Social and Family Networks
This chapter summarizes the workshop discussions on the importance of social and family networks for immigrants and their success in this country.
The statistics collected by the government on immigration are limited for many types of social science research. In particular, when migration is viewed as a process, the information contained in administrative records does not include the relevant data. Empirical work suggests that transnational migration unfolds with certain regularities over time. Following the initiation of emigration from a community to another country, migration has a well-defined tendency to become more prevalent and to broaden its social and economic base in the sending community. At the same time, migration tends to focus more selectively on a narrow range of jobs and destinations in the receiving country.
Such regularities have been evidenced by migration from Mexico to the United States over the past two decades. The migratory trends follow from the fact that migration affects individual motivations and social structures in both the sending and receiving countries, in ways that stimulate further migration. Transnational migration is a self-reinforcing process that develops internal momentum. As a result, migration becomes increasingly independent of the condi-
tions that originally caused it, leading to common empirical features across diverse communities.
Transnational migration may originate for a variety of reasons. Potential immigrants may observe wage differentials or employment prospects that differ between origin and destination areas. Households may seek to diversify risks to their economic well-being by sending some family members to work in different regional or foreign labor markets. Immigrants may be recruited by foreign employers seeking to import workers for specific tasks, or they may be impelled to move because structural transformations in the local economy eliminate traditional sources of employment. Migration may also be promoted by a variety of noneconomic factors, including threats to physical well-being by ethnic and religious persecution, civil violence, and local or regional environmental conditions (such as famine or drought).
No matter how transnational migration begins, the first immigrants from a community are likely to experience it as a costly and risky enterprise, both in monetary and psychological terms. They have little knowledge of the conditions in the host country and are largely ignorant of its culture, language, and employment practices. In most cases, they incur the expenses of the trip and absorb the opportunity costs of foregone income while moving and looking for work. Lacking knowledge about prevailing wage rates, work habits, legal conventions, and social expectations, these first immigrants are vulnerable to exploitation and mistreatment, particularly if they are illegal aliens.
There is a growing amount of empirical research literature on the migration process, reporting on Mexican emigration (Massey et al., 1987). Most emigration begins with people in the lower-middle and middle ranges of the socioeconomic scale, generally married men, who emigrate after recruitment. Such people have enough resources to absorb the costs and risks of the trip but are not so affluent that the idea of moving to another country to work is unattractive. The first immigrants are usually in their prime years of labor force participation. Men are better able than women to absorb the physical risks of emigration, and married men have a family support network that can offer some foundation for them while they search for employment away from home.
The earliest emigrants leave their families and friends and strike out for solitary work in a foreign land. Most of the first solitary workers are target earners, who seek to earn as much money as quickly as possible in order to recoup their initial investment, attain a predetermined income goal, and return home. They have little interest in permanent settlement abroad. Once more people have come and gone in this fashion, the situation in the sending community begins to change. Each migratory act generates a set of irreversible changes in individual motivations, social structures, cultural values, and knowledge of the destination community that alters the context within which future migration decisions are made. These changes accumulate to create conditions that make additional migration more likely.
At the individual level, participation in a high-wage economy encourages changes in tastes and motivations that turn people away from target earnings toward longer-term migration. The firsthand experience gained from migration makes the satisfaction of new wants increasingly feasible. Someone who has migrated and returned home has direct knowledge of employment opportunities, labor market conditions, and ways of life in the destination country and can use these understandings to migrate again with fewer risks and costs than before. Once experienced, migration becomes familiar and a reliable resource that can be used again as new needs arise and motivations change. From empirical studies in Mexico, it is clear that the more often a man migrates, the more he is likely to continue migrating (Massey, 1985; Massey et al., 1987).
During the first trips, immigrants tend to live under spartan conditions, sharing apartments with other men and sleeping in shifts to save money. They work long hours, have little social life, and repatriate most of their earnings. Immigrants see themselves as members of their home communities and not as participants in the destination community.
As immigrants spend more time abroad and as the number of their trips to the destination country increases, the pressure to bring along family members grows. The first relatives to accompany an immigrant are typically unmarried sons of working age, because they have the greatest earning potential and migration is consistent with their sex roles. Over time, unmarried working-age daughters are likely to join as well, and finally wives and younger children. As a result, the demographic base of migration steadily widens and the mean age of migration drops.
Not only does the act of migration induce changes in individual immigrants and their family, but it also changes social structure in ways that spread migration within the originating community. Each immigrant is linked to a group of nonimmigrants through a set of social ties based on friendship and kinship. Nonimmigrants draw on these social ties to gain access to employment and assistance abroad, substantially reducing the costs and risks of movement experienced by the earlier immigrants. Thus, the demographic base broadens, as many in the community begin to emigrate. A network develops as the community of immigrants develops ties at both the origin and the destination, the network offering a form of social capital.
Each new immigrant reduces the costs and risks of migration for a set of friends and relatives and, with these lowered costs and risks, some of them are induced to emigrate, which further expands the set of people with ties abroad, which in turn reduces costs and risks for a new set of people, causing them to emigrate. Once the number of network connections reaches a critical threshold, migration becomes a self-perpetuating phenomenon because each act of movement creates the social structure necessary to sustain additional migration.
If the process of migration continues long enough, eventually networks reach a point of saturation. Larger and larger shares of the transnational community
reside in the foreign branch communities, and virtually all of the people in the home community are connected to someone living abroad or to someone with substantial immigrant experience. When networks reach this point of development, the costs of immigration stop falling with each new immigrant and the process of immigration loses its dynamism. The rate of entry into the immigrant labor force begins to decelerate, perhaps trailing off, at this point.
In a sample of Mexican communities studied by Massey and his colleagues, the prevalence of ever having been in the United States for the population in each community suggests that emigration began with men in the early 1900s, and some women began to go to the United States after about 1920. The pattern of emigration can be described in groups of low, middle, high, and mass movement (mass movement is defined as U.S. experience by 30 percent or more of the residents of a Mexican community).
The proportion of Mexicans with U.S. experience or with relatives with such experience increases dramatically with the transition from low to mass emigration. The diversity of immigrants with regard to age, sex, and household membership also increases. Geographic diversity first increases, then decreases with the low to mass transition. Diversity in trip duration, legal status, strategy, and occupation increases with the transition. At mass levels of emigration, some emigrants have become foremen with better connections who can hire members of their social network. This narrows both geographic diversity and the targeting of the emigrants. In terms of social class, emigration starts in the middle of the socioeconomic scale and broadens. At mass levels of emigration, many more immigrants are from agricultural backgrounds.
Based on the study of Mexican communities, Massey and his colleagues report considerable evidence of common patterns of migration that are in line with the empirical observations of other investigators. These results suggest that common migration processes occur across a wide range of Mexican communities, yet their expression is shaped by structural factors operating at the aggregate level. The next step of research in this area is to study the structural factors and population processes that govern the spread of migration within communities, seeking to understand why some communities rapidly attain a state of mass emigration, whereas others develop more slowly and achieve only modest rates of emigration.
Immigration and the Family2
Family reunification and formation are the cornerstone of U.S. immigration law. A large proportion of visas are issued on these grounds. For an immigrant,
the strategy of immigration depends on family. The family also affects immigration after arrival. And the family affects the children of immigrants. In short, the family is key to the study of immigration.
It is useful to highlight some ways in which the family matters for migration. In order to gain permanent residence, an immigrant needs a sponsor, who is often a relative. Another way to gain permanent residence is to marry a U.S. citizen—which might suggest higher rates of marital dissolution among marriages entered into for immigration purposes. In cases of marriage and naturalization, naturalization becomes an important part of family dynamics. The children of immigrants are important: there is a consistent finding that immigrants invest more in schooling for their children than comparable native-born parents. And U.S. citizens living abroad sometimes marry local residents and thereby establish a chain of potential new immigrants.
Immigrant success, and hence the effects of immigration on the U.S. economy and society, depends on who the immigrants are and thus how they are selected. It is not easy to obtain legal permanent residence in the United States. Except for refugees, legalized aliens, and the new ''lottery'' immigrants (i.e., immigrants who gain a visa based on the random selection of applicants from selected countries), the vast majority of adult immigrants require a sponsor. Under current U.S. law, the prospective immigrant has entitlement to an immigrant visa through kinship or occupational ties to a U.S. citizen or resident, who in turn verifies the relationship and petitions for the immigration of the relative or employee. The familial route requires that the prospective immigrant either already has or acquires the "right" type of relative. Appropriate U.S.-citizen relatives include parents, adult offspring, siblings, and spouses.
One of the most important policy debates about the criteria of U.S. immigration law concerns the degree of importance that should be accorded family reunification principles, which confer admission advantages for blood relatives of U.S. residents or citizens, over criteria that emphasize an individual's potential economic contribution. The fact that the economic success of immigrants, as measured by their earnings, may be importantly related to the route by which they migrated, as indicated by whether or not they had a family sponsor, is at the heart of the policy debate about immigrant selection criteria. Existing data, however, provide limited information for analysis of this important policy question. Questions on large surveys, such as the Current Population Survey and the decennial census, on the type of visa of the legal immigrant at entry, indicating sponsorship, could help predict immigrant economic success and provide important information on the determination of admission criteria.
Emigration by immigrants—the return of foreign-born people to their country of origin—is a neglected phenomenon in immigration literature, due mostly to the lack data (Warren and Kraly, 1985). Yet emigration is an important measure of assimilation, and it importantly affects the net impact of immigration. Analysis of data for the period 1960 to 1980 suggests that emigration is substan-
tial—up to 1.2 million permanent resident aliens may have left the United States between 1960 and 1969, and up to 1.9 million between 1970 and 1979 (Jasso and Rosenzweig, 1990:124–138). Thus, more than 40 percent of immigrants may have left the United States in the 1960 to 1979 period. Moreover, emigration appears to be selective: immigrants from Canada and Europe appear to have the highest propensity for leaving the United States, and people from Asia the lowest, with Latin American and Caribbean immigrants in an intermediate position. As a consequence, data on gross immigration flows, which are used in most analyses of immigration, tend to overstate the relative growth in the contribution of immigrants, especially the proportions of European-origin immigrants. Net immigration figures display a much greater growth in the volume of Asian immigration than of European-origin immigration in recent years than do gross immigration data.
Census-based data on the foreign-born suggest that emigration, besides being selective with respect to country of origin, is also selective with respect to age, sex, and schooling (Jasso and Rosenzweig, 1990:138–144). Emigration is higher for men, for example, than for women. But without regular longitudinal data on immigrants and without information on the role of the family in both the United States and the country of origin, it is not yet possible to measure with much precision the magnitude of emigration or to predict who among immigrants will leave.
Marriage to a U.S. citizen is a potential route to immigration for someone without the requisite skills or blood relatives in the United States or for someone from a country from which the number of qualified visa applicants exceeds country ceilings. And, indeed, as U.S. immigration law has become more restrictive in terms of both family reunification and skill criteria, the importance of international marriage as a route to immigration has increased. Analysis of data for the early 1980s provides a clear conclusion (Jasso and Rosenzweig, 1990): international marriage provides a way to achieve emigration to the United States when it would otherwise be quite difficult. More refined analysis of the relationship of spouses of foreign-born U.S. citizens, the timing of the marriage, and of subsequent marital disruption requires data that are not currently available.
The decision to naturalize is also intimately related to family matters. One of the most important rights conferred by naturalization is the right to sponsor the immigration of immediate and close family members; hence, naturalization may be particularly attractive to individuals seeking to bring in a spouse, parent, or other relative. The attractiveness of naturalization will thus depend on the visa status of the immigrant and on the number and type of relatives residing in the country of origin. The ideal way to study the naturalization of immigrants is, first, to obtain a family history that includes information on the location of the immigrant's parents and siblings and, second, to track the immigrants and those family members over time. Other pertinent evidence could be obtained by following a cohort of immigrants over time, using available administrative records
(such a study was carried out for a 1971 cohort of immigrants; see Jasso and Rosenzweig, 1990:110–120).
In the long run, perhaps the greatest impact of immigration is the investments that immigrants make in their children. Workshop participants noted that, historically, the children of immigrants have been higher achievers than the children of native-born parents. Data from the 1960 and 1970 censuses show that native-born children of foreign-born parents have higher levels of schooling and achieve higher earnings than the native-born children of native-born parents. Unfortunately, the 1980 and 1990 censuses did not collect information on the nativity of parents, so these census data cannot be used to study this phenomenon for recent cohorts of the foreign-born.
A second long-run impact of immigration arises from the family reunification provisions of U.S. immigration law. Because immigrants are more likely to have immediate relatives residing abroad than are native-born U.S. citizens, immigrants play a large role in determining who is able to immigrate to the United States. It is thus important to measure the number of immigrants admitted by each immigrant (a phenomenon that is referred to as the immigration multiplier in the demographic literature). Whereas the potential multiplication of immigration visa entitlements can be substantial because of family reunification provisions, the actual number depends on the propensity of immigrants to naturalize and to sponsor new immigrants. There is currently inadequate information on the rates at which immigrants sponsor new immigrants and thus on the size of the immigration multiplier.
In discussing social and family networks, workshop participants reached agreement on five points, including topics not directly raised in the two papers summarized above:
- The environment faced by immigrants is important. Migration is a transnational process with lots of circulation. It is important to consider the beginning and ending of migration. Because emigrants are highly selective (usually highly educated in their countries of origin, for example), there are enormous problems with survey data on the general population of immigrants. Any cross-sectional data on immigrants must be initially regarded in the context of the recency of immigrant arrivals and their countries of origin. Immigrants are motivated by push and pull factors, especially at the local level. These factors affect the choice of destination, as well as the prospects for return migration.
- The environment of immigrants includes not only the social and economic conditions at the places of origin and destination, but also the family ties and the social networks of alternative places of destination. The social capital of immigrants constitutes the resources that they can bring to bear on economic
- success. Family ties can help the initial adjustment and may offer social capital for economic achievement.
- For immigration in the United States, the type of visa a person has and the duration of the visa status are very significant. Immigrants are by definition a population of movers, and movers are likely to be more mobile socially and economically. But the visa status at entrance to the United States affects eligibility for programs (for refugees, for example) and for employment (illegal aliens have a more curtailed labor market, for example).
- Family and social networks matter greatly, and there is a need to know how they work. Family-based immigration may provide additional support to immigrants; in fact, they may do better than occupationally related immigrants. An interesting idea for research is to make this type of comparison for a sample of immigrants.
Workshop participants noted some important conclusions about the study of social networks. Foreign-born families typically have a larger number of members, a higher proportion of relatives other than immediate family members, and a greater number of adults in the labor force. Whereas the higher labor force participation rates among foreign-born families help to increase overall household income, the complexity of family structure may limit the mobility of individual members. There has been relatively little recent systematic study of immigrant families. Foreign-born families experience both negative and positive outcomes, which would be revealed through careful study of the diverse ways in which they figure into social and economic adjustment. The study of the immigrant family and its consequences for assimilation is an important research question.
Studies of Mexican immigration to the United States illustrate that it is an evolutionary process that becomes self-sustaining. There has been an expansion of emigration from a broader base of Mexican communities. This pattern reinforces the need to collect data in the United States and Mexico in a coordinated fashion.
Some participants argued, however, that networks may no longer be so important a part of the Mexican immigration process as they have been in the past. The reasons for this possibility are as follows: as Mexican migration shifts from rural to urban origins, it is less likely that those leaving for the United States will be able to draw on community-based networks. Urban neighborhoods are not the kind of communities that rural villages and small towns are. Several new studies being conducted in rural Mexico suggest ongoing changes there, too: the village is becoming less the basis for family survival strategies and is more a place of residence, like towns and cities. The research issue is that not all immigrants. from Mexico or other countries, are able to use social networks. Rather, social networks are a "social capital" variable that needs to be explored in accounting for differences in immigration levels and patterns, as in immigrant adjustment.
Discussions with the Mexican government for more than a decade about cooperative data collection suggest strong Mexican support for cooperative work with the United States. A simultaneous Mexico-United States survey is a possibility now worth exploring.3 More money for research is available in Mexico at this time, and there is considerable interest among government and university researchers. Mexico is an urban nation, and urban emigration from Mexico is an open issue for its future development. There is special value in looking at the Mexico-United States border because about 60 percent of the manufacturing jobs created in Mexico in the 1980s were on that U.S. border. In the context of emigration from other Mexican communities, the border area deserves further study for questions of the migratory movement of labor.
The migration process itself provides an important theoretical framework for analyzing immigration into the United States. As described in this chapter, the migration process involves a social network, with the immigrant situated in a complex set of family and relational ties. The available immigration data from administrative records, however, provide information on individual events, abstracting the immigrant from the temporal and social networks. Whereas such agencies as the Immigration and Naturalization Service have a legal interest in such individual events as naturalization, federal immigration policy should also have a broad interest in the interconnection of events in the migration process, including what happens to immigrants after entering the United States.
If significant progress is to be made in the understanding of immigration in the context of U.S. immigration law, new data are needed that take into account the fundamental role of the family—both in the country of origin and in the United States—that follow the immigrant and relevant family members over time, and that are much more comprehensive than existing data in terms of immigrant behavior and immigrant legal status. Moreover, to avoid the biases associated with cross-sectional data, especially those due to emigration selectivity, such data must be longitudinal. Workshop participants suggested that one data base for studying migration processes would consist of an individual survey, drawn from successive cohorts of new immigrants (and possibly illegal aliens and legal nonimmigrants) who are reinterviewed periodically, along with associated family information. Currently, one principal data source for studying immigration is the administrative records of the Immigration and Naturalization Service These data bases, while quite valuable, have many well-known limitations
that arise principally from the fact that they are not designed specifically for the study of immigrants or immigration.
Workshop participants considered possible improvements of Immigration and Naturalization Service administrative records for the study of social networks. There are some limits about what available records can provide: (a) not everyone wants to migrate to the United States, including people with relatives in the United States and people eligible to sponsor their relatives abroad; (b) the sponsor of an immigrant may be only the most convenient of many potential sponsors; and (c) occupationally related immigrants could, in some hypothetical circumstances, have entered as family-based immigrants.
In general, workshop participants urged the Immigration and Naturalization Service to consider enhancements to its the creation of family records. Linkage of current Immigration and Naturalization Service administrative records is feasible in general, although it is complicated by name changes in some records. Also, some records are not automated at present and require a search through paper records. It would be useful to ensure that key records for family study are automated for retrieval and possible linkage. The relatives of an immigrant present special problems because there may be duplicate records, with different relatives reporting on the same person. Increased linkage and exploitation of the usefulness of Immigration and Naturalization Service records would help to improve the quality of the records, including methods for dealing with possible duplication.
One item would be especially useful to improve and attach to the current administrative records: occupation both upon entrance to the United States and after arrival (at the time of naturalization, for example). Occupational data are collected at the time of application for a visa or change of visa status, and again at time of naturalization (for those immigrants who remain in the United States and seek naturalization). Occupational data are currently provided by only about 50 percent of the respondents. Little investigation has been conducted on the selectivity or quality of the reported INS occupational information (one exception is a study by North, 1995). Such added information, especially when more complete and linked to other records, would greatly enhance the value of Immigration and Naturalization Service administrative records for studies of the family and social networks.
In sum, there are significant enhancements to these existing data sources that could potentially produce large gains at relatively little cost:
- Creation of Immigration and Naturalization Service sponsorship files would provide machine-readable records for linking data on immigrants with data on other immigrants they have sponsored.
- Addition of information to the New Immigrant File of the Immigration and Naturalization Service (the principal's A-number, identifiers for both visa sponsor and financial sponsor, information on offsprings and siblings, schooling,
- and number of marriages for self and spouse) would increase the usefulness of current administrative data on new permanent resident aliens. Although the collection of additional information at the time of visa application would increase respondent burden, it would enhance the research value of INS records for important policy analysis. Initiation of data collection, workshop participants argued, should include an assessment of data accuracy.
- Additional enhancements of Immigration and Naturalization Service data (inclusion of more occupational categories and publication of sex-specific counts) would further increase the usefulness of the data.