Mexican Migrant Integration in the United States,1965

Abstract

This article uses 20 years of survey data from the Mexican Migration Project to describe how the experiences of Mexico-U.S. migrants have changed over time. I use survey questions about migrants’ experiences on their last U.S. trip to develop composite indices of six integration domains, and then estimate random intercept linear regression models with individual-, family-, and community-level characteristics to explain variations in Mexican migrant integration. I find that, over time, Mexican migrant linguistic and social integration has steadily increased, whereas integration in other family and economic domains changed little or not at all. Results from the multivariate models show the importance of human capital to integration across the multiple domains. Higher education, more time spent in the country of destination, and being male are all strongly associated with higher levels of integration. I also find evidence that both family and community migration networks facilitate integration.

Keywords

immigration integration assimilation Mexico U.S.

Mexican migration to the United States has evolved over the last 75 years from a circular flow of predominantly men working in seasonal agriculture and other low-skilled occupations to a diverse population of temporary migrant men working alongside long-term and settled immigrant men, women, and families (Massey, Durand, and Malone 2002; Riosmena and Massey 2012). Numerically, the number of Mexican-born persons in the United States has grown from 0.5 million in 1950 to approximately 11.5 million in 2013 (Lowell, Perderzini, and Passel 2006, 2; López 2015, 1). While there is a growing literature on the trajectories of settled immigrants and their children with respect to economic status (Connor and Massey 2010; Hall, Greenman, and Farkas 2010), educational attainment (Baum and Flores 2011; Portes and Rivas 2011; Rumbaut and Komaie 2010; Suarez-Orozco et al. 2011), and health (Singh, Rodriguez-Lainz, and Kogan 2013), little is known about the daily experiences and social interactions of recent arrivals, long-term migrants, and returned migrants in Mexico, and how their incorporation into U.S. society has changed over time.

Anti-immigrant policies and movements at local and national levels affect all Mexican migrants in the United States (see Butz and Kehrberg 2016; Jaret 1999; Muste 2013; Wilson 2000; Ybarra, Sanchez, and Sanchez 2016), yet little is known about how rising nativism is affecting the incorporation of recent migrants (Salas, Ayón, and Gurrola 2013). In this study, I use 20 years of survey data from the Mexican Migration Project (MMP) to follow trends over time in a dataset that includes the experiences of long-term, settled, and returned Mexico-U.S. migrants. I use questions about migrants’ experiences on their last U.S. trip to develop composite indices of six integration domains (linguistic, social, family, employment, financial, and assets), and then estimate random intercept linear regression models with individual-, family-, and community-level characteristics to explain cross-sectional and longitudinal variation in Mexican migrant integration. The MMP is one of the few, if not the only, data sources for studying the integration experiences of recently arrived, long-term, and settled Mexico-U.S. migrants alongside returned migrants. Migrant integration experiences likely shape subsequent decisions about settlement versus return and the migration of other family members. Understanding the trend in Mexican migrant experiences over the last five decades offers insights into potential future trends.

Background

Immigration is transforming the demographic and ethnic profile of many high-income countries. In the United States, Canada, and most of Europe, immigrants constitute 10 to 20 percent of the national populations (United Nations 2017). Young immigrants and the children of immigrants often constitute an even larger proportion of the youth populations in these countries. The successful integration of immigrants and their children is a major policy concern in host countries (Kurthen and Heisler 2009). The human capital that immigrants possess, the conditions under which they arrive, and the degree to which they become linguistically and socially integrated impact processes of labor market insertion and wage determination (Connor and Massey 2010; Hall, Greenman and Farkas 2010), health outcomes (Singh, Rodriguez-Lainz, and Kogan 2013), and, most importantly for the future, the educational success of their children (Baum and Flores 2011; Crosnoe 2005; Portes and Rivas 2011, Suárez et al. 2011). How immigrants become a part of the host society has repercussions not only for the immigrant generation and their children, but for the host society as well.

Assimilation theory developed by U.S. scholars contemplates the gradual disappearance over time, and across generations, of differences between immigrants and the native population on a wide range of economic, social, cultural, and political dimensions (Alba and Nee 1997; Gans 1979, 1992; Gordon 1964). Because assimilation encompasses not only the adoption of the host society culture and values, but also the loss of forms of association based on ethnicity, it is heavily focused on long-term and intergenerational change (Gans 1999). Although Gordon suggested that acculturation (the adoption of language, customs, norms and values) and structural assimilation (the development of close social relations with the host society) occur in the immigrant generation, the experiences of temporary and returned migrants are generally not a concern of assimilation theory.

The concept of integration has enjoyed growing use, especially among migration policy experts and researchers in Europe (Hellgren 2015). In the broadest of terms, integration refers to the full participation of immigrants in the host society (Bijl et al. 2008), which, by some scholars, is interpreted as the attainment of parity between immigrants and natives (Kurthen and Heisler 2009). An important part of the discussion and literature on integration defines it as a two-way, interactive process that involves acceptance and accommodation both by immigrants and the host society (Givens 2007). In contrast to assimilation theory, integration does not require the disappearance of all differences nor the abandonment of one’s culture of origin (Berry 1997; Phalet and Swyngedouw 2003; Van Oudenhoven, Prins and Buunk 1998).

Scholars have identified multiple domains of integration—socioeconomic, cultural, legal political, and attitudinal (Entzinger and Biezeveld 2003). Cross-national studies of immigrant policies and incorporation in Europe have developed indicators and data collection systems to measure integration policy and experiences (Sam 2000). Common indicators of integration include paid employment, income, language skills and usage, housing quality, home ownership, contacts with the native population, and intermarriage (Bijl et al. 2008; Di Bartolomeao and Kalantaryan 2017; Entzinger and Biezeveld 2003; Huddleston, Niessen, and Tjaden 2013; Kurthen and Heisler 2009). The definition of integration as full participation in the host country’s institutional and social life is especially well suited to the study of temporary and settled Mexican migrant experiences in the United States. Assimilation theory’s emphasis on intermarriage and educational attainment are not relevant for most first-generation migrants who arrive in the United States as adults, nor is the emphasis on the abandonment of co-ethnic forms of association to understanding the experiences of temporary and returned migrants.

In this study, I define six integration domains that are especially well suited for describing the range of integration experiences among Mexican migrants in the United States, both across distinct migration periods and different migration and settlement strategies. These integration domains are linguistic, social, family, employment, financial, and assets. Linguistic integration refers to the level of English language skills and levels of English usage in various contexts of daily life. Competence in the language of the host society is critical to successful incorporation into the formal labor market, and long-term occupational mobility and wage growth (Borjas 1999; Chiswick 1991; Lindstrom and Massey 1994). It is also necessary for meaningful social interaction with members of the host society, participation in nonethnic-based associations, and access to a broad range of institutions. Linguistic integration involves more than competency in the host country language, it also encompasses regular usage in a wide range of social contexts, including the private spheres of home and friends.

I define social integration as having regular interactions with members of the host society outside of one’s own ethnic group. Interactions with people beyond co-ethnics provide migrants with a sense of belonging to the host society. They also provide the opportunity to develop diverse social networks that can lead to better employment opportunities and provide information and advice about dealing with a range of institutions. Working with other ethnicities, for example, is an indication of employment outside of the immigrant community, which also increases the range of subsequent employment opportunities. Developing friendships with members of other ethnic groups is an especially important indicator of social relations extending beyond the immigrant community. Such relationships increase the likelihood of intermarriage among immigrants and their children (Gsir 2017).

Having a spouse and children in the United States are indicators of family integration. Among migrants who begin as solo labor migrants, bringing a spouse and children into the destination country marks a major transition from temporary to long-term residence. Family life entails interactions with a broad range of institutions, such as schools, medical services, and social-welfare services as well as a change in housing arrangements, all of which increase social interactions and relationships outside the ethnic immigrant community. Having a family in the destination also tends to signify a commitment to remaining in the host society, which provides a strong incentive for developing destination-specific knowledge and social ties.

Employment in the formal sector is an important goal of labor migrants, and it is a key indicator of integration. Being paid by check rather cash, and having social security and federal tax withholdings are good indicators of formal employment in the United States. Work in the formal sector is typically more stable and better paid than informal employment. It also offers access to a number of private as well as public benefits, such as health insurance, disability and unemployment insurance, social security and other retirement benefits. In the case of Mexican migrants in the United States, formal employment also signals legal authorization to work in the United States or the possession of high-quality counterfeit documents. Possession of these documents opens up a wider range of employment opportunities and can lead to wage growth over time that is not available to undocumented workers.

Financial integration involves the use of formal financial services and products that are an integral part of economic life in the host society. Having a bank account makes it easier to manage personal finances, to pay bills, to accumulate savings securely, and to build-up assets and wealth. It also provides access to credit for buying an automobile or home and possibly financing a business (Newberger, Rhine, and Chiu 2004; Rhine and Greene 2006). Possession of a credit card is another financial tool that provides a safe alternative to cash, instant purchasing access to goods and services, and a means to smooth out consumption across variability and interruptions in earnings. Access to banking and credit services requires legal identification and the declaration of one’s presence in the host society (Bair 2003). In the United States, this normally entails getting a Social Security card, a driver’s license, and other forms of identification, and submitting a federal tax return. Establishing a financial identity provides access to the consumer economy, and is essential for long-term wealth accumulation. Research finds that immigrant use of financial services increases with duration after migration (Delener and Katzenstein 1994; Johnson 2007).

Asset integration encompasses home, land, and business ownership in the host county. As Constant et al. note “immigrant homeownership is paramount in the development of immigrants’ economic and social incorporation and advancement in the host country; it represents a stepping stone in their settlement process” (Constant, Roberts, and Zimmermann 2009, 1880; also see Myers, Megbolugbe, and Lee 1998). Studies from a number of countries find lower rates of homeownership among immigrants compared with the native population (Kalantaryan, Gidley, and Caputo 2017). Homeownership is also a measure of immigrant economic success and an indicator of a commitment to remain in the country of destination. It tends to be the single most important source of wealth for both immigrants and the native-born (Alba and Logan 1992). Business ownership is another important indicator of economic integration into the host society. Business ownership requires knowledge of host society laws and regulations, in addition to sufficient savings and access to credit markets. As is the case with home ownership, it represents a commitment to remaining in the host country and creates a larger stake in the local economic and political contexts.

Together these six integration domains cover a broad range of daily experiences in the United States. They also overlap with many measures of assimilation, such as occupational attainment, social interaction with the native population, contact with core social institutions, and the acquisition of assets. Although the domains are related to one another, the level of integration in one domain does not imply integration in another (Fokkema and de Haas 2011). Variation in levels of integration is related to human and social capital, as well as to a period of observation. At the individual level, the same human capital variables that the literature identifies as important determinants of employment outcomes and economic success also apply to successful integration and in some cases to greater acceptance by the host society.

Education is a key predictor of attaining better jobs, better English skills, more chances for interaction with the native population, greater experience and more confidence in working with a wide range of institutions, and, by virtue of the higher status that is accorded to higher levels of education, greater acceptance by the host society (Espinosa and Massey 1997; Fokkema and de Haas 2011). Duration in the United States is another key determinant of integration. Longer durations are associated with higher levels of integration in all domains. Longer durations provide greater exposure to the host society, more time to develop host society language skills, more time in the destination labor market, greater opportunities for occupational mobility and wage growth, and more time to develop relationships with people outside of one’s own ethnic group (Akresh, Massey, and Frank 2014).

Gender is another important variable. Although immigrant occupational niches in host societies frequently have a gender dimension to them that can favor either males or females, many studies of immigrant adaptation find that being male is associated with a higher degree of integration (Venturini 2017). Given the male bread-winner role that is common in most immigrant groups, labor force participation tends to be nearly universal among working age male immigrants, whereas women frequently stay home to care for children. These gender differences in family roles generate substantial differences in opportunities for interaction with the host society, leading to poorer host society language skills, fewer chances for formal employment, and fewer opportunities for interaction and friendships with members of the native population among female migrants compared with males (Fokeema and de Haas 2011).

Age at arrival also plays a role in migrant integration. Older age at arrival is associated in the literature with poorer host society language skills. Migrants who arrive at older ages have less time and fewer opportunities for the development of location-specific human capital and job-based skills. They also tend to be at more advanced stages in the family life cycle and therefore less likely to uproot their families and settle in a new location. Older migrants also have less time for asset accumulation in the place of destination.

Social capital in the form of migration networks plays a powerful role at all stages of the migration process. Migrant networks provide assistance with finding housing and employment; they help migrants navigate the rules and regulations of host country institutions; and they provide emotional support and comfort (Massey and España 1987; Pessar 1999). Whether migrant networks in the end facilitate or hinder the full integration of immigrants into the host society is not always straightforward. Although research clearly shows that migrant networks provide valuable support to newly arrived migrants, an over reliance on co-ethnic migrant networks can also limit the opportunities for interaction with the native population, create excessive dependence on co-ethnics, and foster a degree of insularity with respect to the host society (Fokkema and de Haas 2011; Li 2004; Nee and Sanders 2001; Portes 1998; Venturini 2017).

Finally, the legal and political context of reception matters (Portes and Rumbaut 2014). Immigration policy defines the opportunity structure for legal migrants, the types of jobs they can legally access, the opportunities for family migration and reunification, and the penalties associated with being unauthorized. The period of migration is also associated with the structure of economic incentives for migration, including the demand for immigrant labor in the receiving country and the wage differential between the origin and destination country. The presence and settlement patterns of conational immigrant communities in any given period are the product of migration in prior periods.

Immigrants arriving in well-established immigrant communities will have a very different integration experience compared with the initial founders or pioneers of an immigrant community. In the case of Mexico-U.S. migration, we expect the changing dynamics of Mexico-U.S. migration flows over the past 50 years (post-Bracero) to be associated with increasing integration in all domains. The movement away from agriculture and into urban jobs provided greater opportunities both residentially and in the workplace for contacts with natives, greater opportunities for using English, more stable formal employment opportunities, and better prospects for the migration of relatives and family reunification. Of course, the rise in anti-immigrant sentiments and the accumulation of more restrictive immigration policies over the last several decades have likely had a negative effect on integration.

Data and Methods

I use data collected by the MMP in 109 communities in Mexico (communities 53–161) and 33 destination communities in the United States. The communities were surveyed over a 20-year period from 1997 to 2016. My analysis includes 4,137 male and female household heads with U.S. migration experience. I do not include communities surveyed before 1997 (communities 1–52) because the MMP questionnaire applied in those communities did not include questions on the ownership of residential property, businesses, and land in the United States; the possession of a credit card; submission of a U.S. federal tax return; and use of English during the last U.S. trip. I use the questions on asset ownership in the United States, and the questions on experiences during the last U.S. trip added in 1997, along with other questions on U.S. experience to construct indices for my six integration domains: linguistic, social, family, employment, financial, and assets.

Linguistic integration is measured by the five questions: Do you speak and understand English? Do you use English at home? Do you use English at work? Do you use English with friends? and Do you use English in your neighborhood? The ordinal response set for English ability is “doesn’t speak nor understand,” “doesn’t speak but understands a little,” “doesn’t speak but understands well,” “both speaks and understands a little,” and “both speaks and understands well.” The ordinal response set for English use is “none,” “sometimes,” “often,” and “always.” Social integration is measured by three questions: What kind of relations do you have with blacks? What kind of relations do you have with Asians? and What kind of relations do you have with Gringos? The ordinal response set for these questions is “none or casual,” “workplace,” “friendship,” “very close,” and “other.” Family integration is measured by whether the respondent’s spouse is in the United States and the number of children in the United States.

Employment integration is measured by indicators of whether the respondent was paid with a check, had Social Security taxes withheld, and had federal taxes withheld on the most recent U.S. job. Financial integration is measured by whether the respondent had a bank account in the United States, had a credit card in the United States, and filed a federal tax return. Finally, asset integration is measured by whether the respondent owned a house in the United States, owned a business in the United Sates, and the number of hectares of agricultural land owned in the United States. I used principal components analysis to construct standard normal indices from the question sets for each integration domain, with high values on the indices corresponding to high levels of integration.

Table 1 presents Cronbach’s alpha for each of the six sets of questions. Cronbach’s alpha is a measure of internal reliability that ranges from zero to one. A high value of Cronbach’s alpha indicates the variables are highly correlated with one another and together offer a good measure of a common, latent construct. Cronbach’s alpha ranges from a low of 0.58 for asset integration to a high of 0.93 for employment integration. On all six of the question sets, Cronbach’s alpha indicates a moderate-to-very-high level of internal reliability. I also constructed an overall integration index based on all nineteen questions. Cronbach’s alpha for all nineteen questions is 0.84.

Table 1

Measures of Integration, Last U.S. Trip for Household Heads and Spouses, MMP Communities 53–161, Mexico and U.S. Samples, 1997–2016

	Cronbach’s Alpha	Percent
Linguistic integration^a	0.886
Speaks and understands English		25.3%
Uses English at home		18.3%
Uses English at work		48.8%
Uses English with friends		22.4%
Uses English in neighborhood		34.1%
Social integration^b	0.766
Social relations with blacks		36.1%
Social relations with Asians		30.6%
Social relations with Whites		72.0%
Family integration	0.690
Spouse in U.S.		16.1%
Children in U.S.^c		24.1%
Employment integration	0.934
Paid with check		63.2%
Social Security taxes withheld		60.9%
Federal taxes withheld		58.4%
Financial integration	0.836
Bank account in U.S.		14.8%
Credit Card in U.S.		9.7%
Filed federal tax return		20.2%
Asset integration	0.578
Owns house in U.S.		3.4%
Business in U.S.		1.1%
Land in U.S.^d		0.2%
Overall integration	0.844
Number of observations	4,137

Speaks and understands English a little or well, uses English sometimes, often, or always.

Social relations in workplace, friendship, very close, other.

One or more children in the U.S.

Has any agricultural land in the U.S.

I use the six integration indices and the overall integration index to follow trends in Mexican migrant integration from 1965 to 2015. I also use these indices as dependent variables to estimate random intercept linear regression models using individual measures of human capital, immigrant social capital, period, and U.S. destination states or regions as independent variables. The random intercept models include a random error component at the community level, and robust estimates of the standard errors. The human capital variables include gender (male), highest completed year of schooling, age, and duration and legal status (unauthorized) of last U.S. trip. The social capital variables measure migration networks at the household and community levels. Family networks is the number of adult siblings of the household head and the spouse of the head who have U.S. migration experience prior to, or during, the most recent U.S. trip. The prevalence ratio is the percentage of adults in the community of origin who had U.S. migration experience measured in the most recent year of the last U.S. trip.

I define four periods, 1965–1985, 1986–1995, 1996–2005, and 2006–2015. The first period, 1965–1985 corresponds to the two decades between the end of the Bracero Program and the passage of the Immigration Reform and Control Act (IRCA) of 1986. This is a period of relatively lax border control and predominantly male circular migration. The second two post-IRCA periods, 1986–1995 and 1996–2005, correspond to the intensification of border control efforts, a continued increase in the number of Mexicans in the United States, and a shift away from predominantly male, temporary, and circular migration toward family migration and settlement. The fourth period, 2006–2016, is characterized by declining Mexican migration to the United States, a decline in the stock of Mexicans in the United States, and a substantial decline in U.S. border apprehensions of Mexicans. The period also overlaps with the great recession of 2008 and increasingly more visible anti-immigrant public sentiment.

To control for regional variation in the immigrant-receiving context, I define four destination states and five regions in the United States. California, Illinois, New York and, Texas are the four most common destination states for migrants in the MMP samples (California, Illinois and Texas are the three states with the largest Mexican-born populations). I group the other destination states into four regions: other west, other mid-west, other south, and other northeast, and I put the 1.1 percent missing destination cases into a missing category. Finally, I include a dummy variable for U.S. sample. The U.S. sample is designed to capture migrants who are resident in the United States and unlikely to be captured in the Mexico community samples.

Table 1 also presents the percentage distributions for the nineteen variables used to measure integration. Only one in four migrants in my MMP analytical sample spoke and understood English a little or well, yet close to one in five used English with some frequency at home, and close to 50 percent used English at work. However, a majority of migrants did not use English in any context, which signals a high degree of relative isolation from non-Spanish speakers, at least in terms of interactions that are more than rudimentary. Even so, 72 percent of migrants reported some degree of social relations with whites, and approximately one-third reported social relations with blacks and Asians.

The large majority of migrant household heads from the MMP were in the United States without a spouse and without children. Only 16 percent of migrant heads were accompanied by a spouse on the last trip, and 24 percent had a least one child with them. An important indicator of formal employment in the United States is payment by check, and social security and federal tax withholdings. Approximately 60 percent of migrants reported being paid by check with social security and federal tax withholdings. In contrast to formal employment, relatively few of the migrant household heads actually filed a federal tax return (20 percent), and even fewer used financial services such as a bank (15 percent) or had a credit card (10 percent). Asset ownership in the United States is especially rare among migrant household heads in the MMP sample. Only 3 percent of household heads owned a home in the United States on their last trip, and 1 percent or less owned a business or agricultural land.

Trends in Integration and Its Correlates

Figure 1 presents graphs of the six integration indices and the overall integration index for the years 1970 to 2015. Sample estimates for the years 1965 to 1969 are not shown due to the relatively small number of observations. There are three distinguishable patterns in the graphs. First, there is a gradual increase over time in all of the indices. Second, there is a trend toward convergence in the different graphs up until around 2005, after which they increasingly diverge from one another. Third, employment and assets show the smallest amount of change over time, with virtually no change in assets up until the end of the 1990s, which is consistent with the very low level of asset ownership in the sample. There is an upward spike in several of the graphs in the last year, which may reflect features specific to the migration streams from one or more of the communities that were surveyed in 2015 and 2016, rather than a departure from the overall trend.

Figure 1

Mean Integration Indexes by Most Recent Year in the United States (1970–2015), Household Heads and Spouses, MMP Communities 53–161, Mexico and U.S. Samples, 1997–2016

Along with the gradual rise in the level of integration of migrant household heads in the MMP sample, there have been some important changes in several of the key determinants of the migration experience. Table 2 presents the measures of human and social capital by period of last year in the United States. Over the roughly 50-year period covered by the data, migrant household heads surveyed by the MMP have remained overwhelmingly male, ranging from 94 to 96 percent. This pattern reflects both the predominance of male household headship in Mexico and the continued predominance of males among circular and retuning Mexican migrants (Aguilar 2017; Lowell, Perderzini, and Passel 2006).

Table 2

Human and Social Capital by Period of Last Year in the United States, Household Heads and Spouses with U.S. Migration Experience, MMP Communities 53–161, Mexico and U.S. Samples, 1997–2016

		Means and Percentages
		Period of Last Year in United States
	1965–1985	1986–1995	1996–2005	2006–2016
Human capital
Male	93.7%	95.1%	95.8%	94.3%
Education	4.8	6.6	6.7	7.1	**
Age trip	28.8	29.6	32.3	32.1	**
Duration trip	1.9	2.7	4.5	8.9	**
Unauthorized	80.9%	78.3%	68.5%	71.8%	**
Social capital
Family network	0.5	0.8	1.0	0.7	**
Prevalence ratio	15.4%	17.7%	19.8%	16.2%	**
Number of cases	650	761	1899	827

Significance levels of difference of means/proportions tests: **p<0.01, **p<0.05.

The mean number of years of schooling completed rose from 4.5 years in the earliest post-Bracero period to 7.1 years in the most recent period. This increase in education parallels a comparable rise in education over the same period in Mexico (Alcántara 2008). Over the first three periods from 1965 to 2005, the mean age at the time of departure on the last U.S. trip rose by more than three years from 28.8 to 32.3 years. The most dramatic change in migration experience over the 50-year period is the increase in mean trip duration from 1.9 years in the first period to 8.9 years in 2006–2016. This large increase in the length of trips reflects the movement away from temporary, circular migration toward settled and long-term migration that followed the U.S. government’s growing efforts at tighter border control and interdiction.

Along with the increase in trip duration there has been a decline in the prevalence of unauthorized migration. Roughly 80 percent of migrant household heads who were in the United States on their last trip lacked authorization to work during the years 1965 to 1995. The percentage of unauthorized migrant heads dropped to approximately 70 percent in the last two periods covering 1996–2016. At the same time, there was a small, but statistically significant, rise in the level of migrant social capital. The mean number of siblings with U.S. migration experience increased as did the prevalence of U.S. migration in the communities sampled by the MMP.

The MMP is not a nationally representative sample of Mexican migrants resident in the United States, of course, nor is it nationally representative of all returned migrants in Mexico. Nevertheless, in general, the levels and patterns of change in human capital and trip characteristics among household heads in the sample are broadly consistent with other data sources. For example, Feliciano (2005), using U.S. census data, found a rise over time in the educational levels of Mexican migrants to the United States. Sáenz (2015), using data from the American Community Survey (ACS), found a rise in both educational levels and age in the first decade of the twenty-first century. Using data from the Survey of Migration on the Northern Border [Encuesta sobre Migración en la Frontera Norte de México (EMIF)], Gomez del León and Tuirán (2000) found a tendency in the last decades of the twentieth century toward longer durations in the United States. My review of historic trends in the indices constructed here reveals an increase in integration over time. Based on findings from other studies and my predictions, I expect the increases in the integration indices to be linked to increases in education, trip duration, and authorized migration. In the next section, I use multivariate regression models to estimate the relative effects of the human and social capital variables on integration experiences in the United States, and to test for period effects net of changes in the composition of migrant human and social capital.

The Determinants of Integration

Table 3 presents the coefficients from the random intercept linear regression models predicting integration during the last U.S. trip. I begin with the overall integration index that is based on all nineteen integration questions. As expected, being male, having more years of schooling, and longer durations in the United States are all associated with significantly higher levels of integration, as is being part of the U.S. sample. The U.S. sample captures migrants who are resident in the United States, many of whom no longer maintain households in the community of origin. These migrants have selected themselves into settled residence in the United States rather than returning regularly to Mexico, and, therefore, they have a higher degree of commitment to making a new life in the United States than migrants who still return to Mexico.

Table 3

Parameter Estimates from Random Intercept Linear Regression Models Predicting Integration on Last U.S. Trip, Household Heads and Spouses, MMP Communities 53-161, Mexico and U.S. Samples, 1997–2016

	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7
	Overall	Linguistic	Social	Family	Employment	Financial	Asset
Human capital
Male	0.330**	0.212**	–0.027	0.082	0.544**	0.154*	0.174**
Education	0.041**	0.056**	0.029**	–0.007*	0.001	0.017**	0.002
Age trip	–0.001	–0.008**	0.002	0.013**	–0.000	0.002	0.003**
Duration trip	0.052**	0.029**	0.028**	0.063**	0.015**	0.051**	0.037**
Unauthorized	–0.363**	–0.205**	–0.029	–0.195**	–0.314**	–0.418**	–0.029
Social capital
Family network	0.076**	0.049**	0.048**	0.030*	0.051**	0.071**	0.017*
Prevalence ratio	0.005*	0.001	–0.002	0.010**	0.007**	0.004*	0.002
U.S. Sample	0.380**	0.028	–0.185	0.667**	0.207**	0.565**	0.597**
Period
1965–1985 (ref)
1986–1995	0.071*	0.089*	0.058	0.078*	0.028	–0.011	–0.031
1996–2005	0.120*	0.154**	0.125*	0.080*	0.023	0.012	–0.100**
2006–2016	0.267**	0.330**	0.166*	0.078	0.076	0.127*	–0.017
Destination
California (ref)
Illinois	0.044	–0.018	0.005	–0.004	0.170*	0.026	0.018
New York	–0.095	0.046	0.086	–0.029	–0.334**	–0.150*	–0.049
Texas	–0.070	–0.037	0.016	–0.082	–0.118	–0.110*	0.128
Other west	–0.038	–0.044	0.030	–0.060	–0.019	–0.058	0.065
Other mid-west	0.047	–0.022	0.050	–0.109*	0.217**	0.034	–0.024
Other south	–0.098	–0.114*	0.005	–0.130**	0.014	0.106**	0.081
Other northeast	0.174*	0.184	0.096	–0.068	0.078	0.094	0.162**
Missing	–0.682**	–0.632**	–0.426**	–0.180**	–0.664**	–0.202**	0.015
ρ	0.060	0.049	0.093	0.012	0.054	0.023	0.000
R ²	0.341	0.173	0.074	0.377	0.113	0.324	0.134
Observations	4,137	4,137	4,137	4,137	4,137	4,137	4,137

p<0.01. **p<0.05.

As expected, unauthorized status is associated with significantly lower levels of integration. The magnitude of the effect is approximately equivalent to that of being in the U.S. sample, but in the opposite direction. Both family migration networks and the prevalence of migration in the community of origin are associated with significantly higher levels of integration. Net of human and social capital, the period effects are positive and significant, and increase in magnitude across periods. Compared with the first post-Bracero period, migrant household heads in the MMP sample show higher levels of overall integration as the periods progress toward the present.

Most of the destination effects are not statistically significant, with one important exception. Migrants with missing destination information on last U.S. trip have significantly lower levels of overall integration compared with other migrants. Although this group constitutes only 1 percent of the sample, the difference is notable. Migrants with missing destination information may have been very mobile and lived in multiple states on their last trip—which is the experience of many agricultural workers. The model provides a very good fit to the data: R² is 0.341, indicating that the human capital, social capital, and period measures explain a substantial proportion of variation in the overall integration index.

Turning to the indices for each of the six integration domains, both trip duration and family networks have significant positive effects on all six indices. Longer durations and having more siblings in the United States increase the chances of integration in all domains. The other human and social capital variables are significant in most but not all of the domains. Being male, having a higher education, and originating in a community with a higher prevalence of migration are all associated with a significantly higher degree of integration across a number of domains.

Also consistent with the results for the overall integration index, unauthorized status in the United States is associated with significantly lower levels of integration in four of the domains. Being in the U.S. sample is not a significant factor for linguistic or social integration, but it is significant for the domains of family, employment, financial, and asset integration. Indeed, the magnitudes of the coefficients for U.S. sample are among the largest of all coefficients in the models. The period coefficients provide very clear evidence of increasing integration over time in the linguistic and social domains. However, family integration increases from the first to the second period but then goes stagnant; and there is no evidence of change over time with respect to the employment, financial, and asset domains once we take into account human and social capital. Of the nine period coefficients corresponding to these three domains only two are statistically significant, and they are in opposite directions.

In summary, I find very clear evidence of significant human and social capital effects on migrant integration in the United States. Male migrants, those with higher levels of education, longer trip durations, and those in authorized status experience the highest levels of integration. Age at time of migration has mixed effects. Older arriving migrants have lower levels of linguistic integration, but they are more likely than their younger counterparts to be accompanied by a spouse and children and to have assets in the United States.

A key finding is that both family migration networks and community migration networks facilitate integration. There is no evidence of negative network effects in any of the six integration domains. The positive effects on integration of having siblings in the United States, along with trip duration, are the most consistent results across all domains. Migration networks facilitate not only migration, but also integration into the host society. The linguistic and social integration of Mexican migrants in the MMP sample has steadily increased over the last 50 years, even after taking into account change over time in background characteristics that are associated with higher levels of integration. The levels of linguistic and social integration also do not seem to be affected by whether a migrant was in the Mexico or the U.S. sample, which is a good proxy of returned and settled migration.

Thus over the last 50 years, the general trend for Mexican migrants is one of increasing contact and interaction with people outside the Mexican community, regardless of whether they were temporary, long-term, or settled migrants. On the other hand, employment, financial, and asset integration all appear to be strongly associated with being in the U.S. sample, and net of changes in other factors, do not appear to have changed over time. Similarly, family integration, which rose following IRCA, has remained relatively unchanged for the last 30 years.

Discussion and Conclusion

Whether immigrants remain on the margins of society or become full participants of social and institutional life is an important issue in immigrant receiving countries. The concept of integration is especially important for considering the experiences of temporary, long-term, or settled migrants. Indicators of integration capture the daily experiences of migrants and the long-term processes of language acquisition, the development of new social networks, family building, and the accumulation of assets. Early integration experiences are also likely to factor into migrants’ decisions about returning to their country of origin, especially in migration streams that have a large circular component.

The MMP data are uniquely suited to study the integration experiences of returned and settled migrants, and change over time in those experiences. The binational nature of the sample captures migrants at different stages of the settlement process in the United States, and includes both authorized and unauthorized migrants, as well as returned migrants in Mexico. I took advantage of the richness of the MMP to develop composite measures of the multifaceted domains of migrant integration. My six integration domains cover the diverse ways in which temporary, long-term, or settled migrants experience life in the United States. They include English language skills and usage, different contexts of social interaction with non-Mexicans, having a spouse and children in the United States, having a formal job, using financial services and submitting income tax returns, and owning a home or other assets.

These domains of integration capture experiences that facilitate and overlap with assimilation, but are also relevant to the lives of temporary migrants. High levels of integration often coexist with cultural change and the adoption of new national identities, but they do not require these changes. The results from the multivariate models show the clear importance of human capital variables for integration across the multiple domains. The results are consistent with findings from other studies that show that higher education, more time spent in the country of destination, and being male are all strongly associated with higher levels of integration.

I also find that, contrary to concerns that migrant networks might encourage insularity from the host society, networks actually facilitate integration. Family networks in particular are associated with higher levels of integration in all domains. The prevalence of migration in the community of origin is significantly associated with higher levels of integration in the family, employment, financial, and asset domains. These are the same domains in which being in the U.S. sample also makes a significant difference. Having a family in the United States, owning a home, having a formal job, and using formal financial services are all indicators of settlement. Access to a broad network of migrants from the community of origin is especially important for accessing good, stable jobs that are also integral to family settlement, home ownership, and financial integration.

A potential source of concern for the long-term integration of Mexican migrants is my finding that integration in the family, employment, financial, and asset domains has either stagnated or not improved over the 50-year period covered by the sample. In contrast, net of other factors, linguistic and social integration has steadily risen over the same period. Linguistic and social integration involve a combination of opportunities for interaction with the native population, and the willingness of migrants to learn and use English, while developing relationships with non-Mexicans.

My findings provide evidence that the structural conditions affecting Mexican migrants in the United States are increasingly favorable to English language usage and social interaction with other ethnic groups, and Mexican migrants are embracing these opportunities for linguistic and social integration. In addition, the movement of Mexican migrants into nonagricultural jobs and urban areas has brought them into contact with a wider cross-section of U.S. society. The hardening of the U.S. southern border and the rise of family and long-term settled migration has also increased the incentives for integration. However, there does not appear to have been a comparable change in the structural conditions that promote employment, financial, and asset integration beyond the effects of changes in the composition of Mexico-U.S. migration streams. A major impediment to integration in all domains is unauthorized status, which is associated with significantly lower integration on the overall index and on indices for four of the six domains. It has an especially large effect on employment and financial integration, both of which require legal forms of identification.

There are important limitations to this work. First, the MMP is not nationally representative of Mexicans in the United States, nor of all returned migrants in Mexico. It would be wrong to use these data to make statements about the integration of all Mexicans in the United States. At the same time, however, nationally representative sources of data on the Mexican-born population in the United States, such as the census and the American Communities Survey, do not capture temporary and returned migrants, and under-enumerate unauthorized migrants. Because many long-term and settled Mexican migrants begin as temporary migrants, it is important to study the experiences of temporary and returned migrants. Their experiences of marginalization or integration influence decisions about staying in the United States and are also shared by other Mexican migrants who decide to remain and become a part of U.S. society.

A second limitation is the absence of information on other domains of integration that are equally important for understanding the migrant experience. Integration is a two-way process that involves both the migrant’s efforts to be a part of the host society and the host society’s willingness and efforts to accept the migrant as a full member of society. The survey does not measure experiences and perceptions of discrimination, or feelings of insecurity and fear of deportation that are associated with unauthorized status. What my results do show is that, despite the rise in anti-immigrant public sentiment and policies designed to marginalize unauthorized migrants, temporary, long-term, and returned migrants surveyed over the last 20 years are learning and using English at increasingly higher rates, and are growing less isolated from non-Mexicans over time. These trends provide evidence of structural factors at work that facilitate the gradual integration of Mexican migrants into U.S. society, albeit unevenly across different domains, and the desire of Mexican migrants to be part of U.S. society.

Footnotes

David P. Lindstrom is a professor of sociology and a faculty associate of the Population Studies and Training Center, Brown University. His research examines the determinants and consequences of migration in economically developing societies, the transition into adulthood, and the changing dynamics of reproductive health and behavior.

References

Aguilar

Lorena

. 2017. La situación demográfica de México, 2016. México, DF: Consejo Nacional de Población. Available from https://www.gob.mx/conapo/documentos/44667.

Akresh

Ilana R.

Massey

Douglas S.

Frank

Reanne

. 2014. Beyond English proficiency: Rethinking immigrant integration. Social Science Research 45 (1): 200–10.

Alba

Richard D.

Logan

John R.

. 1992. Assimilation and stratification in the homeownership patterns of ethnic groups. International Migration Review 26 (4): 1314–1341.

Alba

Richard D.

Nee

Victor

. 1997. Rethinking assimilation theory for a new era of immigration. International Migration Review 31 (4): 826–74.

Alcántara

Armando

. 2008. Políticas educativas y neoliberalismo en México: 1982–2006. Revista Iberoamericana de Educación 48:147–65.

Bair

Sheila C.

2003. Improving access to the U.S. banking system among recent Latin American immigrants. Washington, DC: Inter-American Development Bank.

Baum

Sandy

Flores

Stella M.

. 2011. Higher education and children in immigrant families. The Future of Children 21 (1): 171–93.

Berry

John W.

1997. Immigration, acculturation and adaptation. Applied Psychology: An International Review 46 (1): 5–68.

Bijl

Rob V.

Zorlu

Aslan

Jennissen

Roel P.W.

Blom

Martine

. 2008. The integration of migrants in The Netherlands monitored over time: Trend and cohort analyses. In International migration in Europe: New trends and new methods of analysis, eds. Bonifazi

Corrado

Okólski

Marek

Schoorl

Jeanette

Simon

Patrick

, 199–223. Amsterdam: Amsterdam University Press.

10.

Borjas

George J.

1999. Heaven’s door: Immigration policy and the American economy. Princeton, NJ: Princeton University Press.

11.

Butz

Adam

Kehrberg

Jason E.

. 2016. Estimating anti-immigrant sentiment for the American states using multi-level modeling and post-stratification, 2004–2008. Research and Politics. DOI: 10.1177/2053168016645830.

12.

Chiswick

Barry R.

1991. Speaking, reading, and earnings among low skilled immigrants. Journal of Labor Economics 9 (2): 149–70.

13.

Connor

Phillip

Massey

Douglas S.

. 2010. Economic outcomes among Latino migrants to Spain and the United States: Differences by source region and legal status. International Migration Review 44 (4): 802–29.

14.

Constant

Amelie F.

Roberts

Rowan

Zimmermann

Klaus F.

. 2009. Ethnic identity and immigrant homeownership. Urban Studies 46 (9): 1879–1898.

15.

Crosnoe

Robert

. 2005. Double disadvantage or signs of resilience? The elementary school contexts of children from Mexican immigrant families. American Educational Research Journal 42 (2): 269–303.

16.

Delener

Nejdet

Katzenstein

Herbert

. 1994. Credit card possession and other payment systems: Use patterns among Asian and Hispanic consumers. The International Journal of Bank Marketing 12 (4): 13–24.

17.

Di Bartolomeo

Anna

Kalantaryan

Sona

. 2017. Counting the uncountable: Measuring migrant integration between origin and destination. In Migrant integration between homeland and host society, volume 1: Where does the country of origin fit? eds. Weinar

Agnieszka

Unterreiner

Anne

Fargues

Philippe

, 29–37. Cham, Switzerland: Springer.

18.

Entzinger

Han

Biezeveld

Renske

. 2003. Benchmarking in immigrant integration. Rotterdam: European Research Centre on Migration and Ethnic Relations, Erasmus University.

19.

Espinosa

Kristen E.

Massey

Douglas S.

. 1997. Determinants of English proficiency among Mexican migrants to the United States. International Migration Review 31 (1): 28–50.

20.

Feliciano

Cynthia

2005. Educational selectivity in U.S. immigration: How do immigrants compare to those left behind? Demography 42 (1): 131–152.

21.

Fokkema

Tineke

de Haas

Hein

. 2011. Pre- and post-migration determinants of socio-cultural integration of African immigrants in Italy and Spain. International Migration 53 (6). doi:10.1111/j.1468-2435.2011.00687.x.

22.

Gans

Herbert J.

1979. Symbolic ethnicity: The future of ethnic groups and cultures in America. Ethnic and Racial Studies 2 (1): 1–20.

23.

Gans

Herbert J.

1992. Second-generation decline: Scenarios for the economic and ethnic futures of the post-1965 American immigrants. Ethnic and Racial Studies 15 (2): 173–92.

24.

Gans

Herbert J.

1999. Toward a reconciliation of “assimilation” and “pluralism”: The interplay of acculturation and ethnic retention. In The handbook of international migration: The American experience, eds. Hirschman

Charles

Kasinitz

Phillip

DeWind

Joshua

, 161–71. New York, NY: Russell Sage Foundation.

25.

Givens

Terri E.

2007. Immigrant integration in Europe: Empirical research. Annual Review of Political Science 10:67–83.

26.

Gómez de León

José

Tuirán

Rodolfo

. 2000. Patrones de continuidad y cambio de la migración hacia Estados Unidos. In Migración México-Estados Unidos presente y future. eds. Tuirán

Rodolfo

19-28. México, D.F.: Consejo Nacional de Población.

27.

Gordon

Milton M.

1964. Assimilation in American life: The role of race, religion, and national origins. New York, NY: Oxford University Press.

28.

Gsir

Sonia

. 2017. Do countries of origin contribute to socio-cultural integration of migrants abroad? In Migrant integration between homeland and host society, volume 1: Where does the country of origin fit? eds. Weinar

Agnieszka

Unterreiner

Anne

Fargues

Philippe

149–72. Cham, Switzerland: Springer.

29.

Hall

Matthew

Greenman

Emily

Farkas

George

. 2010. Legal status and wage disparities for Mexican immigrants. Social Forces 89 (2): 491–514.

30.

Hellgren

Zenia

. 2015. Immigrant integration as a two-way process: Translating theory into practice. Interdisciplinary Research Group on Immigration Working Paper Series 23. Barcelona: Universitat Pompeu Fabra.

31.

Huddleston

Thomas

Niessen

Jan

Tjaden

Jasper D.

. 2013. Using EU indicators of immigrant integration: Final report for directorate-general for home affairs. Luxembourg: European Commission.

32.

Jaret

Charles

. 1999. Troubled by newcomers: Anti-immigrant attitudes and action during two eras of mass immigration to the United States. Journal of American Ethnic History 18 (3): 9–39.

33.

Johnson

Phyllis J.

2007. Credit card practices of Vietnamese and Laotian newcomers to Canada: A 10-year longitudinal perspective. Journal of Family and Economic Issues 28 (2): 227–46.

34.

Kalantaryan

Sona

Gidley

Ben

Caputo

Maria Luisa

. 2017. Residential integration: Towards a sending country perspective. In Migrant integration between homeland and host society, volume 1: Where does the country of origin fit? eds. Weinar

Agnieszka

Unterreiner

Anne

Fargues

Philippe

, 117–47. Cham, Switzerland: Springer.

35.

Kurthen

Hermann

Heisler

Barbara S.

. 2009. Immigrant integration: Comparative evidence from the United States and Germany. Ethnic and Racial Studies 32 (1): 139–70.

36.

Peter S.

2004. Social capital and economic outcomes for immigrants. Journal of International Migration and Integration 5 (2): 171–90.

37.

Lindstrom

David P.

Massey

Douglas S.

. 1994. Selective emigration, cohort quality, and models of immigrant assimilation. Social Science Research 23 (4): 315–49.

38.

López

Gustavo

. 2015. Hispanics of Mexican origin in the United States, 2013. Washington, DC: Pew Research.

39.

Lowell

B. Lindsay

Perderzini

Carla

Passel

Jeffrey S.

. 2006. The demography of Mexico/U.S. migration. Washington, DC: Institute for the Study of International Migration, Walsh School of Foreign Service, Georgetown University.

40.

Massey

Douglas S.

Durand

Jorge

Malone

Nolan J.

. 2002. Beyond smoke and mirrors: Mexican immigration in an era of economic integration. New York, NY: Russell Sage Foundation.

41.

Massey

Douglas S.

García España

Felipe

. 1987. The social process of international migration. Science 237 (4816): 733–38.

42.

Muste

Christopher P.

2013. The dynamics of immigration opinion in the United States, 1992–2012. Public Opinion Quarterly 77 (1): 398–416.

43.

Myers

Dowell

Megbolugbe

Isaac F.

Lee

Seongwoo

. 1998. Cohort estimation of homeownership attainment among native-born and immigrant Populations. Journal of Housing Research 9 (2): 237–69.

44.

Nee

Victor

Sanders

Jimy

. 2001. Understanding the diversity of immigrant incorporation: A forms-of-capital model. Ethnic and Racial Studies 24 (3): 386–411.

45.

Newberger

Robin

Rhine

Sherrie L.W.

Chiu

Shirley

. 2004. Immigrant financial market participation: Defining the research questions. Chicago Fed Letter 199:1–4.

46.

Pessar

Patricia R.

1999. The role of gender, households, and social networks in the migration process: A review and appraisal. In The handbook of international migration: The American experience, eds. Hirschman

Charles

Kasinitz

Phillip

DeWind

Joshua

, 53–70. New York, NY: Russell Sage Foundation.

47.

Phalet

Karen

Swyngedouw

Marc

. 2003. Measuring immigrant integration: The case of Belgium. Studi Emigrazione/Migration Studies 40 (152): 773–803.

48.

Portes

Alejandro

. 1998. Social capital: Its origins and applications in modern sociology. Annual Review of Sociology 24:1–24.

49.

Portes

Alejandro

Rivas

Alejandro

. 2011. The adaptation of migrant children. The Future of Children 21 (1): 219–46.

50.

Portes

Alejandro

Rumbaut

Rubén G.

. 2014. Immigrant America: A portrait, 4th ed. Berkeley, CA: University of California Press.

51.

Rhine

Sherrie L. W.

Greene

William H.

. 2006. The determinants of being unbanked for U.S. immigrants. The Journal of Consumer Affairs 40 (1): 21–40.

52.

Riosmena

Fernando

Massey

Douglas S.

. 2012. Pathways to El Norte: Origins, destinations, and characteristics of Mexican Migrants to the United States. International Migration Review 46 (1): 3–36.

53.

Rumbaut

Rubén G.

Komaie

Golnaz

. 2010. Immigration and adult transitions. The Future of Children 20 (1): 43–66.

54.

Sáenz

Rogelio.

2015. A transformation in Mexican Migration to the United States. Carsey Research, National Issue Brief #86, University of New Hampshire.

55.

Salas

Lorraine M.

Ayón

Cecilia

Gurrola

Maria

. 2013. Estamos traumados: The effect of anti-immigrant sentiment and policies on the mental health of Mexican immigrant families. Journal of Community Psychology 41 (8): 1005–1020.

56.

Sam

David L.

2000. Psychological adaptation of adolescents with immigrant backgrounds. The Journal of Social Psychology 140 (1): 5–25.

57.

Singh

Gopal K.

Rodriguez-Lainz

Alfonso

Kogan

Michael D.

. 2013. Immigrant health inequalities in the United States: Use of eight major national data systems. The Scientific World Journal. doi.org/10.1155/2013/512313

58.

Suárez-Orozco

Manuel M.

Darbes

Tasha

Dias

Sandra I.

Sutin

Matt

. 2011. Migration and schooling. Annual Review of Anthropology 40:311–28.

59.

United Nations. 2017. International Migration 2017. United Nations, Department of Economic and Social Affairs, Population Division. https://www.un.org/en/development/desa/population/migration/data/estimates2/estimates17.asp

60.

Van Oudenhoven

Jan Pieter

Prins

Karen S.

Buunk

Bram P.

. 1998. Attitudes of minority and majority members towards adaptation of immigrants. European Journal of Social Psychology 28 (6): 995–1013.

61.

Venturini

2017. Immigrant assimilation in the labour market: What is missing in economic literature. In Migrant integration between homeland and host society, volume 1: Where does the country of origin fit? eds. Weinar

Agnieszka

Unterreiner

Anne

Fargues

Philippe

, 21–42. Cham, Switzerland: Springer.

62.

Wilson

Tamar D.

2000. Anti-immigrant sentiment and the problem of reproduction/maintenance in Mexican immigration to the United States. Critique of Anthropology 20 (2): 191–213.

63.

Ybarra

Vickie D.

Sanchez

Lisa M.

Sanchez

Gabriel R.

. 2016. Anti-immigrant anxieties in state policy: The great recession and punitive immigration policy in the American states, 2005-2012. State Politics & Policy Quarterly 16 (3): 313–39.

Mexican Migrant Integration in the United States,1965–2015

Abstract

Keywords

Background

Data and Methods

Trends in Integration and Its Correlates

The Determinants of Integration

Discussion and Conclusion

Footnotes

References