Re-conceptualizing the economic incorporation of immigrants: A comparison of the Mexican and Vietnamese

Abstract

Using data from the 2000 5 per cent Integrated Public Use Microdata Series, this article advocates three shifts in our theoretical and empirical approaches to understanding immigrant economic incorporation. First, through a comparison of Mexican and Vietnamese immigrants, these findings highlight the importance of an immigrant population's relationship to the state for economic outcomes, and cautions against analyses that aggregate the foreign-born population. Second, through a joint analysis of unemployment and poverty outcomes, these findings call for researchers to be specific about the varied aspects of “economic incorporation” and distinguish between factors that drive labor market access, and those that foster material well-being. Lastly, by examining three state economic, demographic and policy variables, this article promotes an approach that takes human capital into account, while also heeding the immigrant context of reception.

The Persistent Significance of Assimilation

Scholars are consistently concerned about what promotes immigrant assimilation, in particular, immigrant economic incorporation. Many assimilation trajectories have been proposed and debated ranging from Gordon's (1964) “straight line assimilation” model, to Portes and Zhou's (1993) revisionist “segmented assimilation” model. Alba and Nee (2003), who reaffirm the importance of assimilation for contemporary immigration streams, define assimilation as the “social distance separating immigrants and their children from the mainstream of American society.” In the current multicultural climate, there have been calls to discard the debate on assimilation, yet the discussion continues. This persistent focus is motivated by a variety of concerns, including the effect immigrants have on native workers, the fiscal strain immigrants may pose to local economies, and the well-being of immigrants and their families.

This article assesses a key aspect of the immigrant assimilation process, economic incorporation, and revisits the standard models proposed for immigrant incorporation.¹ While immigrant assimilation is multi-faceted, the various ways in which immigrants navigate the US labor market are fundamental indicators of overall well-being. Generally, most empirical assessments of economic incorporation focus on the impact of human capital on individual labor market outcomes, with less attention paid to the importance of immigrant mode of entry, family resources, and the context of reception in the host society. I argue that conventional studies of economic incorporation have been too narrow, and identify three specific limitations. First, existing scholarship tends to treat the immigrant population monolithically (for example, Kossoudji, 1989; Bratsberg et al, 2002), or analyze particular groups individually (for example, Aguilera and Massey, 2003), rather than comparatively recognizing the heterogeneity of the foreign-born population and the distinct processes of economic incorporation that different national origin groups may experience. Second, many analyses assess economic incorporation solely on the basis of individual labor market outcomes (for example, Borjas and Tienda, 1993; Portes and Zhou, 1996; Bean and Stevens, 2003), rather than also assessing the key determinants for family well-being. Third, I argue that past studies disproportionately attend to individual human capital determinants (for example, Chiswick et al, 1997; Trejo, 1997; Borjas, 2001), while insufficiently attending to the institutional and structural factors which may also influence economic outcomes.

This article broadens our understanding of economic incorporation through a statistical analysis that takes these additional factors into account. By taking this expanded approach, I highlight the importance of refining our economic incorporation models, based on the particular national origin group in question. I also find that the key factors driving individual entry into the labor market differ from those that promote family well-being for immigrants in the labor force. Lastly, I argue that individual human capital variables, though essential, do not alone explain immigrant economic outcomes.

Expanding Our Analytical Frame

Economic incorporation, while typically measured by one or two indicators, perhaps more accurately represents a multi-faceted process that includes various aspects of the work experience. While employment represents a positive success for immigrants, certainly not all jobs are created equal, and there is significant occupational stratification and earnings disparity in the foreign-born (as well as native-born) population. However, while success in the labor market drives financial security, non-labor market resources (such as family assets and public transfers) also shape an immigrant's material well-being, measured most extremely by their risk of poverty. Figure 1 outlines the various elements that shape this economic incorporation process, including individual human capital, the host society's context of reception, and an immigrant's mode of entry (or documentation status.)

Figure 1

Elements of the economic incorporation process.

One of the most discussed elements is human capital, which Borjas (2001) defines as “a person's endowment of ability and acquired skills.” This is typically measured by years of schooling and/or educational credential, language proficiency and individual labor market experience. Existing evidence shows that human capital does not necessarily have the same returns for all workers, and not all human capital is created equal. Foreign-born workers pay an “immigrant tax.” Their schooling and added labor market experience, especially if it is acquired abroad, brings immigrants workers smaller returns compared to native-born workers (Chiswick et al, 1997; Aydemir and Skuterud, 2005). This is due, in part, to the difficulty of transferring a foreign credential; the story of the highly educated immigrant who was a professional back home, but now drives taxis or cleans buildings in the United States, is not an uncommon one (Mattoo et al, 2008). The effect of human capital also differs for particular national-origin groups, and presents a barrier especially for immigrants who lack documentation status (Duleep and Regets, 1996; Kossoudji and Cobb-Clark, 1996; Cortes, 2004). While some research has highlighted the “negative socialization” of an American educational system for racial and ethnic minorities (Ogbu, 2008), American schooling also imparts key skills and norms that may aid immigrants in the economic incorporation process (Ainsworth-Darnell and Downey, 1998; Glick and White, 2003).

However, despite the complexity of the transmission of human capital to economic outcomes, the context of reception in the immigrant's host country is also a key factor. Past comparative research has highlighted the importance of national policies that promote incorporation (Reitz, 1998; Bloemraad, 2003), and state and local policies are increasingly relevant as well (Murthy, 2007). Other contextual factors, such as employer preferences (Kirschenman and Neckerman, 1991; Waldinger and Lichter, 2003), local economic conditions (Raijman and Tienda, 1999), and the history of immigration in a particular place (Portes and Stepick, 1985; Sanders and Nee, 1987) have also been identified as significant determinants.

Though an immigrant's mode of entry is known to mediate eligibility for employment, services and benefits (Van Hook and Balistreri, 2002), as well as shape economic mobility processes (Tienda and Singer, 1995; Phillips and Massey, 1999), it is a factor that has been, nevertheless, largely omitted from research on economic incorporation. This is understandably due, in large part, to data limitations; most surveys simply do not query mode of entry or documentation status. Citizenship status is more often collected, and consequently most research focuses only on the effects of naturalization (for example, Chiswick, 1978; Bratsberg et al, 2002; Kwon et al, 2004). Yet, very few US jobs specify citizenship status as an explicit eligibility requirement.² Conversely, immigrants who lack a work permit (obtained either through refugee status, legal permanent residence or a temporary visa), are not authorized to work in the United States at all. Estimates find that close to a third of immigrants fall into this unauthorized category, constituting 5 per cent of the civilian work force (Passel, 2006).

Existing economic incorporation scholarship also tends to focus solely on the individual labor market access, with less attention paid to material well-being. While key individual outcomes such as employment, earnings and occupation depend on successful entry into the labor market, poverty risk is contingent on family context and also depends on access to non-market resources.³ This access is mediated by eligibility, which can be based in large part by an immigrant's documentation status. Unauthorized economic migrants are barred from legally working the United States, and are largely ineligible for services that may aid in immigrant incorporation. Conversely, political migrants (that is, refugees) receive resettlement assistance and resources that may aid in their economic incorporation process. Though the refugee experience is not uniform, refugees receive work authorization upon arrival, and the Office of Refugee Resettlement (a department within the US Department of Health and Human Services) coordinates programs for refugees such as vocational training and English instruction largely through grants to non-profit organizations who work in these communities (Newland et al, 2007).

Although past economic research has compared outcomes for economic immigrants and refugees, most studies highlight the individual characteristics of each group. For example, Cortes (2004) focuses on the effect of the “implicit time horizon” each group has, highlighting the fact that refugees lack the option to return to their country of origin and may in turn be more inclined than economic migrants to assimilate. However, sociological research emphasizes that the primary difference between these two groups is not necessarily these attitudinal perspectives, or other demographic and human capital factors, but rather their relationship to the states. Hein finds that refugee status constitutes a “relationship to the state that takes a number of forms during the process of uprooting, migration, and adaptation,” and particularly shapes access to the social welfare system (Hein, 1993, 55).

In sum, the effects of human capital on economic incorporation, while critical, are far from uniform. Its effects are mediated and complemented by the context of reception and an immigrants’ mode of entry to the United States. A more accurate understanding of economic incorporation requires us to disaggregate the various elements of this process and explicitly include these institutional and structural factors in our theoretical and empirical models of economic incorporation. This article advances our understanding of immigrant economic incorporation in the following ways. First, by comparing results for Mexican immigrants (the largest national origin group of economic migrants in the United States) and Vietnamese migrants (the largest national origin group of refugee migrants in the United States), I assess how an immigrant's relationship to the state mediates the effects of human capital. Second, I overtly shift the traditional discussion of economic incorporation away from just focusing solely on labor market success, by comparing what matters for employment,⁴ versus the poverty risk.⁵ Lastly, I move beyond the focus on individual human capital, to assess the role that context of reception plays for a range of economic outcomes.

Methods and Data

This article relies on a statistical analysis of aggregate census data. To be sure, statistical analyses can be problematic, particularly for studying vulnerable populations such as racial/ethnic minorities and immigrants (Bonilla-Silva and Baiocchi, 2001). However, I argue that a quantitative approach provides a useful tool for comparing outcomes for large population groups (such as economic migrants and refugees), while controlling for important competing determinants (such as individual human capital characteristics versus context of reception factors), in order to adjudicate between competing theories. The data for this analysis are drawn from the 2000 Census 5 per cent Integrated Public Use Microdata Series.⁶ The universe is the non-institutionalized foreign-born population aged 25–64 years.⁷

I control for standard demographic variables such as gender, marital status and citizenship status. In addition, I control for key human capital variables, including labor market experience, years of residence in the United States, self-reported English proficiency and educational attainment. To measure the effect of US socialization on economic incorporation, I also control for whether the respondent migrated as a school-aged child (age 10 years or younger). In addition to these individual factors, I assess the effects of three state context of reception factors: economic climate and job competition (measured by the unemployment rate), history of immigration (measured by the percent of the state population that is foreign-born) and immigrant policy context (measured by the presence of an English-Only law.)

I address four different economic outcomes: poverty status, employment status, earnings and occupation. Following the federal standard, individuals are classified as living in poverty if their family income falls below 100 per cent of the poverty threshold.⁸ An unemployed person is defined as someone who does not currently have a job, is looking for a job, and has not yet found one. All working age individuals who were in the labor force, and either had a job or were in the armed forces, were coded as being employed. I estimate these dichotomous outcomes independently with a standard logistic model, and jointly with a bivariate probit model. Earnings are measured based on an individual's aggregate wage and business income from the previous year, and is limited to non-zero earners.^9,10 Using the approach advocated by Petersen (2002), earnings are estimated with a general linear model (GLM), with a logarithmic link function and gamma distributed error term.¹¹ The coefficients of a GLM regression model can be correctly interpreted as the natural logarithm of the factor by which the predicted mean earnings differs from each unit change in the dependent variable. This earnings analysis is limited to those employed individuals who earn at least $1, and all relevant discussions of earnings rely on the estimates from this non-zero earner sample. To assess occupational outcomes, I followed a similar approach to Hout et al (1995), and group the approximately 500 census groups into six main occupational categories: Professional, White Collar, Upper Blue Collar, Lower Blue Collar, Service and Agriculture. Analyses of occupational outcomes employ a multinomial logit model. The base category for all estimates presented is agriculture. (For a detailed summary of all the variables used in this analysis, refer to Appendix A.)

Findings I: Differential Benefits of Human Capital for Economic Migrants and Refugees

Table 1 presents a general profile of the foreign-born population in the United States. On average, 11 per cent of all immigrants arrived when they were of 10 years or younger, 42 per cent have naturalized, and less than 10 per cent claim not to speak any English. Roughly a third of these immigrants have less than a high school education, whereas a quarter have completed at least a Bachelor's degree. Yet, to be sure, there is significant heterogeneity within the immigrant population. Table 2 provides a comparison of the two populations that this article focuses on. On average, Mexican immigrants migrate with less human capital than do their Vietnamese counterparts. A greater percentage of Mexican immigrants lack English proficiency (21 per cent), than do Vietnamese (5 per cent). Similarly, 69 per cent of working age Mexican immigrants completed less than a high school diploma, compared to 36 per cent of Vietnamese. At the other end of the educational spectrum, less than 5 per cent of Mexican immigrants have a Bachelor's degree or more, compared to 20 per cent of Vietnamese immigrants. The occupational clustering of each group is also distinct, with higher percentages of Vietnamese immigrants in white collar and professional jobs. Given this variation, statistical analyses allow us to control for the apparent differences and heterogeneity within these two populations in order to assess whether or not gains in human capital transmit equally to gains in labor market access for economic, versus political, migrants.¹² In this section, I highlight the different processes of human capital transmission for economic migrants and refugees. To do so, I focus on employment outcomes and assess the effects of educational attainment and English proficiency for Mexican immigrants versus Vietnamese immigrants.

Table 1 Descriptive statistics – Foreign-born versus native-born population

Table 2 Descriptive statistics – Mexican versus Vietnamese immigrants

Much of the discussion around the importance of human capital has traditionally assumed that these effects are transmitted equally across all segments of the immigrant population. However, as Cortes (2004) and others have found, outcomes for refugee populations differ from those of economic migrants. In this section, I re-examine this finding in the context of employment, while focusing on the largest economic migrant group (Mexican), and the largest refugee group (Vietnamese) (Malone et al, 2003).¹³ In addition to their demographic significance, I opt for a comparison between Mexican and Vietnamese immigrants for several analytical reasons. First, Mexico and Vietnam are both in the list of top five sending countries according to data published by the Bureau of Citizenship and Immigration Services. Second, while the history of Mexican migration to the United States is longer, and Mexican immigrants are further dispersed throughout the nation, California and Texas are the top destination states for over half the population of both groups. Lastly, and most importantly, Mexican and Vietnamese immigrants represent a theoretically important comparison between economic migrants and refugees. One in five Mexican immigrants are citizens (Batalova, 2008), and one in two are estimated to be unauthorized (Passel, 2004). Conversely, over 70 per cent of Vietnamese immigrants are naturalized, and close to half of the Vietnamese migrants who are admitted today do so as immediate relatives of naturalized citizens (Terrazas, 2008). Today the Vietnamese represent less than 8 per cent of all refugee admissions (O’Donnell and Batalova, 2007), however the legacy of the large waves of Vietnamese refugees following the Vietnam War, and the concomitant resources provided by the Office of Refugee Resettlement has created significant community resources in the Vietnamese community for economic incorporation (Bloemraad, 2006). While it is wholly inaccurate to claim that all Mexican immigrants are unauthorized, and all Vietnamese immigrants are refugees, what is clear is that each community has a set of barriers and opportunities by virtue of their relationship to the state. As such, these two groups provide a theoretically relevant set of cases to compare.¹⁴

Overall immigrants tend to have lower unemployment rates, than do their native-born counterparts (Bureau of Labor Statistics, 2008). On the whole Vietnamese immigrant men have been found to have a lower labor force participation rate (Terrazas, 2008), and Mexican immigrant men have been found to have a higher labor force participation rate than other immigrants overall. However, after controlling for demographic and human capital factors, Figure 2 shows that while overall more recent immigrant arrivals in both immigrant communities are disadvantaged, Vietnamese immigrants have more access to the labor market than do Mexican immigrants. While the gender gap for Mexican immigrants is much larger than for Vietnamese immigrants,¹⁵ Mexican immigrant men are still disadvantaged as compared to their Vietnamese counterparts.

Figure 2

Predicted probability of being unemployed, by sex.

Furthermore, while English proficiency is a key element of human capital for immigrant economic incorporation (Chiswick et al, 1997; Funkhouser, 2000), these findings suggest that refugee migrants may be somewhat buffered from the deleterious impacts of not speaking English. For example, Figure 3 shows that on average, while recently arrived non-English-speaking Mexican immigrants have about a 12 per cent predicted probability of being unemployed, the risk for their Vietnamese counterparts is only about 9 per cent. Furthermore, these results show that complete linguistic assimilation is not equally beneficial for all immigrants. In fact, bilingual Mexican migrants have a clear advantage over their monolingual English-speaking counterparts. Bilingual English-speaking Vietnamese migrants do not experience this advantage. While this disparity may likely be related to the proliferation of Spanish as a second language across the country (compared to the relative concentration of Vietnamese speakers in key metropolitan areas), these findings signal the need to re-examine the assumptions around the economic benefits of convergence to the American mainstream skill set.

Figure 3

Predicted probability of being unemployed, by English ability.

In addition to formal credentials and language skills, a third important aspect of human capital is the set of skills and norms that an immigrant gains through US socialization. Though socialization can arguably happen at any time in an immigrant's life course, contact with US educational institutions is a particularly salient experience (Charney and Hernandez, 1998). I assess the effects of this socialization process by controlling for whether an individual migrated before the age of 10 years. Figure 4 reveals that the US educational system may be the primary socializing institution for Mexican migrants. While Mexican immigrants who arrived after the age of 10 years are disadvantaged compared to those who migrated earlier, Vietnamese immigrants who arrived after the age of 10 years fare equally well to their child migrant counterparts. While access to a K-12 public education is one of the only institutions undocumented migrants have access to, refugees are able to draw on an array of other resources through the resettlement process as an adult as well. This added support may mediate the potential barriers of low human capital, thus creating greater parity between those Vietnamese refugees who did and did not benefit from an American educational experience.

Figure 4

Predicted probability of being unemployed, by child immigrant experience (entered the United States before age 10).

Findings II: Labor Market Access, versus Risk of Poverty

In the previous section, I have confronted the drawbacks of ignoring the heterogeneity in the immigrant experience, and specifically demonstrated the ways in which an immigrant's relationship to the state may mediate the effects of human capital. In this section, I propose that it is also vital for researchers to be theoretically specific when considering various elements of the immigrant economic incorporation process. In particular, I argue that the factors influencing access to the labor market may be distinct from the factors that influence their material well-being. Employment alone is not sufficient for evading poverty. Furthermore, family resources may provide additional resources for immigrant survival, which traditional human capital theories do not take into account.

In order to distinguish the key factors driving entry to the labor market, versus material well-being for immigrants in the labor force, I utilize a bivariate probit model to simultaneously estimate the probability that an individual immigrant is unemployed, versus whether they are living in poverty.¹⁶ A bivariate probit model tests whether the key determinants for one outcome significantly differ from another.¹⁷ These results are presented in Table 3 and reveal that while Mexican immigrants are no more likely than other immigrants to be unemployed, all else equal, they are still significantly more likely to live in poverty. Conversely, also controlling for key factors, Vietnamese immigrants are significantly less likely to be unemployed than other immigrants, and the findings in Table 3 suggest that they are also be less likely to be in poverty. In light of these findings , it is important to consider that refugee migrants may have access to the non-market institutions during the resettlement process that are not available to economic migrants.

Table 3 Bivariate probit estimates: Unemployment versus living in poverty

Findings III: The Importance of Context of Reception

Thus far, I have proposed two amendments to the standard conception of immigrant economic incorporation. I have highlighted the ways in an immigrant group's relationship to the state can mediate the effects of human capital, as well as the importance of differentiating between labor market access and material well-being. Lastly, in this section I argue that a nuanced understanding of economic incorporation should also take into account an immigrant's context of reception. Economists typically argue that an immigrant's destination is an endogenous decision he or she makes (that is, immigrants self-select themselves into particular contexts based on their rational calculation of their perceived future success in that place), and consequently factors related to an immigrant's host society may not be an independent determinant of his or her economic outcomes (for example, Brezis and Krugman, 1993; Llull, 2008). However, geographers and sociologists have long looked at the structural and institutional effects of the economic, demographic and policy context, and there is theoretical justification for assuming that an immigrant's destination matters (regardless of whether they made a concerted rational calculation to move there).

Yet, while several immigration scholars have addressed the importance of the context of reception with a cross-national lens (for example, Reitz, 1998; Van Tubergen et al, 2004; Kesler, 2006; Zuberi, 2006), sub-national variation is often ignored. In the wake of welfare-reform in 1996, much focus was placed on the differential access immigrants had for public benefits, depending on their state of residence (Borjas, 2002; for example, Capps et al, 2002).¹⁸ However, an immigrant's context of reception is multi-faceted. First, although individual factors shape an immigrant's desirability as a worker, the supply of jobs also shapes their access to the labor market. The classic economic model predicts that outcomes are negatively impacted in markets where unemployment is high. Low employment rates are generally associated with a healthy, growing economy (Summers, 2007). Second, the history of immigration in a place is important from both a social capital and institutional perspective. Scholars debate the specific impact of a co-ethnic community for immigrant incorporation (Portes and Stepick, 1985; Sanders and Nee, 1987). However, places where immigrants have a longer history and have become integrated in the state's institutions are likely to provide additional resources for later cohorts. Lastly, while the federal government has ultimate authority to set immigration law, states have become increasingly significant in terms of setting immigrant policy (NCSL, 2007). To this end, this analysis assesses the importance of state context of reception, through three measures: (1) labor market competition (measured by the state unemployment rate); (2) the history of immigration in a place (measured by the percent of the state population that is foreign-born); and (3) the policy context for immigrants (measured by the presence of an official English-Only law). This analysis is conducted for a range of economic outcomes in the incorporation process: unemployment, earnings, poverty status and occupational status. The utility of this approach is twofold. First, it assesses the importance of the context of reception, controlling for variation in human capital. Second, it allows us to specifically assess what aspects of economic incorporation these contextual factors matter for. Taking such an expanded approach certainly risks simplicity, yet this discussion highlights the theoretical clarity it also provides. A summary of these findings can be found in Table 4, while additional model results for these four outcomes are provided in Appendix B.¹⁹ While it is clear that state context of reception is relevant, even after controlling for individual characteristics, these effects are not uniform for all aspects of economic incorporation.

Table 4 Context of reception effects for immigrant economic incorporation

The first contextual variable I examine is the economic contexts for immigrant economic incorporation. These results conclude that a poor economic context negatively impacts an individual's labor market access and poverty status, but not necessarily their earnings or occupational outcomes. While immigrants in more robust state economies may not necessarily get paid more or experience more occupational mobility, those whose job prospects are limited, or who have fewer family resources to rely on, are more vulnerable to losing their job or falling into poverty.

Second, I evaluate the impact of the level of immigration in a state, and find significant effects for employment and earnings, while not for poverty or occupational status. The presence of an existing immigrant community may present competition for new arrivals, yet those who are already employed earn more in these states. Specifically, a 10 per cent increase in the percent foreign born in an immigrant's state of residence increases his or her odds of being unemployed 1.18 times. This same demographic shift results in a 5 per cent increase in earnings, all else equal. Thus, these findings call into question the assumption immigrants can always successfully call upon co-ethnic networks for job opportunities. Yet, a larger immigrant workforce may also create a critical mass that can demand higher wages and minimize exploitation.

Lastly, I evaluate the policy context for immigrants. Immigrant advocates such as the American Civil Liberties Union caution that “English-Only laws” represent a hostile policy context for immigrants. Before the 2000 Census, there were 16 states that had adopted an “English-Only” law, which the American Civil Liberties Union summarize as a diverse set of statutes that range from simply declaring English as the “official” language of the state, to limiting the provision of non-English language assistance and services (ACLU, 2008).²⁰ While this variation makes it difficult to speak specifically about the impacts of each type of law, I argue that more importantly, these provisions represent a context where immigrants have become a significant subject of political debate and weigh prominently in the eyes of policymakers (irrespective of the size of that population.) In general, I find that immigrants who live in English-Only states are only 0.91 times as likely to be unemployed as their counterparts in states without these provisions.

However, assessing the impact of immigrant policies on the aggregate foreign-born community may be misleading. For example, through “a difference-in-differences-in-differences” approach using 1980 and 1990 data, Zavodny (2000) finds that the passage of such legislation is particularly adverse for the earnings limited English proficient workers. After controlling for human capital, and allowing effects to vary for Mexican and Vietnamese workers, a dual disparity emerges in these findings.²¹ Figure 5 shows that Mexican immigrants in an English-Only state are more likely to be unemployed than their Vietnamese counterparts. Furthermore, the employment gap between immigrants who reside in English-Only states, versus those who do not, is also slightly larger for Mexican immigrants. This suggests that the context of reception may create differential disparities not only between different immigrant populations, but also within them as well.

Figure 5

Predicted probability of being unemployed, by presence of state English-Only Law.

The proliferation of local immigrant policies also suggests that scholars concerned with the impact of immigrant policy should not only be attentive to the state policy context, as argued by this article, but also to those policies enacted by cities and counties (Murthy, 2007; Rodriguez et al, 2007). Such an expanded approach also calls for innovative data collection strategies, both qualitative and quantitative, that can take into account the variation in place.²²

Evidence for a Multifaceted Process of Immigrant Economic Incorporation

The findings presented in this article support the argument that human capital alone, though integral to economic outcomes, does not solely determine immigrant's labor market outcomes. Institutional and socialization factors also can significantly impact the economic incorporation process. Furthermore, by comparing outcomes for the largest economic immigrant group (Mexicans) to the largest refugee group (Vietnamese), this analysis reaffirms the importance of an immigrant population's relationship to the state. In addition to highlighting the limitations of existing scholarship on immigrant economic incorporation, this article offers suggestions for future research. By expanding our analysis to include structural and institutional factors, in addition to individual characteristics such as human capital, we may better understand what influences immigrant well-being and economic success. These findings also suggest that a “single indicator” approach, while useful for understanding the dynamics of that outcome, may not be generalizable for all aspects of the economic incorporation process. In particular, the factors driving labor market access and success, may not also solely determine poverty risk. Lastly, recognizing the variation in the immigrant experience, we make it possible to avoid over-generalizing the incorporation process for all immigrants.

Appendices

Appendix A

See Tables A1–A2.

Table a1 Explanatory model variables

Table a2 Outcome model variables

Appendix B

See Tables B1–B2.

Table b1 Selected regression results for unemployment, earnings and poverty

Table b2 Selected regression results for occupation