Introduction

While much is known about the effects of migrant networks on the labor outcomes of Mexican migrants in the United States (Aguilera & Massey, 2003; Amuedo-Dorantes & Mundra, 2007; Kossoudji & Cobb-Clark, 2000; Munshi, 2003), we know less about the occupational mobility of Mexicans moving to the U.S., and even less about their mobility within the U.S. This paper analyzes the occupational mobility of Mexicans who migrated to the U.S. after 1965. Using life history data from household heads and their spouses in 154 Mexican communities from the Mexican Migration Project, we compare the first occupation attained in the U.S. to the last occupation in Mexico and estimate the determinants of upward and downward mobility. Then, we examine whether their occupation changes over time by comparing their first and last occupations in the U.S. and analyze the determinants of post-migration occupational mobility for migrants who had at least 5 years of experience in the U.S.

International migrants face many challenges in being incorporated into the labor market at their destination. One issue is the extent to which human capital can be transferred from place of origin to place of destination. More specifically, Mexican migrants without documents, English language skills, or relevant local experience would have more limited access to employment commensurate with their previous work when they arrive in the U.S. Besides individual characteristics, other structural circumstances influence the incorporation of Mexican migrants into U.S. labor markets. For one, migrant workers often group into specific occupations – so-called “immigrant jobs.” Furthermore, migrant networks – both at place of origin and at destination – influence labor opportunities that migrants have access to in the U.S. New migrants are likely to obtain employment in sectors of the economy where their social contacts – and other migrants – are already established. These structural forces work together to create occupational niches defined by national origin or ethnicity (Bohon, 2005; Masseyet al., 1998; Munshi, 2003). It is also important to consider the post-migration occupational trajectories of Mexican migrants and whether their employment opportunities improve as they spend time in the U.S. and gain local experience. Our research questions are as follows: (1) Are Mexican migrants to the U.S. able to remain in the same occupational category that they were in in Mexico; if not, what accounts for their downward (or even upward) occupational mobility? (2) Once in the U.S., what factors contribute to upward and downward occupational mobility of workers who had immigrated from Mexico?

We build on recent work about occupational trajectories in the U.S. and Europe. Helgertz (2013) found that immigrants to Sweden had a lower return on their skills both in terms of occupational status and income. A study of occupational trajectories of Senegalese immigrants in Europe (Obućina, 2013) showed that they experienced downward occupational mobility upon arrival, and that their first job in Europe was a better predictor of their subsequent occupational trajectory than their past occupation in Senegal. Toussaint-Comeau’s (2006) study of occupational assimilation of Hispanic immigrants indicated that initially the loss of immigrants’ wages is greatest in the highest-status occupations, but that for all occupations, that loss decreases with time in the U.S. And among immigrants in Spain, Vidal-Coso and Miret-Gamundi (2014) found that women were more likely than men to experience downward occupational mobility at the time of migration, with only a small proportion able to later move beyond jobs traditionally held by female immigrants such as house cleaning and domestic service.

The Determinants of Occupational Status of Migrants

International migrant workers face important barriers to occupational attainment in the place of destination as the jobs available to them do not necessarily depend on their educational attainment or previous work experience but on the types of jobs where migrants concentrate locally (Amuedo-Dorantes & Mundra, 2007; Rainer & Siedler, 2009). Hagan et al. (2015) find that the unskilled are a heterogeneous category and possess many skills that they can transfer, as migrants, to their place of destination. Since the process of labor market segmentation is closely related to the existence of migrant occupational niches and to the spread of migrant networks, we assume these two theoretical perspectives to influence the occupational attainment of migrants jointly (Vono-de-Vilhena & Vidal-Coso, 2012).

Migration and Social Networks

The social ties connecting relatives, friends or community members in places of origin and destination provide important support for the movement of migrants, goods and information across borders (Masseyet al., 1987). Previous research has found that migrant networks mitigate the costs and risks of migrating and increase the economic benefits of migration. As new migrants arrive in places of destination, they have access to a reliable source of information and job search assistance through migrant networks (Durand, 1994; Massey et al., 1998). These networks include members of the community with current or previous migration experience who can provide economic and logistical assistance to cross the border and then find an appropriate job. The participation of community members and relatives in the migration process can be extensive; it may go from covering the costs of travel and lodging or loaning migrants the money to pay a smuggler, to providing information, references and assistance to obtain a job. Existing research has established that migrants with extensive social networks have access to better paying jobs, and that these positive effects are stronger for undocumented migrants (Aguilera & Massey, 2003; Amuedo-Dorantes & Mundra, 2007; Granberry & Marcelli, 2011; Munshi, 2003; Palloni, Massey, Ceballos, Espinosa & Spittel, 2001). This network of support and assistance is transnational, initially acquired through connections in the home community, then spread out to places of destination, and their influence is in both directions across borders.

The effects of migrant networks are likely to differ by documentation status (Aguilera & Massey, 2003). Undocumented migrants have limited employment opportunities since not all employers are willing to hire them; using migrant networks may improve the types of jobs that undocumented migrants have access to (Kossoudji & Cobb-Clark, 2000; Munshi, 2003).

Despite the positive effects of social networks on migration, other research has found that the use of migrant networks may result in the concentration of migrants in particular sectors of the labor market. When that is the case, the use of networks may negatively impact the probabilities of upward mobility and could increase the risk that migrants end up in lower-status occupations in ethnically dominated sectors of the economy (Portes & Sensenbrenner, 1993; Vono-de-Vilhena & Vidal-Coso, 2012). Since migrants tend to be disproportionately concentrated in low prestige occupations, getting a job through social networks may eventually result in their concentration in so-called “immigrant jobs” and in limited opportunities for occupational mobility (Mahuteau & Junankar, 2008; Portes & Sensenbrenner, 1993; Vono-de-Vilhena & Vidal-Coso, 2012).

In addition to employer preference and access to networks, the structure of the labor market and the concentration of migrants in specific occupations are likely to influence occupational attainment. In the next section we discuss the role of these migrant job niches on the occupational mobility of immigrants.

Migrant Job Niches

When used to understand labor migration processes, the segmented labor market theory posits that the structure of the economy and of the labor market in the place of destination is closely related to the kinds of jobs accessible to migrants. Mexican migrants enter the U.S. labor market at a disadvantage. Some of their limitations are related to not having local work experience, lacking the necessary certifications or training, not having migration documents, and not speaking the local language. As a result, migrants tend to concentrate in secondary and tertiary sectors of the economy and in less stable and less prestigious jobs. Migrants are a source of low-skilled work, and part of their migration strategy may aim at earning higher salaries – relative to their salaries in Mexico – in lower skilled occupations in the U.S., regardless of the loss of prestige and social status (Masseyet al., 1998; Rooth & Ekberg, 2006; Vono-de-Vilhena & Vidal-Coso, 2012). This is particularly true for migrants who do not expect to stay at the destination permanently and are working toward a specific economic goal.

Once a sizeable proportion of migrants from a country or ethnic group are employed in a particular type of job – e.g., domestic or agricultural work –, those migrants who follow them will be more likely to work in the same type of job. A consequence of this occupational concentration is that once an occupational “niche” has consolidated, new migrants will find it harder to obtain jobs in other occupations, and their social networks will place them in minority or ethnicity-dominated jobs (Vidal-Coso & Miret-Gamundi, 2014). As a result of a highly segmented labor market, migrant women and ethnic minorities are more likely to end up in low-prestige occupations in destination countries (Reyneri & Fullin, 2011;Rooth & Ekberg, 2006).

Existing research has documented these limitations. For instance, even when taking into account sociodemographic characteristics and human capital, migrants experience disadvantages in the labor market, particularly in obtaining high-skilled employment (Bernardi, Garrido & Miyar, 2011; Veira, Stanek & Cachón, 2007). This negative effect is especially pronounced for migrant women (and worse for female undocumented migrants) who disproportionately hold jobs in domestic work and care activities, and who have low probabilities of occupational mobility even when they have advanced degrees and training (Barone & Mocetti, 2011;Vidal-Coso & Miret-Gamundi, 2014).

Given the expected effects of the migrant job niches and migrant networks, our first research question asks: Are Mexican migrants to the U.S. able to remain in the same occupational category that they were in in Mexico, and what accounts for their occupational mobility in the U.S.? We hypothesize that recently arrived Mexican migrants to the U.S. are more likely to enter occupations of lower status than the ones they had in Mexico. Moreover, upward mobility would be less likely for those who hold higher status occupations in their place of origin. We also expect these negative effects to be more pronounced for females who, regardless of previous occupation or human capital, concentrate in domestic work or services occupations. Men will likely be concentrated in agricultural, low-skilled, construction, or services occupations.

Occupational Mobility after Migration

Recent migrants face important disadvantages in the labor market in destination countries, mostly due to the lack of opportunities to get employment commensurate with their previous work experience and skills. However, we can expect this negative effect to fade as migrants spend more time in the destination country and become better incorporated into the local labor market, or acquire the necessary skills or resources to get better jobs (Chiswick, Lee & Miller, 2005). This positive effect is particularly pronounced for migrants whose skills are easily transferrable to the local labor market (Akresh, 2006, 2008; Chiswicket al., 2005; Vidal-Coso & Miret-Gamundi, 2014). However, it is also possible that the negative impacts of being in so-called “immigrant jobs” are more permanent, especially for undocumented workers, such that, regardless of years of experience in the U.S. labor market, Mexican migrants will remain in the occupational status they achieved upon arrival. Given these expectations, it is important to investigate whether time spent in the U.S. labor market helps overcome the limitations of a migrant niche labor market or if migrants become stuck in lower-level occupations. Our second research question therefore asks: What are the factors that contribute to upward and downward post-migration occupational mobility?

Data and Methods

We use life history data from the Mexican Migration Project (mmp) to analyze the determinants of occupational mobility between the first and last occupations in the U.S. The mmp collects information from 25,658 households in 154 communities throughout Mexico and the U.S. We select households where the household head or their spouse had any migration experience to the U.S., and use their labor and migration histories to determine the type of occupation they had before migration, after their first migration trip in the U.S., and also in the last year spent working in the U.S. Our analysis takes advantage of this unique source of life history data to pinpoint different moments in the occupational trajectories of Mexican migrants. The life history data also provides us with year-specific measures for most of our indicators.

We select respondents who migrated to the U.S. for the first time after 1965 to exclude migrants during the Bracero Program under which migrants were neither free to select a job nor to choose the state where they wanted to work. We also select only those who migrated to the U.S. after age 15. The analysis further excludes individuals who were unemployed or not in the labor force before leaving Mexico, because there is no initial point of reference for comparison of their occupational attainment.³ Finally, we exclude cases with missing data in the variables of interest.

Dependent Variables

We identify the last occupation held in Mexico in the year before migration; for those not working in that year, we identify the most recent occupation up to the previous 5 years. We then identify the occupations in the first and last years in the U.S. These occupations are classified into the following eight categories:

Professional, managerial or technical;

Skilled worker;⁴

Administrative worker;

Sales and services worker;

Low-skilled worker;⁵

Unskilled construction worker;

Agricultural worker; and

Domestic worker.

The basis for these eight occupational categories is the Clasificación Mexicana de Ocupaciones (inegi, 1996), used by the Mexican Migration Project to code all occupations in the data. We used inegi’s two-digit categories or grupos principales but because of sample size issues and to better reflect substantive occupational changes, we combined some contiguous categories into larger groups. For instance, we joined the separate categories of sales workers and services workers into one category, and professional, managerial and technical workers into another. The category “construction workers” is made up of only unskilled workers (code 546 in inegi 1996); skilled construction workers (code 526 in inegi 1996) are included in the category “skilled workers.”⁶

Using these eight occupational codes, we create two variables that compare the pre- and post-migration occupational status in three categories: 1) upward mobility, 2) lateral mobility (no mobility)⁷, and 3) downward mobility. The first dependent variable compares the last occupation in Mexico to the first occupation in the U.S. The second dependent variable compares the first occupation in the U.S. to the last occupation in the U.S. for those migrants who stayed in the U.S for more than 5 years. Migrants with fewer than 5 years of migration experience in the U.S. were excluded.

Independent Variables

Individual characteristics

We account for individual characteristics such as age, level of education, and union status in the year of reference for each analysis. We also control for household headship and occupation, and stratify the analysis by sex. We expect that older migrants will show higher probabilities of upward occupational mobility.⁸ We control for the highest level of education in the first and last years of U.S. migration in each analysis, respectively, to account for the effects of human capital on the probabilities of having a better job in the place of destination. The variable is classified into four categories: elementary school or less, middle school, high school, and university or more. Our expectations regarding the effects of education are twofold. First, if the labor market in the place of destination is segmented and immigrants are concentrated in occupational niches, education will not have a significant effect on occupational mobility, especially for those who are downwardly mobile. Second, migrants with relatively more education will have higher probabilities of upward mobility as they spend more time in the U.S. and acquire the cultural and social capital necessary to obtain employment commensurate with their educational credentials and previous experience.

The union status variable indicates whether the respondent is in a marital or cohabiting union in the corresponding year. We also include a variable that indicates whether the respondent is the head of household.

Migration characteristics

Our study accounts for the characteristics of the migration trip, including documentation status, period of migration, region of origin in Mexico, region of destination in the U.S., U.S. unemployment rate,⁹ and prevalence of migration in the community of origin in the corresponding year. For the analysis that compares the first and last occupation in the U.S., we further control for the cumulative years of migration experience and the number of trips to the U.S., to account for length and continuity of experience in the U.S. labor market. The combined duration in the U.S. and the number of trips would give us an idea of how consistent or sporadic the experience of the migrant in the U.S. labor market is. For instance, a migrant with a lengthy cumulative duration and a small number of trips had a more established presence in their U.S. job than a worker with a short duration or a worker with many short trips.

Documentation status is a dichotomous variable indicating if the migrant was unauthorized to work in the U.S. in the pertinent year. This category is comprised of undocumented migrants (including those using false documents) and those with tourist/visitor visas, since their visa status does not allow them to work legally in the U.S. Legal status categories include legal residents, temporary workers, refugees, and U.S. citizens. Since unauthorized migrants are more likely to obtain less prestigious jobs and are less likely to translate previous occupational experience into job opportunities in the U.S., we hypothesize that unauthorized migrants will show a significantly higher risk of downward mobility.

We differentiate between three periods of migration, following previous classifications (Durand, 1994; Durand & Massey, 2003; Masseyet al. 2016): (1) 1965-1985, following the end of the Bracero Program, for the period of “undocumented migration” when the U.S. border was fairly open and the risk of apprehension was low; (2) 1986-2001 for the period after the passing of the Immigration Reform and Control Act (irca), representing the peak of Mexico-U.S migration; and (3) the 2002-2015 period to capture the post-9/11 immigration environment characterized by increased concerns for national security and close border surveillance that made undocumented migration more difficult and costly than ever before.

To measure the existence and diffusion of migrant networks in the community of origin, we use year- and community-specific rates of migration prevalence. Migrants from a particular place of origin often concentrate in one or a couple of places of destination, and it is in these places where other members of the community are available to provide assistance to the new migrant (Durand, 1994; Lindstrom & Lopez Ramirez, 2010; Masseyet al., 1987). We calculate the rates of migration prevalence following Lindstrom and López Ramírez’s (2010) methodology. We first excluded individuals and households interviewed in the U.S. and then calculated rates of migration experience at the community level, using information on the dates of first migration to the U.S. for individuals listed in the household roster for each household in the community. We use the rate of male migration prevalence, because relatively few women are migrants.

The denominator for this index of migration prevalence is the number of live men in each year where information is available, and the numerator is the number of men aged 15 years and over with migration experience in the corresponding year for each community. The calculations include a few restrictions: years are included if: 1) there were at least 50 people alive in the community; 2) at least 2 inhabitants with U.S. migration experience existed; and 3) if migration prevalence is higher than 0.01 (Lindstrom & Lopez Ramirez, 2010) can we empirically identify a juncture in the historical development of community-based migration that marks the transition from an initial stage of low levels of migration and gradual growth into a takeoff stage in which the prevalence of migration grows at a more accelerated rate? Second, does this juncture point exist at roughly similar migration prevalence levels across communities? Third, are first-time migrants in the initial stage (pioneers. In the years where these restrictions are not met, the migration prevalence rates are set at 0.

Next, we account for the region of origin in Mexico, classified into the main regions of U.S. migration:

1. 1. Historic region: Durango, Nayarit, Zacatecas, Aguascalientes, San Luis Potosí, Guanajuato, Jalisco, Colima, and Michoacán;
2. 2. Central region: Querétaro, Hidalgo, Estado de México, Distrito Federal, Tlaxcala, Morelos, Puebla, Guerrero, and Oaxaca;
3. 3. Border region: Baja California Norte, Baja California Sur, Sonora, Sinaloa, Chihuahua, Coahuila, Nuevo León, and Tamaulipas; and
4. 4. Southeast region: Veracruz, Tabasco, Chiapas, Campeche, Yucatán, and Quintana Roo.

This variable accounts for the different spread of migrant networks in different regions of the country. For instance, the Historic region of migration has a long tradition of U.S. migration spanning over one hundred years; hence we may expect that migrants coming from this area are more likely to have access to a more established migrant network, compared to migrants from, for example, the Southeast region, where migration has only been widespread in recent decades.

We classify the region of destination in the U.S. into the following categories:

1. 1. Borderland (California, Arizona, New Mexico, and Texas);
2. 2. Great Lakes (Indiana, Illinois, Wisconsin, Michigan, and Ohio);
3. 3. Northwest (Washington, Oregon, Idaho, Nevada, and Utah);
4. 4. Great Plains (Montana, Wyoming, South Dakota, North Dakota, Minnesota, Colorado, Nebraska, Kansas, Oklahoma, Missouri, and Iowa);
5. 5. Southeast (Maryland, DC, West Virginia, Virginia, South Carolina, North Carolina, Georgia, and Florida);
6. 6. South (Arkansas, Louisiana, Mississippi, Alabama, Tennessee, Kentucky); and
7. 7. Northeast (Maine, Vermont, New Hampshire, Massachusetts, Rhode Island, Connecticut, New York, New Jersey, Pennsylvania, and Delaware).¹⁰

In the same way as networks are more likely to be widespread in some Mexican states, some destination regions have more developed migrant labor markets and migrant networks. We also expect that the more popular destination regions such as the Borderland and the Great Lakes will be those where migrants are more concentrated in ethnic niches and where migrant networks may result in more limited occupational choices; as a result, they may be less likely to achieve upward occupational mobility.

To further account for the effect of migrant concentration in specific labor niches, our models include a few dummy variables that indicate whether the migrant works in an “immigrant job” in the U.S. (i.e., a job where Mexican migrants are highly concentrated). For males these two categories are agricultural work and sales/services; for females they are domestic work and sales/services. These indicators will help us estimate whether downward mobility and lateral mobility are associated with being employed in jobs associated with Mexican migrants.

Method

The first part of the analysis compares the last job in Mexico to the first job in the U.S. We first present two-way tables to compare the last occupation in Mexico and the first occupation in the U.S. for men and women. We then estimate a multinomial logistic regression to determine the relative probability of achieving upward or downward mobility, compared to staying in the same occupational category, controlling for the characteristics listed above. The models are stratified by sex and use Huber-White robust standard errors to account for clustering at the community level.

The second part of the analysis compares the first and last occupation in the U.S. for migrants who have more than 5 years of accumulated experience in the U.S. We use this cut-off point to exclude one-time and temporary migrants who may not be seeking to change jobs or improve their occupational status because they may not plan to work in the U.S. in the long term; this selection also excludes those migrants who may not yet be at risk of mobility since they have not spent a significant amount of time in the country. We also exclude those who did not get a job upon arrival in the U.S. because they have no initial point of comparison. The analysis presents two-way tables to compare first and last U.S. occupation by sex. We estimate multinomial logistic regression models for the relative probability of occupational mobility. As in the first analysis, we account for the covariates presented above, though in this case we also control for cumulative years of experience in the U.S. and the number of U.S. trips for each respondent to account for the labor experience acquired in the U.S. The models are stratified by sex and use robust standard errors. For detailed characteristics of the samples for each analysis, refer to Table 1 in Appendix 1.

Findings

Occupational Mobility in the First U.S. Trip

Regarding the distribution of occupations (see Table B in Appendix 1 for figures), almost half of all male migrants were agricultural laborers in Mexico, but only 30% worked in that occupation in the U.S. as their first job. About the same percentage of men were skilled workers both in Mexico and in the U.S. Men in the U.S. were twice as likely, after considering other characteristics, to be sales/services workers than they were back in Mexico, and the overall size of the “low-skilled” category in this sample was about 50% larger than had been the case in Mexico. Few male migrants were listed as unemployed in the first year of migration to the U.S. The last jobs held by female migrants in Mexico were mostly as sales and services workers (28%), skilled workers (25%), and domestic workers (16%), with 8% having been professionals.¹¹ Once in the U.S., close to one-third were either unemployed or not in the labor force. Domestic service was the only category that employed similar proportions of female migrants in both Mexico and the U.S., even with the high proportion of females not working in the U.S. As in the case of males, only a small share of female migrants with a professional occupation in Mexico was able to obtain a professional job once in the U.S.

We analyzed the probabilities of being unemployed or not in the labor force for female migrants at the time of arrival in the U.S. Of the 173 female migrants without a reported occupation (32.8% of all female migrants), over three-quarters became homemakers or helped out in the household without pay; only 5% were either unemployed or became students.

Tables 1 (males) and 2 (females) show the flow from the last Mexican occupation to the first U.S. occupation. For males, migrants who remained in the same occupation type were the largest group in the categories of skilled, sales/services, low-skilled, and agricultural workers (see the diagonal of Table 1). But the results also show that only 11% of all migrants who were professionals in Mexico were able to maintain that occupational status, and only 10% of administrative workers. The category of construction workers stands out: it is the only category where a larger proportion ended up in a higher-status occupation (skilled with 32%). Female migrants, regardless of their occupation in Mexico, were more than likely not working once they arrived in the U.S. (with the exception of those in skilled employment and construction work, which is a miniscule number of women, and domestic work). Women who entered the labor force in the U.S., in general, were most likely to work in the occupational category that they reported for the last job in Mexico, except for professionals; as is the case for males, female professionals in Mexico were generally unable to transfer their credentials when they entered the U.S. labor force.

The occupation matrices in Tables 1 and 2 show three mobility outcomes: lateral mobility (the diagonal cells); downward mobility (the cells above the diagonal); and upward mobility (the cells below the diagonal). Because several of the adjacent occupational categories are unlikely to involve significant mobility, we treated occupational change between the following sets of occupations as lateral mobility: professionals and administrative workers; skilled and sales/service workers; low-skilled; and construction, agricultural, and domestic workers. While the last group includes three categories, Tables 1 and 2 showed that virtually no male migrants are domestic workers and virtually no female migrants are construction workers.

Overall, 25% of male migrants but over half of female migrants experienced downward occupational mobility from their last job in Mexico to their first job in the U.S. (Tables 1 and 2). In contrast, 36% of males but only 13% of females had upward occupational mobility following migration. Examining the sources of mobility of male migrants showed that 65% of downwardly mobile men formerly had skilled and sales/services occupations in Mexico. In contrast, 77% of upwardly mobile migrants had been agricultural workers in Mexico; 76% of those who moved between comparable occupational categories were skilled and agricultural workers in Mexico.

The lower occupational mobility of female migrants is largely due to the high proportion exiting the labor force after migration to the U.S. Downwardly mobile female migrants had mostly professional (14%), skilled (28%), and sales/services (33%) occupations in Mexico. Almost three-quarters of upwardly mobile female migrants had been domestic and agricultural workers. Females who were in comparable occupational categories in both Mexico and the U.S. were skilled (29%), sales/services (30%), and domestic (42%) workers in Mexico.

Table 3 presents results of the multinomial logistic regression. Results show important sex differences in the effects of the independent variables on downward and upward mobility. Being in a union increased the odds of downward mobility and decreased the odds of upward mobility for males but had no effect for females. A higher level of education is related to higher odds of downward mobility and lower odds of upward mobility for men, whereas it does not have the same significance for female migrants. The strongest effect for women is most concentrated in the increased odds of downward mobility for university educated women. These findings suggest that Mexican migrants are generally unable to transfer their credentials from Mexico to the U.S., especially males and more educated women.¹²

The effect on mobility of having a so-called “immigrant job” in the U.S. depends on the type of job. Following the trends observed in tables 1 and 2, working in agriculture (for males) and domestic service (for females) decreases the chances of upward mobility, (given the concentration of males in agriculture and females in domestic service, the effects are quite large). This effect is an example of how migrant occupational niches may sort migrant workers in specific occupations where other Mexican migrants also concentrate. However, obtaining a sales or services occupation in the U.S. had a negative effect on downward mobility for both sexes and positive effect on upward mobility for female migrants. Being an unauthorized migrant worker, contrary to expectations, showed no effects on mobility for female migrants and only increased the chances of upward mobility for male migrants, but this effect is only marginally significant. Compared to the 1965-1985 period, those who migrated after the irca (1986-2001) were less likely to experience upward mobility, though the effect for women is only marginally significant; males who migrated during 1986-2001 were also less likely to be downwardly mobile, albeit with marginal significance.

The effects of region of destination on occupational mobility differ by sex and outcome. Male migrants in the Great Lakes region, when compared to the U.S. Borderland region, are less likely to experience downward mobility. Female migrants, on the other hand, have greater odds of being downwardly mobile if they move to the Northwest and lower odds if the Northeast is their destination. But the South and Northwest as destination lowers the odds of female migrants being upwardly mobile, whereas females are more likely to experience upward mobility if they migrate to the Northeast.

Finally, male migrants who were originally in the Southeast and Central regions of Mexico (as compared to the Historic region) had a lower risk of downward mobility (albeit only at the 0.1 level of significance); for males, originating from the region also increased their chances of upward mobility. Contrary to expectations, we find that the US. unemployment rate is only marginally related to an increased odds of downward mobility for men. Our social networks proxy, the prevalence of migration in the community of origin, is significantly associated with increased odds of upward mobility for both sexes.

Occupational Mobility in the Last Year in the U.S.

We now turn to the analysis of occupational mobility between the first and the last occupation held by Mexican migrants with more than 5 years of accumulated experience in the U.S. The main change at their last occupation was an increase of males working in skilled occupations and a decrease in those working as agricultural workers. Similarly, three-fourths of female migrants were also employed in three occupation types – skilled, sales/services, and domestic work – in their first job in the U.S. At the time of their last reported job, 47% of female migrants continued to work in sales/services and skilled occupations, but the two main changes in the distribution of occupations between the first and last year in the U.S. for females is the decrease in the share of workers reporting a skilled occupation, and the substantial proportion (24%) being either unemployed or not in the labor force. Almost all women (95%) who did not report an occupation in the last year in the U.S. were homemakers, and none reported being unemployed (see Table C in Appendix 1 for more details on these distributions).

Compared to the occupational mobility from the last job in Mexico to the first U.S. job, both males and females were far more likely to remain in their occupational category between their first and last reported job in the U.S. (see the large proportions in the diagonals in tables 4 and 5). While only 40% of male migrants remained in their occupation after the move from Mexico to the U.S., 57% did so from their first to their last job in the U.S. The corresponding figures for female migrants are 34% and 62%. During their work in the U.S., male as well as female migrants experienced both less downward and less upward mobility when compared to the occupational change between their last job in Mexico and first job in the U.S. But the sex differential in occupational mobility remained: whereas men were more likely to be upwardly mobile than to experience downward mobility both for the shift from the last Mexican job to the first job in the U.S. and for the shift from the first to the last U.S. job, the opposite is the case for women for both mobility measures.

For men, skilled and sales/services workers accounted for 73% of those who were downwardly mobile from first to last occupation in the U.S. Of all men who experienced upward mobility, 59% had been agricultural workers and 33% low-skilled workers in their first U.S. job. For women, skilled and sales/services workers were also the two largest occupational categories (69%) for those who experienced downward mobility. Regarding women’s upward mobility, half of female migrants were domestic workers at their first U.S. job, and another 32% were low-skilled workers.

We present the results of the multinomial logistic regression for first and last occupation in the U.S. in Table 6. They show that occupational mobility of both males and females in the U.S. is associated with different factors than is mobility from last Mexican job to first job in the U.S. The statistical significance of the effects both of being a household head and of marital status is lower between first and last job in the U.S. But for mobility in the U.S., education does not matter as much as it did for the first job in the U.S.: having a university education or more increases the likelihood of downward mobility for female migrants only. For males, there is change in the effects of the individual characteristics on the likelihood of mobility: compared to the mobility from the last Mexican job to the first U.S. job, being in a marital or cohabiting union no longer increases the odds of downward mobility.

Having an “immigrant job” has outcomes similar to those for mobility from the last Mexican to the first U.S. job: for males, those who were an agricultural worker in the last job in the U.S. are more likely to have experienced downward mobility and less likely to have experienced upward mobility; the same holds for females who are domestic workers in their last U.S. job. But the effects on mobility of being a sales/services worker in the last U.S. job differ from the Mexico-to-U.S. mobility experience. In the U.S., it decreases the chances of upward mobility for males; for females, they were more likely to have experienced both downward mobility and upward mobility (at the 0.1 level of significance).

Our findings further showed that once in the U.S., documentation status has no effect on occupational mobility; the same holds for period of migration.

Region of destination continues to have limited effect on the occupational mobility of males: migrants going to the Great Plains and the South are more likely to experience downward mobility, while those in the Great Plains and the Southeast (at the 0.1 level) have a higher probability of experiencing upward mobility. Women whose last job in the U.S. was in the Northwest and Great Plains are less likely to have experienced downward mobility (compared to those in the Borderland), whereas being in the Great Lakes region reduced their chances of having been upwardly mobile and being in the Northwest increases them, although these effects are only marginally significant. Region of origin in Mexico is only relevant for occupational mobility during the time spent in the U.S. for migrants coming from the Central region and females from the Southeast, who are less likely to be upwardly mobile.

Migration prevalence in the community of origin has no effect on occupational mobility, except for female migrants who are less likely to be upwardly mobile if they come from communities of high migration. The unemployment rate in the U.S. substantially lowers the odds of males becoming upwardly mobile during their stay in the U.S. The number of trips to the U.S. and the years of migration experience in the U.S. both have no bearing on occupational mobility of female migrants. But for male migrants, U.S. trips are negatively associated with downward mobility and positively related to upward mobility, and spending more years in the U.S. is related to higher odds of upward mobility.

Conclusions

In this paper we analyze the occupational mobility of Mexican migrants to the U.S., using data from the Mexican Migration Project. Occupational attainment of migrants is determined by their own individual characteristics, but also by the circumstances of their trip. Migrants can make use of social networks to access better jobs upon arrival at destination; however, their ability to take advantage of their human capital in the labor market of the destination country limits their choices as it prescribes specific occupational niches where they can enter the labor market. Previous research found that migrants are likely to end up in the lower ranks of the occupational ladder even after accounting for their previous occupational experience. Although the expectation is that downward occupational mobility will diminish as migrants spend more time at the destination, evidence has been mixed on the ability of migrants to move beyond ethnic occupational niches.

Our findings are consistent with the existence of migrant occupational niches and relatively low opportunities for occupational mobility, especially for women. Overall, most men experience lateral mobility upon arrival in the U.S., but more than half of women end up in lower status jobs. Downward mobility for women is mainly driven by those in higher status occupations getting lower status jobs in the U.S. Further, this occupational sorting remains as migrants spend more time in the U.S. The findings show that both male and female Mexican migrants to the U.S. encounter barriers to transferring their credentials from Mexico to the U.S., but these barriers are greater for women.

Consistent with previous research, we find that university educated women are more likely to experience downward mobility in their first U.S. job than those with less education. For men, education is related to higher probabilities of downward mobility and lower odds of upward mobility, which is consistent with the argument that immigrants are not able to translate previous skills and experience into an appropriate job in the U.S. In addition, we find support for the impact of occupational niches. Being in jobs where migrants concentrate, or so-called “immigrant jobs,” is related to lower probabilities of upward mobility. This concentration of migrant workers in specific niches is consistent and is independent of documentation status. Taken together, these findings help paint a picture of a rigid occupational structure for Mexican migrants in the U.S. where placement in occupational niches is a stronger determinant of occupation than human capital or documentation status (which only marginally increased the odds of upward mobility for male migrants). It is possible that migrants are unable to translate human capital into occupational status, either because they do not speak English or because they do not have the necessary certifications to work in their area of expertise.

Following our expectation, the rate of migration prevalence in the community of origin, which we use as a proxy of the spread of migrant networks, is positively associated with the log odds of upward mobility, while the rate of unemployment had no significant effects (it is only marginally related to higher odds of downward mobility for men). However, our indicators for region of origin do affect occupational mobility. Male migrants from non-traditional sending communities are less likely to be downwardly mobile. These findings provide indirect evidence that more developed migrant networks of persons from traditional areas of origin may be associated with migrants entering an ethnic niche instead of jobs they are better suited for. This is consistent with previous research that found that even though migrant networks aid in finding employment, they tend to result in lower status jobs, disproportionately affecting migrants with higher human capital.

The dynamics of occupational mobility from first to last job in the U.S. differ in some respects from occupational mobility regarding the last Mexican and first U.S. jobs. Once in the U.S., educational attainment increases female migrant’s odds of downward mobility. Thus, while migrants could not transfer their human capital from Mexico to the U.S., some were eventually able to have their previous qualifications and experience applied in the U.S. (about 12% of men and 10% of women). This finding is unrelated to the time spent in the U.S. The effect of working in a migrant niche on occupational mobility continued in the U.S.: it increased the odds of downward mobility and decreased the odds of upward mobility. Coming from a non-Historic region of origin in Mexico was no longer positively associated with upward mobility; in fact, migrants originating from the Central region and females from the Southeast had lower odds of upward mobility than those from the Historic region of origin. In sum, the main finding of our analysis is (a) the inability of migrants to transfer their human capital and occupational status from Mexico to the U.S.; and (b) the effect of occupational niches in the U.S. which increase downward mobility and lower the odds of upward mobility. Thus, niche jobs might provide employment, but they do so at a cost to migrants in terms of occupational status.

Our analysis has a few limitations. First, we are not accounting for direct measures of migrant network resources. While the mmp data includes questions on the use and the nature of migrant networks, these are only available for the most recent trip of the household head to the U.S. and not for the entire migration history. However, we use widely accepted proxies to measure their impact. Second, we are not taking into consideration that a lateral or even downward occupational move may still entail increased wages for Mexicans in the U.S. Considering differences in both wages and standard of living between the two countries, migrants could earn a much higher real salary (i.e., controlling for cost-of-living differentials) for the same job in the U.S. than they would in Mexico. Future work should explore these salary differentials and how they may offset the loss of status and downward mobility. Lastly, although we are able to analyze the impact of union status on occupational mobility, we cannot control for the nationality of the spouse of the individual, as numbers of non-Mexican spouses are very small in our sample.

References

Aguilera, M. B., & Massey, D. S. (2003). Social Capital and the Wages of Mexican Migrants: New Hypotheses and Tests. Social Forces, 82(2), 671-701. doi: http://doi.org/10.1353/sof.2004.0001.

Akresh, I. R. (2006). Occupational Mobility Among Legal Immigrants to the United States. International Migration Review, 40(4), 854-884. doi: http://doi.org/10.1111/j.1747-7379.2006.00046.x.

Akresh, I. R.(2008). Occupational Trajectories of Legal US Immigrants: Downgrading and Recovery. Population and Development Review, 34(3), 435-456. doi: http://doi.org/10.1111/j.1728-4457.2008.00231.x

Amuedo-Dorantes, C., & Mundra, K. (2007). Social Networks and Their Impact on the Earnings of Mexican Migrants. Demography, 44(4), 849–863. doi: http://doi.org/10.1353/dem.2007.0039

Barone, G., & Mocetti, S. (2011). With a Little Help from abroad: The Effect of Low-skilled Immigration on the Female Labour Supply. Labour Economics, 18(5), 664–675. doi: http://doi.org/10.1016/j.labeco.2011.01.010

Bernardi, F., Garrido, L., & Miyar, M. (2011). The Recent Fast Upsurge of Immigrants in Spain and Their Employment Patterns and Occupational Attainment. International Migration, 49(1), 148-187. doi: http://doi.org/10.1111/j.1468-2435.2010.00610.x

Bohon, S. A. (2005). Occupational Attainment of Latino Immigrants In the United States. Geographical Review, 95(2), 249-266. doi: http://doi.org/10.1111/j.1931-0846.2005.tb00365.x

Chiswick, B. R., Lee, Y. L., & Miller, P. W. (2005). A Longitudinal Analysis of Immigrant Occupational Mobility: A Test of the Immigrant Assimilation Hypothesis1. International Migration Review, 39(2), 332-353. doi: http://doi.org/10.1111/j.1747-7379.2005.tb00269.x

Durand, J. (1994). Más allá de la línea. Patrones migratorios entre México y Estados Unidos. Ciudad de Mexico: Consejo Nacional para la Cultura y las Artes.

Durand, J. & Massey, D. S. (2003). Clandestinos. Migración México-Estados Unidos en los albores del siglo xxi. Ciudad de México: Universidad Autónoma de Zacatecas and Miguel Ángel Porrúa.

Granberry, P. J., & Marcelli, E. A. (2011). Social Capital is Associated with Earnings among Foreign-born Mexican Men but not Women in Los Angeles County. International Migration, 49(6), 113-128. doi: http://doi.org/10.1111/j.1468-2435.2011.00710.x

Hagan, J., Hernandez-Leon, R., & Demonsant, J.-L. (2015). Skills of the “unskilled”: Work and Mobility Among Mexican Migrants. Oakland: University of California Press.

Helgertz, J. (2013). Pre- to Post-Migration Occupational Mobility of First Generation Immigrants to Sweden from 1970–1990: Examining the Influence of Linguistic Distance. Population Research and Policy Review, 32(3), 437-467. doi: http://doi.org/10.1007/s11113-013-9274-9

Instituto Nacional de Estadística Geografía e Informática (INEGI) (1996). Clasificación Mexicana de Ocupaciones (cmo) - Histórica, vol. 1. Mexico.

Kossoudji, S. A., & Cobb-Clark, D. A. (2000). IRCA’s Impact on the Occupational Concentration and Mobility of Newly-Legalized Mexican Men. Journal of Population Economics, 13(1), 81-98. doi: http://doi.org/10.1007/s001480050124

Lindstrom, D. P., & Lopez Ramirez, A. (2010). Pioneers and Followers: Migrant Selectivity and the Development of U.S. Migration Streams in Latin America. The ANNALS of the American Acadamy of Political and Social Science, 630(July), 53-77. doi: http://doi.org/10.1177/0002716210368103

Mahuteau, S., & Junankar, P. N. (Raja). (2008). Do Migrants get Good Jobs in Australia? The Role of Ethnic Networks in Job Search. Economic Record, 84, S115–S130. doi: http://doi.org/10.1111/j.1475-4932.2008.00488.x

Massey, D. S., Alarcon, R., Durand, J., & Gonzalez, H. (1987). Return to Aztlan: The Social Process of International Migration from Western Mexico. Berkeley: University of California Press.

Massey, D. S., Arango, J., Hugo, G. J., Kouaouci, A., Pellegrino, A., & Taylor, J. E. (1998). Worlds in Motion: Understanding International Migration at the End of the Millennium. Oxford: Oxford University Press.

Massey, D. S., Durand, J., & Pren, K. A. (2016). Why Border Enforcement Backfired, American Journal of Sociology, 121(5), 1557-1600.

Munshi, K. (2003). Networks in the Modern Economy: Mexican Migrants in the U. S. Labor Market. The Quarterly Journal of Economics. 118(2). doi: http://doi.org/10.1162/003355303321675455

Obućina, O. (2013). Occupational Trajectories and Occupational Cost among Senegalese Immigrants in Europe. Demographic Research, 28(19), 547-580. doi: http://doi.org/10.4054/DemRes.2013.28.19

Palloni, A., Massey, D. S., Ceballos, M., Espinosa, K., & Spittel, M. (2001). Social Capital and International Migration: A Test Using Information on Family Networks. American Journal of Sociology, 106(5), 1262-1298.

Portes, A., & Sensenbrenner, J. (1993). Embeddedness and Immigration: Notes on the Social Determinants of Economic Action. American Journal of Sociology, 98(6), 1320-1350. doi: http://doi.org/10.1086/230191

Rainer, H., & Siedler, T. (2009). The role of social networks in determining migration and labour market outcomes. Evidence from German reunification. Economics of Transition, 17(4), 739-767. doi: http://doi.org/10.1111/j.1468-0351.2009.00365.x

Reyneri, E., & Fullin, G. (2011). Labour Market Penalties of New Immigrants in New and Old Receiving West European Countries. International Migration, 49(1), 31–57. doi: http://doi.org/10.1111/j.1468-2435.2009.00593.x

Rooth, D.-O., & Ekberg, J. (2006). Occupational Mobility for Immigrants in Sweden. International Migration, 44(2), 57-77. doi: http://doi.org/10.1111/j.1468-2435.2006.00364.x

Toussaint-Comeau, M. (2006). The Occupational Assimilation of Hispanic Evidence from Panel Data. International Migration, 40(3), 508-536. doi: http://doi.org/10.1111/j.1747-7379.2006.00034.x

Veira, A., Stanek, M., & Cachón, L. (2007). Los determinantes de la concentración étnica en el mercado laboral español. Revista Internacional de Sociología, 69(m1), 219-242.

Vidal-Coso, E., & Miret-Gamundi, P. (2014). The Labour Trajectories of Immigrant Women in Spain: Are There Signs of Upward Social Mobility? Demographic Research, 31(13), 337-380. doi: http://doi.org/10.4054/DemRes.2014.31.13

Vono-de-Vilhena, D., & Vidal-Coso, E. (2012). The impact of informal networks on labour mobility: Immigrants’ first job in Spain. Migration Letters, 9(3), 237-247.

Notes

Author notes