital predictors include a recoded ordinal measure capturing survey panelists’ highest level of education; three dummy variables capturing internet access in the household (internet access versus not), passport ownership (holding a US passport versus not), and foreign language (ability to speak at least one least one foreign language versus only English); and two ordinal variables for extent of panelists’ prior travel/tourism experience and prior living abroad experience (ranging from “I have never traveled/lived outside the United States” to “I have toured/lived abroad in more than 20 countries”). 10 Our reasoning is that in the United States, higher education, having access to the internet, having a US passport, the ability to speak at least one foreign language, and prior tourism and/or living abroad experience 11 are all arguably measures of Americans’ embodied cultural capital and class location (Bourdieu 1986), or even cosmopolitanism (Haubert and Fussell 2006), not just their human capital or “hard” professional skills. Furthermore, internet access may reflect stratified access to mass-mediated imagery that circulates to Americans through this medium, which scholarship has shown can encourage both cognitive imagining of migration (Salazar 2011; Koikkalainen and Kyle 2016) and actual migration (Hayes 2014).
Financial capital predictors
We also follow other scholars’ lead in hypothesizing that financial resources — such as income, employment status, or even dissatisfaction or anxiety with one’s personal finances or the broader “quality of life” in the public domain — can shape the “first stage” of migration aspirations (Canache et al. 2013; Van Dalen and Henkens 2013; Docquier, Peri, and Ruyssen 2014; Hayes 2014; Theiss-Morse and Wals 2014; Wals and Moreno n.d.). Here, our financial capital predictors include a continuous measure of annual pre-tax median household income; two dummy variables for currently employed status and homeownership; and two ordinal variables to measure panelists’ subjective ratings of the health of their own personal finances and the health of the U.S. economy, our ways of gauging their personal and sociotropic assessments of economic well-being. We think it equally plausible that Americans’ aspirations could be classed by differential access to financial resources (Appadurai 2004) or, alternately, that such aspirations could be more “open” and less dependent on financial capital compared to later stages of decision and action.
Social capital predictors
A central form of capital (Bourdieu 1986), social capital can be loosely defined as resources that derive from individuals’ relationships with others and that can be converted into value. Docquier, Peri, and Ruyssen’s (2014) study shows that social networks with prior migrants are the “key” factor encouraging the development of migration aspirations worldwide, 12 while the larger literatures on migration and transnationalism show that prior migrants are a key source of imagery about different destinations (Collins 2018). Here, our social capital predictors include a dummy variable measuring survey panelists’ immigrant heritage; two dummy variables measuring panelists’ social networks beyond the household: military service and panelists’ self-reports of direct social networks with other US citizens who have lived or are living abroad; and two dummy variables for household structure: living with partner and the presence of children under 18 in the household. Our reasoning is that Americans’ social ties to US citizens abroad can emerge from a number of sources, including their families; their broader circles of friends, coworkers, and acquaintances; 13 and even formal institutions such as the military, which commonly deploys Americans overseas. We also follow extant research showing that household structure shapes peoples’ norms and obligations, especially by age, gender, and class.
Bivariate Correlates of Americans’ Migration Aspirations
Tables 4 and 5 present bivariate associations between our independent and dependent variables. We focus on the main patterns to save space for our multivariate results, but cross-tabulations for all significant results are located in a Supplemental Appendix (available in the online version of this article) for interested readers.
Bivariate Relations between Independent and Dependent Variables.
a Full sample.
b Only among panelists who reveal aspirations.
Chi-square (χ2) values, *p < 0.05. **p < 0.01. ***p < 0.001.
Bivariate Relations between Independent and Dependent Variables, Continued (Only among Panelists Who Reveal Aspirations).
Chi-square (χ2) values, *p < 0.05. **p < 0.01. ***p < 0.001.
Aspiration
A number of variables are significantly associated with survey panelists’ aspiration to live abroad (see Table 4, Column 1). Among the controls, higher proportions of panelists who are younger, reside in the US Northeast, are politically liberal, and report less than a very strong American identity also indicate stronger aspiration to live abroad. Beyond the controls, all six cultural capital variables, three of the five financial capital variables, and three of the five social capital variables are significantly associated with panelists’ aspiration to live abroad (see Table 4, Column 1). Cultural capital appears to be especially strongly correlated with Americans’ aspirations to live abroad; higher proportions of panelists who are college-educated, have internet access at home, speak at least one language in addition to English, hold a US passport, and have prior tourism experience report higher levels of aspiration. In terms of financial capital, higher mean household income and being employed full- or part-time are positively associated, yet homeownership negatively associated, with migration aspirations. Having social networks with other US citizens who have lived abroad and having recent immigrant heritage (defined as having second- to fourth-generation immigrant heritage) are the two key social capital variables associated with stronger aspiration.
Most of these bivariate associations and significance levels remain unchanged when we run a robustness check dropping aspiration’s Response Categories 2 and 3 (see Table 4, Column 2). We do this because we recognize that these two response categories may capture competing yes/no aspects of aspiration within our compound variable, so the results give us greater confidence in our findings.
Geographic and temporal dimensions of aspiration
Honing in on the survey panelists who revealed some aspiration to live abroad, Table 4 (see Columns 3, 4, and 5) shows two main patterns: first, fewer variables are significantly associated with Americans’ ideal destinations and length of residence abroad than was the case with aspiration itself; and second, no cultural, economic, or social capital variables are consistently significantly associated across all models. For instance, aspirations to live abroad in Canada (see Table 4, Column 3) are significantly associated only with having a weaker national identification. By contrast, aspirations to live abroad in Western Europe, Australia, and New Zealand (see Table 4, Column 4) are significantly associated only with whiteness and having networks with other US citizens who have lived abroad. Still differently, aspirations to live abroad in Mexico, Latin America, or the Caribbean (see Table 4, Column 5) are significantly associated with a wider range of variables: being older and less educated, having internet access in the household, homeownership, currently living with a partner, and not having networks with other US citizens who have lived abroad. Finally, aspirations to live abroad for more than one year (see Table 4, Column 6) are significantly associated with yet a different constellation of variables: being in midlife, rating the US economy’s health as better, currently living with a partner, and service in the US military. 14
Motivations for aspiration
Turning to the different motivations for aspirations to live abroad — working, studying, joining a partner, retiring, exploring, or leaving a bad situation in the United States — once again bivariate analysis (see Table 5) shows that fewer variables are significantly associated with any of these underlying motivations than was the case with aspiration. For example, the motivation to work is significantly associated (see Table 5, Column 1) with being younger, male, and a resident of the US Midwest and, among the types of capital, having internet access in the household, not being a homeowner, and not having served in the military. Aspiration to live abroad to study is significantly associated (see Table 5, Column 2) with being either younger (ages 18–25) or older (over age 55) and having a very strong national identification and also with not being employed, giving a positive rating to the US economy’s health, and not currently living with a partner. Still differently, the motivation to join a partner (see Table 5, Column 3) is significantly associated with being female, nonwhite, and a resident of the US West, plus not being a homeowner.
In comparison, social capital appears more relevant to the motivation to retire (see Table 5, Column 4), whereas cultural capital appears more relevant to the motivation to leave (see Table 5, Column 5). To retire is significantly associated with being older and having a very strong national identification but also with having prior international tourism experience, being a homeowner, currently living with a partner, not having children under the age of 18 living at home, and military service. To leave is significantly associated with weaker American national identification and also with having only a high school degree, not holding a US passport, not having prior tourism experience, and giving a lower rating to the US economy’s health. Finally, the motivation to explore (see Table 5, Column 6) is significantly associated only with whiteness, stronger national identification, and having less prior experience living abroad. Fully 92.4 percent of white aspirants — especially those who identify as “somewhat” (94.4%) and “very” strongly (86.6%) American — rank to explore as a motivation, compared to just 76.6 percent of nonwhite aspirants in our sample.
Predictors of Americans’ Migration Aspirations
Having established the bivariate relevance of many of our independent variables, we now build several multivariate models to examine whether — and if so, how — they effectively predict Americans’ aspirations to migrate.
Aspiration
First, we conducted multinomial logistic regression to predict the presence of aspiration. This method is used to model nominal outcome variables and thus offers an opportunity to examine which independent variables predict each of aspiration’s three clean “yes” response categories (numbers 4–6) relative to its clean “no” response category (number 1), without sacrificing sample size. Since none of the independent variables significantly predict Response Category 6, perhaps because of its small sample size, we focus our discussion here only on the variables that predict Response Categories 4 (“Yes, I have [considered living abroad] and if a special opportunity arose to do so I would consider it”) and 5 (“Yes, I have [considered living abroad] and I want to find a way to make it happen someday”). In the text, we transform the coefficients that appear in Table 6 into odds ratios for ease of discussion.
Predictors of Americans’ Aspirations to Live Abroad.
Note. The reference category is 1: “No, I have not and realistically do not expect to do so.”
a B-values from multinomial logistical regression models, *p < 0.05. **p < 0.01. ***p < 0.001.
Compared to the bivariate results presented earlier, far fewer independent variables predict aspiration in the multivariate models. Political ideology, of particular note, is not a significant predictor of aspiration, though having less than a very strong American identification does raise aspiration in both Response Categories 4 and 5, as does residing outside the US West in Response Category 5. Among the other significant results, cultural capital appears to be strongest, followed by social capital. 15 Surprisingly, however, it is not education level, knowledge of foreign languages, or household structure that shapes Americans’ aspiration. Rather, aspiration to migrate is greater among panelists who hold a US passport; have internet access in the household, prior living abroad experience, and recent immigrant heritage; have served in the military; and have networks with other US citizens who have lived abroad. 16
Geographic and temporal dimensions of aspiration
Second, we conducted binary logistical regression to predict whether aspirants rank (1) Canada; (2) Western Europe, Australia, or New Zealand; or (3) Latin America or the Caribbean (including Mexico) as their preferred destination and whether they aspire to live abroad for longer than one year (see Table 7).
Predictors of Geographic and Temporal Dimensions of Americans’ Aspirations to Live Abroad (Only among Panelists Who Reveal Aspirations).
a B-balues from binary logistical regression models, *p < 0.05. **p < 0.01. ***p < 0.001.
b Parameter estimates from ordered logit regression models, *p < 0.05. **p < 0.01. ***p < 0.001.
Overall, cultural capital appears to be the strongest predictor of aspirants’ geographic preferences, followed by financial capital. Perhaps surprisingly, knowledge of foreign languages and internet access in the household do not determine where panelists aspire to go. Still, aspirants are 2.6 times more likely to aspire to live in Latin America as their education level decreases, 3.0 times more likely to prefer Canada (though 2.4 times less likely to prefer Western Europe, Australia, or New Zealand) if they hold US passports, and 0.7 times less likely to prefer Canada if they have greater prior international tourism experience. In terms of financial capital, aspirants are 1.8 times less likely to prefer Canada as their evaluation of the health of their own personal finances improves, 2.6 times more likely to prefer Latin America if they are homeowners, and 1.6 times more likely to want to live abroad for longer than one year if they are employed. Only one social capital variable is significant: aspirants are 2.1 times more likely to aspire to live in Latin America, and 1.7 times more likely to aspire to live abroad for longer than one year, if they are currently living with a partner.
Interestingly, three controls — gender, race, and strength of national identification — also shape panelists’ geographic “imagining” of where they wish to go. Aspirants are 2.8 times more likely to prefer Canada if they are male. If they are non-Hispanic white, they are 1.8 times more likely to prefer Western Europe, Australia, or New Zealand but 2.39 times less likely to prefer Latin America. Finally, as national identification intensifies, aspirants are 1.5 times more likely to prefer Western Europe, Australia, or New Zealand but 2.9 times less likely to prefer Canada. In sum, whereas we saw earlier that certain elements of cultural and social capital, including networks with prior migrants, best predicted the presence versus absence of aspiration among Americans in our sample, in tandem with weaker American national identification and non-West US region residence, here we see that the geographic dimensions of such aspirations are best predicted by a combination of both the cultural and economic aspects of social class, alongside race and national attachment.
Motivations for aspiration
Third, we conducted binary logistical regression to predict whether aspirants rank (1) to work, (2) to study; (3) to join a partner; (4) to retire; (5) to explore; or (6) to leave among their top three motivations (see Table 8).
Predictors of Motivations behind Americans’ Aspirations to Live Abroad (Only among Panelists Who Reveal Aspirations).
B-values from binary logistical regression models, *p < 0.05. **p < 0.01. ***p < 0.001.
We find that the controls — specifically, age, race, US region of residence, and strength of national identification — help predict several motivations, but in different constellations. For example, as age increases, aspirants are 1.4 times less likely to rank to work but 1.4 times more likely to rank to retire as a motivation. Aspirants are 2.0 times more likely to rank to join a partner if they are female. If they are nonwhite, aspirants are 2.7 times more likely to rank to join a partner and 2.1 times more likely to rank to retire, but 3.1 times less likely to rank to explore. If they live in the US Northeast or West, aspirants are 2.9 times less likely to rank to work as a top motivation; those in the Northeast are also 2.2 times more likely to rank to leave. Finally, as national identification strengthens, aspirants are 3.0 times more likely to rank to study, but 4.0 times less likely to rank to leave.
Beyond these controls, three cultural capital and three financial capital variables predict just one motivation each, while one social capital variable (currently living with partner) predicts two. None of the forms of capital help predict whether aspirants rank to work or to explore. Aspirants are 2.4 and 1.8 times more likely to list to study as a top motivation if they are unemployed or not currently living with a partner. As their education declines, if they lack access to internet at home or are not homeowners, aspirants are 2.2, 3.0, and 2.4 times more likely, respectively, to list to join a partner. Aspirants are 2.6 times more likely to list to retire if they are currently living with a partner. Finally, aspirants are 1.9 and 1.7 times more likely to rank to leave as a top motivation if they do not hold US passports and if they give a better (as opposed to worse) rating to the US economy’s health, respectively.
Discussion
This article has measured Americans’ migration aspirations from the point of origin and prior to potential migration behavior, using a nationally representative online sample we commissioned and fielded in summer 2014. We asked, at that point in time, how many Americans had some aspiration to live abroad and, if so, where did they imagine going, for how long, and why. In our analysis, we investigate what forms of cultural, financial, and social capital “embed and inform” (Bal and Willems 2014) these considerations.
Overall, we found that one-third (33.1%) of panelists in our sample expressed some aspiration to live abroad (see Table 1). Likely because our measure does not restrict aspiration to long-term or permanent moves, this proportion is higher than the 10 percent of Americans who indicated a wish to move abroad “permanently, if they had the opportunity to do so,” in the 2007 to 2013 Gallup World Surveys (OECD 2015). We also found that over half of aspirants (52.1%) looked primarily toward Western Europe, Australia or New Zealand; while another quarter (23.3%) were eyeing Latin America, the Caribbean, or Mexico (Table 2). Aspirants were evenly divided between those who envisioned going abroad for less than one year and those who envisioned going abroad for longer; in fact, nearly 15 percent envisioned going abroad for more than 10 years or indefinitely (see Table 2).
Finally, we found that exploration was the leading motivation behind aspirants’ desires to live abroad; fully 87 percent of all aspirants ranked it as one of their top three motivations, and almost half (44%) ranked it as their first. Nevertheless, aspirants’ overall motivations for living abroad varied: approximately half ranked to retire, to leave, or to work among their top three motivations; and even though they were less prominent, roughly one-fourth to one-third of aspirants also ranked to study or to join a partner (see Table 3). On one hand, we interpret these results as fitting nicely with the extant literature on so-called “lifestyle migration,” which shows that Global North migrants see migration as a way to pursue new opportunities for exploration, self-actualization, and fulfillment (e.g., Benson and O’Reilly 2016; Scheibelhofer 2018). On the other hand, we think these results support the effort already under way in that literature to problematize what can sometimes come across as an overly facile depiction of Global North migrants as a singular, homogenous migrant type composed of uniformly “advantaged” and “privileged” people (Croucher 2012; Benson 2014). To be sure, many of our American aspirants, if they were to eventually move abroad, would be “relatively advantaged” (by class, race, and national origin) compared to their host populations, particularly in Latin America, Asia, and Africa (see Croucher 2007, 2009a, 2018; Benson 2013; Spalding 2013; Hayes 2015b; Kordel and Pohle 2018). Still, research shows that American migrants abroad are heterogenous in terms of class status and motivation (Schafran and Monkkonen 2011; Hayes 2014; Klekowski von Koppenfels 2014); our research shows this is also the case among Americans who are considering migration prospectively.
In other words, while exploration is certainly prominent among our aspirants, especially those who are non-Hispanic whites and those who are strong national identifiers, and while some of the other motivations we offered to panelists may also capture elements of an expressive search for self-fulfillment or improved “way of life,” aspirants also exhibit a range of economic and noneconomic motivations for wanting to migrate that belie singular categorization. Indeed, we think our results suggest utility in future research working to better conceptualize exploration as a migration motivation, not only in the literature on flows from the Global North but even more broadly. We know, for instance, that small proportions of US-bound Mexican migrants report wanting to migrate in search of “adventure” (a la aventura) (Hondagneu-Sotelo 1994; Hernández-León 1999; R. Smith 2005; Hagan, Demonsant, and Chávez 2014, 83) or for greater sexual freedom and autonomy (Carrillo 2018). Such motivations, however, remain understudied and undertheorized among a group and stream typically typed as economic “labor” migrants, despite the fact that economic desires for higher wages or risk minimization are often intertwined with a noneconomic desire for an improved “quality of life.”
Setting our panelists’ aspirations and their characteristics within a structure-agency framework, our multivariate results indicate that the presence of migration aspirations among Americans is best predicted by a combination of cultural and social capital, in tandem with weaker American national identification and non-West US regional residence. Our panelists’ aspirations increase when they live outside the US West and express less than a “very strong” national identity; but panelists’ aspirations also increase when they hold US passports, have internet access in the household, have prior living abroad experience, have recent immigrant heritage, have served in the military, and have social networks with other US citizens who have lived abroad. We interpret these findings as lending support to research conducted in other sending contexts that shows a negative impact of national attachment (Theiss-Morse and Wals 2014) but a positive impact of social networks and institutions (such as the military or mass media) (Massey 1999) on the development of migration aspirations. In our data, some Americans do appear to develop “imagery” about a life abroad from prior migrants, having internet access, and serving in the US military.
Further, we interpret these findings as supporting the argument that cultural, not merely economic, elements of social class shape how individuals in the Global North envision their potential lives abroad (A. Smith 2006; Benson 2012). Among panelists in our sample, migration aspirations are certainly “classed” (Appadurai 2004) but by the forms of cultural capital listed above, not by access to financial resources, at least not directly. Of course, financial capital may still come into play more strongly in later stages of decision and action, helping determine which panelists’ migration aspirations get translated into actual behavior in the future, or not (Oliver and O’Reilly 2010; Docquier, Peri, and Ruyssen 2014; Benson and O’Reilly 2016).
Finally, our analysis of the factors predicting the geographic and temporal dimensions of, and motivations underlying, Americans’ migration aspirations tell a less consistent story overall. However, among them we see two noteworthy patterns. First, national attachment and cultural capital continue to be salient; here they shape the “culturally significant imaginings” (Benson 2012) of where Americans wish to go, and they are joined by a few other economic aspects of class and race. Whites, stronger national identifiers, and aspirants who do not hold US passports are more likely to look toward Western Europe, Australia, or New Zealand. Perhaps this reflects dominant views held by many white Americans that Western Europe, Australia, and New Zealand are “civilized” and safe “white” spaces socioeconomically closest in position to the US versus dominant views of Latin America, Asia, Africa, the Middle East, and even Eastern Europe and the (former) Soviet Union as socioeconomically, racially, and religiously inferior or distant. 17 In contrast, aspirants who hold passports and rate the health of their own personal finances as better but are weaker national identifiers and have less international tourism experience are more likely to look toward Canada. Perhaps this echoes long-standing myths and realities about American migrants in Canada, some of whom, despite internally complex “circumstances and motivation” (Dashefsky et al. 1992, 32), went there during the 1960s and 1970s as “draft dodgers, deserters, or political activists who opposed the [Vietnam] war” (Hardwick 2010, 90). Finally, non-whites, less educated aspirants, and homeowners are more likely to look toward Latin America. This could fit well with the extant literature on American retirees and second home owners throughout Latin America (e.g., Ortero 1997; Migration Policy Institute [MPI] 2006; Janoschka 2009; Lizzáraga Morales 2010; Miles 2015). However, the fact that the large majority of American migrants in Latin America today are white (Benson 2013; Hayes 2015b; Croucher 2018) suggests a critical disjuncture in how race shapes Americans’ pre-decisional “mental time travel” (Koikkalainen and Kyle 2016) versus how it may shape their actual migration behavior, which future research could explore further.
Second, we think it noteworthy that race and national attachment also shape Americans’ motivations for living abroad. In our multivariate results, non-Hispanic whites are more likely to rank to explore — the top motivation listed by all aspirants — but less likely to rank to retire or to join a partner, while weaker national identifiers are more likely to rank to leave but less likely to rank to study. While no forms of capital we tested predict motivations to work or to explore, we do find that American aspirants who are motivated to join a partner appear somewhat less advantaged overall (they are less educated, have no access to internet, and are not homeowners), while those who are motivated to leave a bad situation in the United States do not (they do not hold US passports but give a better rating to the US economy’s health). Combined with the finding that aspirants who rank the US economy’s health as better are more likely to want to live abroad for longer, perhaps these results suggest it is the more economically secure (as opposed to insecure) Americans who express the greatest “push” to leave. This would not necessarily invalidate current research documenting narratives of financial insecurity and risk among Americans already abroad (Hayes 2014, 2015b; Hayes and Pérez-Gañán 2017), but it could help put such narratives into broader context by distinguishing between the ways economic outlook shapes Americans’ pre-decisional considerations versus their subsequent decisions to move.
While our study offers these contributions in an effort to enrich the literature on migration aspirations in a Global North context, we do acknowledge its limitations. Most importantly, our measure of aspiration is imperfect, since it does not allow us to examine the full continuum of aspiration or the full range of survey panelists’ desires, intentions, likelihoods, strategies, or plans (Carling and Collins 2018; Carling and Schewel 2018). Carling and colleagues call for surveys that employ multiple questions to better operationalize and capture these dimensions. Further, like Benson and O’Reilly (2016), they suggest that qualitative methods may be even better suited to studying migration aspirations, as they focus centrally on micro-level process-tracing and uncovering meaning, whereas survey methods are more appropriate for identifying broad patterns and establishing causal relationships. Identifying broad patterns and variation among potential migrants within a representative sample was our main focus here, but qualitative methods would certainly help flesh out why panelists did or did not reveal migration aspirations, why they were interested in some places over others, and how they made sense of their underlying motivations.
Further, while we include some measures of Americans’ cultural, financial, and social resources and agree with Oliver and O’Reilly (2010) that these wield the power to shape what humans consider achievable and worth aspiring to in different settings, our list of predictors is by no means exhaustive. Docquier, Peri, and Ruyysen (2014) find little evidence for a range of cultural, institutional, and policy controls on the development of migration aspirations worldwide in the late 2000s; but we agree with existing studies on both migration aspirations and Global North migration that there is utility in future quantitative studies operationalizing not only wider cultural norms and repertoires about migration, perhaps as cultural or social capital (Benson 2012; Timmerman, Hemmerechts, and de Clerck 2014; Carling and Collins 2018) in their models, but also the various kinds of “migration industry” organizations and institutions that grease the wheels of Global North migration, perhaps as social or institutional capital variables (Prado 2012; Hayes 2014; Mescoli 2014; Hayes and Pérez-Gañán 2017). In addition, while political ideology was not significantly associated with migration aspiration or its characteristics in our multivariate analysis — at least not during the midterm election year of 2014 — how political variables shape Americans’ aspirations likely varies over time and could plausibly shift under the contested nature of the current Trump presidency (Kaysen 2018). 18 Nor does our survey include a number of other variables that could shape aspiration, such as personality, environmental disparities, or incentive structure at the place of destination, which could include the range of economic and political incentives that many foreign countries currently offer to recruit Global North migrants (e.g., MPI 2006; Canache et al. 2013; Van Dalen and Henkens 2013).
More broadly, future research could investigate the geographic and temporal dimensions of Americans’ migration aspirations at a finer-grained level, query for a wider variety of motivating factors behind why some individuals wish to leave versus stay, begin tracking aspirations and the aspiration-migration nexus over time (using repeated cross-sectional or longitudinal research designs that allow for better specification of causality than we can provide), and engage in cross-national comparisons with other flows originating in both the Global North and Global South. How economic and noneconomic motivations intersect among various groups’ migration aspiration could be one fruitful area of comparative research (Hayes 2014; Hayes and Pérez-Gañán 2017); how different groups perceive their aspirations on a continuum of (in)voluntariness could be another (Bivand Erdal and Oeppen 2018). As Benson and Osbaldiston (2016) and Hayes (2014) argue, such efforts will help uncover how both material structure and historical time work to shape Americans’ prospective imaginations of themselves, either at home or abroad.
Supplemental Material
Supplemental Material, MRX806852_Supplemental_Appendix - Modeling American Migration Aspirations: How Capital, Race, and National Identity Shape Americans’ Ideas about Living Abroad
Supplemental Material, MRX806852_Supplemental_Appendix for Modeling American Migration Aspirations: How Capital, Race, and National Identity Shape Americans’ Ideas about Living Abroad by Helen B. Marrow and Amanda Klekowski von Koppenfels in International Migration Review
Footnotes
Acknowledgments
We thank Bart Bonikowski, Katharine Donato, Dan Hopkins, Saara Koikkalainen, Natasha Kumar Warikoo, Kerilyn Schewel, Debbie Schildkraut, Cinzia Solari, the Migration and Immigrant Integration Workshop at Harvard University, the IMISCOE network, and the editor and three anonymous reviewers at IMR for helpful feedback.
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 author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: We gratefully acknowledge several funding sources for our survey: the Faculty Research Fund from the Faculty of Social Sciences at University of Kent, the Research Support Fund of the School of Politics and International Relations at the University of Kent, and the Summer Faculty Fellowship from the Faculty Research Awards Committee at Tufts University.
Supplemental Material
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
