Social Science Research 40 (2011) 77–86
Contents lists available at ScienceDirect
Social Science Research journal homepage: www.elsevier.com/locate/ssresearch
Ethnic social capital: Individual and group level sources and their economic consequences Asaf Levanon ⇑ University of Haifa, Department of Sociology and Anthropology, Mount Carmel, 31905 Haifa, Israel University of Haifa, Department of Human Services, Mount Carmel, 31905 Haifa, Israel
a r t i c l e
i n f o
Article history: Received 20 January 2010 Available online 29 September 2010 Keywords: Immigrant incorporation Social capital Multilevel modeling
a b s t r a c t Ethnic social capital shapes economic action by immigrants by providing information, training, and credit that is otherwise unavailable. However, prior research on the effects of ethnic social capital on economic attainment by salaried workers primarily relied on case studies of specific destinations or ethnic groups. Furthermore, prior research focused on group level effects of ethnic social capital while largely ignoring the effects of individual-level utilization of ethnic ties. This study addresses these limitations by combining data from the New Immigrant Survey and the 5% file of the 2000 US Census. This allows controlling for group differences in actual utilization of familial ties and for differences between groups in their demographic makeup, legal status, and human capital when estimating the effect of ethnic social capital. It also allows addressing the potentially divergent implications of ethnic social capital for individual and group economic attainment. Results suggest that, on both the individual and group levels, reliance on ethnic social capital is associated with lower earnings for recent legal immigrants to the US. Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction There is little consensus in immigration literature about the factors affecting the economic incorporation of immigrants. Some scholars focus on individual and group level resources (Alba and Nee, 2003; Nee and Sanders, 2001; Portes and Bach, 1985; Sanders, 2007). Others add a focus on the interaction of immigrants with members of other immigrant and native groups (Model and Lapido, 1996; Portes and Zhou, 1993; Waldinger, 1996; Waters, 1999). Finally, several others have focused on the role of the receiving government (Portes and Rumbaut, 2006; Reitz, 1998; van Tubergen et al., 2004). A common thread, however, binds these distinct theoretical and empirical accounts. They all either assume or explicitly focus on the importance of intragroup interaction in shaping the process of immigrant economic incorporation. Indeed, one of the most important insights of the sociological research on immigration has been that ethnic networks affect economic attainment by immigrants by providing information about employment and business opportunities, access to credit, and exposure to cultural resources that are conducive to economic attainment (Alba and Nee, 2003, pp. 42–46; Light and Gold, 2000, pp. 105–129; Portes and Rumbaut, 2006, pp. 86–87; Waldinger, 1996, pp. 21–28). Ethnic networks, in other words, shape the way in which group resources constrain or enable individual action. Light and Gold (2000) use the term ethnic resources to characterize group attributes that provide benefits for ethnic group members through such networks. These attributes include both cultural practices that are conducive for economic advancement (i.e., ethnic cultural capital), such as particular skills and worldviews, and patterns of in-group interaction – ethnic social capital (Light and Gold, 2000, ⇑ Fax: +972 4 8240819. E-mail address:
[email protected] 0049-089X/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.ssresearch.2010.09.014
78
A. Levanon / Social Science Research 40 (2011) 77–86
pp. 107–108). Most of the literature on the role of ethnic resources focuses on its effects on the propensity to engage in selfemployment (e.g., Light, 1972; Light and Bonacich, 1988; Yoon, 1991). However, several studies focused on effects of ethnic resources on economic attainment among salaried workers (e.g., Logan et al., 2003; Nee et al., 1994; Portes and Bach, 1985; Sanders and Nee, 1987; Waldinger, 1996; Wilson, 1999). Research on the role of ethnic networks and ethnic resources in shaping economic opportunities among salaried immigrants currently suffers from two important deficiencies. First, such research primarily relies on case studies of specific destinations or ethnic groups (e.g., Bailey and Waldinger, 1991; Nee et al., 1994; Portes and Bach, 1985; Sanders and Nee, 1987).1 While case studies provide important in-depth information on the experiences of specific groups they do not allow for producing a generalized account that can be used for understanding the incorporation of new immigrant groups or the experiences of immigrants in new destinations. An exclusive focus on case studies of specific destinations and groups seems especially unwarranted when one considers the radical change in recent decades in the country of origin mix of immigrants to the US and in their settlement patterns (Gibson and Jung, 2006; Massey, 2008; Massey et al., 1998; Waters and Jiménez, 2005). Second, prior research focused primarily on group level effects of ethnic resources while largely ignoring the effects of individuallevel utilization of ethnic ties (e.g., Logan et al., 2003; Portes and Bach, 1985; Sanders and Nee, 1987; Waldinger, 1996; Zhou and Logan, 1989). I address the aforementioned deficiencies in this paper by combining data from the New Immigrant Survey (NIS) and the 5% Public Use Microdata Sample (PUMS) file of the 2000 US Census. NIS is a nationally representative longitudinal study of new legal immigrants to the United States and their family members (Jasso et al., in press). I use data from the first wave of NIS (NIS-2003). Using data from NIS offers an important advantage for analysis of the role of ethnic social capital in shaping the economic attainment of immigrants. Specifically, NIS allowed me to use in-depth individual-level data on the migration history of a nationally representative sample of legal immigrants to the US. This improves on prior studies about the role of personal ties and ethnic resources in the process of economic attainment among salaried workers (e.g., Nee and Sanders, 2001; Portes and Bach, 1985), which generally used information from detailed case studies of specific immigrant groups and destinations. NIS allows me to study the effect of personal ties on the economic attainment of immigrants from 27 groups. Following Nee and Sanders (2001), I specifically focused on obtaining employment through familial ties. I complemented the data from NIS with data from the 2000 US Census, which can be used to provide information on the collective benefits or disadvantages that group resources offer to in-group members. Specifically, I use Census data for information on the level of ethnic concentration of the various immigrant groups in specific labor markets. Using multilevel modeling, I estimate the effect of ethnic social capital on individual and group attainment simultaneously. Multilevel analysis allows disentangling within-group person-level effects from between-group contextual effects (Raudenbush and Bryk, 2002). In the context of the present paper, the personal level effect reflects the impact of personal utilization of ethnic ties on individual earnings attainment within each immigrant group. The contextual effect reflects the impact of group differences in the propensity of reliance on such ties on group earnings attainment. Examining both effects at the same time enables me to address the potentially divergent implications of ethnic resources for individual and group economic attainment. I use this framework in evaluating three accounts of the significance of ethnic ties on economic attainment. These accounts differ in the motivation they identify for the reliance of individual migrants on ethnic ties – normative or instrumental – and the duration of using such ties. The implications of these differences for individual and group attainment are specified below. Together, these improvements offer a much broader perspective on the role of ethnic resources in shaping economic opportunities among immigrants. It is important to note that the scope of the paper is limited by the information available in the aforementioned datasets. Both the NIS and the Census PUMS allow direct and indirect measurement of in-group interaction (i.e., ethnic social capital). However, neither data source contains sufficient information for analysis of the role of ethnic cultural capital. Therefore, in this paper I focus exclusively on the role of ethnic social capital in shaping economic attainment by immigrants.
2. Effects of ethnic social capital on economic attainment by immigrants Economic actions by immigrants are affected by relations within both the family and the ethnic community. Settling in a new country involves considerable costs and risks. This motivates newcomers to rely on resources of coethnics or family members who arrived in prior waves of migration (Massey and Espinosa, 1997). As Nee and Sanders (2001) point out, while it is possible for immigrants to draw on resources provided to them either through their families or through their ethnic communities, immigrants often rely on a combination of the two sources. I will therefore focus in the reminder of the paper on the effect of both familial ties and ethnic community resources on patterns of economic adaptation. The literature on ethnic social capital offers three possible accounts for its role in the process of economic attainment. All three accounts share a focus on the role of intragroup interaction in shaping economic action by immigrants. The accounts differ in the motivation they identify for utilization of ethnic ties by immigrants. They also differ in the implications they identify – positive or negative – for relying on ethnic resources for both individual and group economic attainment. 1 See Wilson (1999) for an exception. However, Wilson focuses on broad racial groups and not on ethnic groups. Research on the impact of ethnic networks on the likelihood of engaging in self-employment generally suffers from a similar deficiency, focusing primarily on case studies. See Light and Rosenstein (1995) for a notable exception.
A. Levanon / Social Science Research 40 (2011) 77–86
79
A primary focus on group level effects of ethnic social capital in prior research did not allow for evaluating the potentially conflicting effects of ethnic social capital on group- and individual-level economic progress. Using information on utilization of familial ties and the degree of geographic concentration of each immigrant group allowed me to address this lacuna. The first account emphasizes shared norms and values within an ethnic group (Alba and Nee, 2003, pp. 43–46; Portes, 1995, 1998; Portes and Sensenbrenner, 1993).2 Individuals can acquire a perception of the importance of sharing resources with ethnic group members through socialization during childhood. Alternatively, they can realize the importance of ethnic solidarity by facing a similar set of adverse circumstances as their coethnics, resulting from either discrimination by natives or competition with other groups. In both cases, this leads to norms of mutual support within the group. Group members, in turn, can draw on resources made available through these norms in order to get access to employment, as well as to capital that can be used for human capital and business investments. Initially, this process results in a positive return to reliance on ethnic ties for individuals within the group alongside negative returns for the group, which reflect discrimination directed toward group members by natives. In the long-run, however, this strategy represents a viable vehicle for group mobility and incorporation (Zhou, 2003). Portes and Sensenbrenner (1993) illustrate this process with the example of Chinese immigrants who arrived at the 19th century (see also Nee and Nee, 1973; Zhou, 1992). Faced with various forms of discrimination and lacking an option for returning home, Chinese immigrants created tightly knit communities, which provided support for their members. The concentration of immigrants in specific communities characterized by dense ethnic networks and norms of mutual support resulted in rapid growth of immigrant enterprise and socioeconomic advancement. The practice of sharing resources within the ethnic community among Chinese immigrants survived to this very day in several major cities both by incorporating new group members and by shaping the values and behaviors of children of immigrants (Zhou, 2003). This view leads to two predictions, depending on the progression of assimilation on the group and individual levels. Initially, reliance on ethnic social capital is associated with a higher level of individual attainment but lower levels of group attainment. In time, groups with a tightly knit ethnic community develop patterns of interaction and mutual support that improve both individual and group mobility. Two considerations lead me to focus on the former prediction. First, recent research documented a radical change in the country of origin mix of immigrants to the US and in their settlement patterns (Gibson and Jung, 2006; Massey, 2008; Massey et al., 1998; Waters and Jiménez, 2005). This suggests that a substantial share of immigrants who recently arrived to the US cannot rely on patterns of interaction and institutions developed by immigrants that arrived in prior waves of migration (Massey, 2008). Second, I employ in this study cross-sectional data on the economic activity of recently arrived immigrants. This prevents me from analyzing temporal changes in the process of group attainment. Hence, this discussion leads to the following hypothesis: H1. Reliance on ethnic social capital will be associated with higher earnings on the individual level but with lower mean group earnings. The second account focuses on instrumental motivations for sharing resources with coethnics (Portes, 1995, pp. 13–14). Instead of a moral or normative obligation to share resources, this account turns the attention to expectations for returns either directly from the recipient of resources or from the group. The expectation for return in both cases depends on the ability to withhold resources from the recipient. This can be achieved either by applying sanctions directly toward the recipient or by relying on the ability of the group to subject the recipient to such sanctions. An expectation for direct return is based on trust, while an expectation for repayment on the basis of the sanctioning capability of ethnic networks reflects enforceable trust. Both types of trust lead to increased cooperation and exchange among group members. Exchange primarily within a specific network, however, leads to retention but not accumulation of resources, as it restrains information about and access to resources that are beyond the groups’ purview (Lin, 2001). Hence, a reliance mainly on exchange within the ethnic network can lead to lower rates of economic progress. The negative impact of reliance predominantly on exchange within the ethnic network can be illustrated with findings on the effects of the concentration of immigrants in ethnic employment niches on economic progress. Ethnic employment niches provide employment and training opportunities for immigrants that are often unavailable in the open labor market (Bailey and Waldinger, 1991; Wilson, 2003). Such opportunities are especially important for immigrants upon their arrival. However, by providing such resources ethnic networks limit the incentives to invest in general human capital that would be appreciated in the non-ethnic labor market (Borjas, 1999; Esser, 2004). Therefore, in the long run, employment in ethnic niches limits the prospects of economic progress among immigrants (Nee et al., 1994). Supporting evidence comes from Wilson (1999), who showed that employment in ethnic niches is associated with lower occupational status and wages than employment in the general labor market.3 This discussion leads to a prediction of a negative effect of reliance on ethnic resources on both group and individual level attainment: H2. Reliance on ethnic social capital will be associated with lower earnings on the individual level and with lower mean group earnings.
2
This account focuses on both ethnic social capital, which is the focus of this paper, and ethnic cultural capital. Additional evidence comes from Logan et al. (2003), who show that Korean and Chinese immigrants in New York who work in ethnic sectors earn less than their counterparts working in the open labor market. 3
80
A. Levanon / Social Science Research 40 (2011) 77–86
A final account suggests that, at least in the short run, reliance on ethnic resources might have conflicting implications for group and individual attainment. Similarly to the second account, this view emphasizes instrumental motivations for reliance on ethnic ties. However, as formulated by Bailey and Waldinger (1991; see also Waldinger (1999)), this view focuses on the temporary nature of such reliance and the mutual benefits it provides for employers and employees. For employers, employing coethnics reduces the risk associated in investing in training new workers, since it improves the quality of information on worker characteristics, helps in establishing a steady stream of job seekers, and creates a sense of mutual obligation between employers and employees. For workers, employment within the ethnic community provides training opportunities that are otherwise unavailable. Hence, while employment in the ethnic sector initially offers lower compensation it enables acquisition of skills that are necessary for upward mobility and, consequently, for group advancement. Bailey and Waldinger illustrate this process in their study of the garment industry in New York. They document the reliance on ethnic networks in recruiting workers for firms in the immigrant sector. In explaining their decision to offer training to their coethnic workers, employers in the ethnic sector refer to decreased risks and costs arising from reliance on information transferred through ethnic networks. According to this argument, reliance on ethnic ties would lead to a positive effect on group attainment alongside a negative but short-lived effect on individual attainment: H3. Reliance on ethnic social capital will be associated with lower earnings on the individual level but with higher mean group earnings. It is important to distinguish the arguments discussed above about the effects of ethnic social capital on economic attainment among immigrants from the debate about the effect of ethnic enclaves. As formulated by Wilson and Portes (1980), an ethnic enclave is a distinct economic sector, characterized by a geographical concentration of ethnic enterprises and coethnicity of employers and workers. Using the place of work and co-ethnicity of workers and employers in identifying enclaves, research on ethnic enclaves found that they do not impede the economic advancement of immigrant workers (e.g., Portes and Jensen, 1989; Wilson and Portes, 1980; Zhou and Logan, 1989). By contrast, research on ethnic enclaves using the place of residence in defining enclave membership documented lower returns to human capital among enclave workers compared to workers in the open labor market (e.g., Sanders and Nee, 1987, 1992).4 I do not engage in the ethnic enclave debate in this paper for both theoretical and practical reasons. From a theoretic perspective, I concur with Light et al. (1994) who argue that an ethnic enclave is a special case of an ethnic economy. While the latter can be identified on the basis of coethnicity of workers and employers, the former additionally requires that ethnic firms will be spatially concentrated. Ethnic enclaves are much less frequent than are ethnic economies (Light et al., 1994). Hence, while a focus on ethnic enclaves is theoretically justified, it provides insight into the experiences of members of only a limited number of groups. A focus on ethnic economies – and the resources they provide to their members – allows painting a much broader picture of the role that ethnic resources play in the economic incorporation of immigrants. From a practical standpoint, NIS does not provide information on the geographical concentration of the firms in which the respondents work. I partially addressed this issue by using data from the 2000 US Census regarding the tendency of each group for concentrating in specific labor markets. This allows me to asses the impact of ethnic social capital on the economic incorporation of group members but not to isolate the effect of enclave membership. 3. Method 3.1. Data and variables Data for this article was obtained from two primary sources. Individual level data on the employment characteristics and wages of legal immigrants was taken from the NIS. NIS is based on a nationally representative sample of 8573 immigrants and their family members who were admitted to legal permanent residence during 2003 (Jasso et al., in press). The sampling frame for NIS was based on electronic administrative records compiled for new immigrants by the US government through the Immigration and Naturalization Service. The sample was stratified by class of admission (i.e., spouses of US citizens, employment-based immigrants, diversity program immigrants, and others). In addition, the sample design included respondents from all top (in terms of total immigrant population) 85 Metropolitan Statistical Areas (MSA’s), a random sample of 10 other MSA’s, the top 38 counties, and a random sample of 15 other counties. The response rate for the survey was 68.6%. The interviews were conducted in the language preferred by the respondent. NIS includes individuals from 19 disaggregate immigrant groups (i.e., groups defined by the country of birth). Immigrants from relatively small groups were aggregated in the survey into eight additional groups (i.e., groups defined by the continent of birth).5 Data on group characteristics was computed from the 5% 2000 US Census Public Use Microdata Sample file (PUMS). Group-level variables were first computed 4 The data used by Sanders and Nee (1987) included information on the place of work for only a limited subset of the sample. Analysis using this information indicated that, in contrary to the findings reported by Wilson and Portes (1980), Cuban enclave workers (but not entrepreneurs) experienced lower returns to their human capital than Cubans in Miami working outside the enclave. A similar finding is reported by Zhou and Logan (1989) for Chinese men in New York, when using a third definition incorporating information on the industrial sector of work. 5 Disaggregate groups included: Canada, China, Colombia, Cuba, Ethiopia, Guatemala, Haiti, India, Jamaica, Korea, Mexico, Nigeria, Peru, Philippines, Poland, Russia, Ukraine, United Kingdom, and Vietnam. Aggregate groups included: Asia (East Asia, South Asia and the Pacific), Arctic region, Europe (including Central Asia), Latin America and the Caribbean, Middle East and North Africa, North America, Oceania, and Sub-Saharan Africa.
A. Levanon / Social Science Research 40 (2011) 77–86
81
on the basis of the country of birth and then aggregated (by taking a weighted mean) on the basis of membership in the 27 immigrant groups identified in NIS. The analysis was restricted to working age (i.e., 25–64) males in the civilian labor force who reported their earnings. I focused only on males to avoid conflating differences among ethnic groups in ethnic social capital with differences in attitudes about the traditional roles of men and women in the public and private spheres. Recent research has documented differences across immigrant groups in women’s economic behavior even after controlling for measureable differences in human capital, family responsibilities, and the degree of ethnic concentration (Read and Cohen, 2007). Case studies of specific immigrant groups suggest that such differences are driven by cultural norms among some groups that emphasize women’s primary responsibility for childrearing (Read, 2004). Future research should therefore focus on differences by gender in immigrant integration. Such analysis will require information on group differences in gender role attitudes. The primary dependent variable in this paper was the log of hourly wage, reported either for the current job or for the last job if the respondent was not employed during the time of the survey.6 The main independent variables on the individual level were dummy variables indicating whether the respondent got the job through help from a relative or has an employer who is a relative (including extended family members). The use of these variables is based on work by Nee and Sanders (2001), who conceptualized both familial and ethnic ties as part of the social capital immigrant workers posses. The main group-level variable is the level of ethnic concentration.7 Ethnic concentration was measured by the average overrepresentation of co-ethnics in one’s local labor market. The census variable used to identify the local labor market is a Public Use Microdata Area (PUMA) of work, which is an area of contiguous territory containing 100,000+ residents within a state. The use of overrepresentation as a measure of ethnic concentration reflects the assumption that ethnic concentration shapes the structure of networks within ethnic communities by increasing their density. The density of social relations within an ethnic community, in turn, can lead either to an increased perception of social solidarity or increased sanctioning capability within the ethnic network (see Portes, 1998). As discussed above, the sanctioning capabilities of ethnic networks and the sense of solidarity among coethnics have different consequences in terms of economic attainment. Following Logan et al. (1994), overrepresentation was calculated as odds-ratios and then averaged (while weighting by the size of the group-by-locality cell) across all members of each immigrant group.8 Several control variables were included in the model. On the individual level I control for age (including a squared term), cumulative time in the US (including a squared term), marital status (whether married or not), hours of work per week,9 citizenship status (a dummy), whether the respondents’ spouse is a US citizen, visa type in the last entrance to the US (five dummy variables for diversity, employment-based, family-based, refugee, or other), whether the last entrance to the US was documented or not, whether the respondent received at least some post-secondary education, language proficiency (1 = does not speak English at all, 2 = does not speak well, 3 = speaks English well, and 4 = speaks English very well),10 union membership, and region (East, Midwest, South, and West). On the group level, I included controls for group size and the economic conditions in the local labor market. Group size can shape patterns of earning adaptation either through effects on incentives to invest in ethnic-specific vs. general human capital (e.g., Esser, 2004) or through its effects on patterns of prejudice toward minority groups (e.g., Quillian, 1995; van Tubergen et al., 2004). Both aspects can affect the propensity of ethnic groups to cluster in specific labor markets but are not assumed here to mediate the effect of ethnic concentration and network density on earnings attainment. Economic conditions can have similar effects – either increasing the incentive to invest in general human capital or increasing the level of solidarity among members of immigrant groups facing adverse economic conditions. The limited number of groups in NIS prevented me from including additional group-level variables. However, I discuss in the results section several additional factors that were considered as sensitivity checks. In addition, correlations among level-2 variables were modest and the variance inflation factor for both the level-1 model and level-2 model did not exceed 2.5, suggesting that multicolinearity did not pose a problem. 3.2. Method of analysis The analysis was carried out using multilevel modeling. The model contained two different levels of analysis: effects of individual characteristics on individual earnings attainment within each group and effects of group attributes on group mean earnings. The first aspect constitutes a level-1 model, which is expressed as follows:
Y ij ¼ b0j þ b1j X 1ij þ b2j X 2ij þ b3j X 3ij þ þ bkj X kij þ r ij ; 6
ð1Þ
Restricting the analysis for those reporting earnings only for the current job did not substantially changed the results reported below. Group concentration provides indirect information on the propensity to rely on ethnic social capital. Unfortunately, direct information is not available in the Census, which is the source for the group-level data. However, I addressed this issue more directly by reporting in the results section the earnings of groups with above average propensity to rely on ethnic social capital, as reported by individuals in the New Immigrant Survey. 8 Logan et al. (1994) defined concentration (overrepresentation) using a dummy variable identifying groups whose representation in a specific sector is 50% higher than in the rest of the workforce (i.e., odds ratio of 1.5 and above). By contrast, I use the odds ratios as measures of concentration without converting them into a dichotomous indicator of overrepresentation because such procedure would involve an unnecessary loss of information. It is important to note than the mean level of concentration for most groups is much higher than 1.5, as is evident from the descriptive statistics in Table 1. 9 I did not find a non-linear effect of usual hours of work per week on earnings. In addition, replacing a continuous hours of work variable with a dummy variable for part-time work did not alter the results reported below. 10 Replacing the four category English proficiency measure with a dummy variable indicating if the respondent speaks English well or very well did not change the results. 7
82
A. Levanon / Social Science Research 40 (2011) 77–86
where Yij refers to the earnings of individual i nested in ethnic group j, b0 represents group level intercepts (i.e., means), X1ij and b1j indicates whether the individual received help in finding the job from a relative and the regression coefficient (i.e., slope) associated with it, X2ij and b2j correspondingly relate to the indicator and slope for ‘‘relative is an employer”, X3ij through Xkij are a set of individual level control variables, b3j through bkj are the regression coefficients associated with the control variables, and rij denotes an individual level random effect. In order to asses the effect of ethnic resources on the group mean level of earnings, the intercept (i.e., b0j) is modeled as random (i.e., it is allowed to vary across groups). All control variables are centered around their grand mean and are treated as fixed, thereby adjusting the intercept (i.e., group mean level of earnings) for differences across groups in measurable demographic, legal status, and human capital attributes. The remaining aspect of the theoretical model – effects of ethnic social capital on the group mean of earnings – is represented in the following level-2 equation:
b0j ¼ c00 þ c01 W j þ u0j ;
ð2Þ
Here, c00 is the average immigrant group intercept, Wj refers to a vector of immigrant-group predictors, c01 represents the effects of the immigrant group variables on the group mean earnings, and u0j is an immigrant group random effect.
4. Results Table 1 provides descriptive statistics on the variables used in the analysis. The mean hourly wage for the workers in the sample is $12.06, which is about $5 lower than the mean hourly earnings of the total civilian workforce in private establishments in 2003 (US Census Bureau, 2006, Table 629). The table shows that a substantial share (19%) of the workers in the sample found their jobs with the help of relatives. The share of the sample with relatives as employers is much lower (6%). Most of the immigrants initially arrived with a family-based visa (41%), concurring with prior evidence on the predominance of family-based migration to the US (Martin, 2004; Smith and Edmonston, 1997). In addition, for most of the sample the last entrance to the US was documented (78%). The level of English proficiency seems fairly high (i.e., a mean of 3.01,
Table 1 Means/proportions and standard deviations for individual and immigrant group-level variables.
Dependent variable Logged hourly wage Independent variables Level-1 Help from relative in finding a job Employer is a relative Age Time in the US Marital status Hours of work per week Citizenship status Spouse is a US citizen Visa type Diversity Employment (reference) Family Refugee Other Last entrance to the US was documented At least some post-secondary education English proficiency Union membership Region East Midwest South (reference) West Ni Level-2 Group concentration in specific labor markets (odds ratios) Group size Mean level of unemployment at local labor market Nj Sources: Level-1: NIS and level-2: 5% 2000 US Census PUMS.
Mean/proportion
SD
2.49
1
0.19 0.06 35.9 1.51 0.77 43.07 0.01 0.22
9.89 3.38 12.32
0.17 0.41 0.03 0.16 0.78 0.62 3.01 0.09
0.95
0.37 14 0.27 1499 3.05 46162.2 4.27 27
3.09 47600.77 1.22
A. Levanon / Social Science Research 40 (2011) 77–86
83
corresponding to speaking English well). Similarly high (62%) is the share of the sample with at least some post-secondary education (cf. US Census Bureau, 2003, Table 1.5). In order to address the research hypotheses, I present in Table 2 results from the multilevel model described above. Two models were estimated. A first model included only individual-level attributes. Examination of this model suggested that mean group earnings (i.e., the group earnings intercept) significantly differed from zero. This allowed me to move to model 2, which included group level characteristics.11 To account for group differences in measurable demographic and human capital attributes, as well as for differences in prior legal status, all control variables were centered around their grand mean. As the results for the level-1 predictors are similar for models 1 and 2, I will discuss in the text only the estimates from model 2. Turning first to the individual level predictors of wage, Table 2 shows that workers relying on help from relatives for obtaining their job have hourly wages that are about 12% lower than those who do not rely on such strong ties. Having a relative as an employer, on the other hand, is not associated with lower earnings. The addition of group characteristics in model 2 allows estimation of equation 2 specified above. The table documents a negative effect of group concentration in specific labor markets on the mean level of hourly wage. More concretely, combining information from Table 2 on the coefficient of group concentration with information from Table 1 on its variance shows that a standard deviation increase ingroup concentration is associated with a decrease of about 12% in-group mean hourly wage. The findings suggest that reliance on ethnic social capital among recently arrived legal immigrants is associated with considerably lower earnings. The results are inconsistent with Hypothesis 1, which predicted a positive impact of reliance on ethnic ties for individual workers. By contrast, Hypothesis 2 was based on an account that focused on trust as the motivation for sharing resources with coethnics. This account maintains that relations based on trust increase cooperation and exchange among group members at the cost of restricting access of group members to resources beyond the reach of the ethnic network (Borjas, 1999; Esser, 2004; Nee et al., 1994). This account therefore suggests that reliance on ethnic social capital limits the prospects of economic advancement by immigrants. The findings reported in Table 2 are in line with such account. Similarly to Hypotheses 2 and 3 predicted a negative impact of reliance on ethnic ties for wages of individual workers. However, Hypothesis 3 was based on a view that stresses the temporary nature of the reliance on ethnic ties and the mutual benefits it provides to co-ethnic employers and employees. Contrary to the prediction of Hypothesis 3, workers in-groups with a substantial degree of ethnic concentration had a lower mean level of economic attainment. The results do not suggest, however, that ethnic employers are negatively impacted by reliance on ethnic social capital. Such inference requires analysis of factors affecting success among ethnic entrepreneurs. This issue falls beyond the scope of the current paper – a point I will discuss further at the discussion section. To provide some additional insight into the results I examined which groups score below the mean in terms of hourly wage and above the mean on the level of group concentration. Seven groups fit the description, including immigrants from Cuba, the Dominican Republic, Ethiopia, Haiti, Sub-Saharan Africa, the Middle East and North Africa, and Oceania. Four groups remained when I additionally narrowed this down to groups who also score above the mean in terms of the proportion relying on relatives in finding a job: Cuba, the Dominican Republic, Ethiopia, and the Middle East and North Africa. Considering the demographic composition of these groups, I ran an additional model with a group control for the proportion black.12 The negative effects of ethnic concentration persisted even after controlling for proportion black. Considering past research on the effect of ethnic concentration on economic adaptation for members of this group (e.g., Portes and Bach, 1985), the finding for Cubans seems surprising (but see Sanders and Nee, 1987, p. 756). However, recent research suggests that recent immigrants from Cuba are quite different from earlier migrants in terms of their human capital attributes and their economic success (Portes and Shafer, 2007). Several sensitivity checks were preformed for the models presented in Table 2. First, replacing the mean level of unemployment in the local labor market with alternative measures of local labor market conditions – specifically the median wage and the level of inequality (90/10 percentile ratio) – did not change the results. In addition, the results were also insensitive to changing the definition of the local labor market to Super PUMA (an area of more than 400,000 residents) instead of PUMA. I also replaced the measure for group size with indicators for birth in Mexico and added an individual level control for birth in Mexico and an interaction between birth in Mexico and help from relatives in finding a job. Models including these variables produced similar results to those reported above.
5. Discussion The main finding of this article is that the use of ethnic social capital is associated with lower hourly earnings for recent legal immigrants to the US. Utilization of strong/familial ties in obtaining employment is associated with substantially lower earnings. This finding is in line with prior evidence showing that utilization of such ties is associated with a higher likelihood of employment in low-skilled jobs (Nee and Sanders, 2001). Concomitantly, groups with higher levels of ethnic concentration have considerably lower earnings than groups with lower levels of ethnic concentration, even after accounting for group differences in measurable human capital, prior legal status, and demographic attributes, as well as economic conditions in 11 Additional analysis suggested that the variance of the effect of help from relatives in finding work and of the effect of employer is a relative were not significantly different from zero. Therefore, these coefficients were not allowed to vary in model 2. 12 Information on race was not available in NIS. The variable measuring proportion black is therefore based on, 2000 US Census data.
84
A. Levanon / Social Science Research 40 (2011) 77–86 Table 2 Multilevel regression of log hourly wage: effects of individual and group characteristics.
Level-1 Help from relative in finding a job Employer is a relative Age Age2 Time in the US Time in the US2 Marital status Hours of work per week Citizenship status Spouse is a US citizen Visa type Diversity
Model 1
Model 2
0.136* (0.061) 0.026 (0.105) 0.042** (0.015) 0.001* (0.0) 099*** (.014) 0.005*** (0.001) 0.069 (0.052) 0.008*** (0.001) 0.359 (0.238) 0.117 (0.089)
0.127* (0.061) 0.021 (0.105) 0.042** (0.015) 0.001* (0.0) 0.099*** (.014) 0.005*** (0.001) 0.064 (0.052) 0.008*** (0.001) 0.334 (0.238) 0.129 (0.089)
0.786*** (0.066)
0.765*** (0.066)
0.611*** (0.075) 0.661*** (0.075) 0.546*** (0.074) 0.048 (0.053) 0.262*** (0.051) 0.166*** (0.027) 0.016 (0.071)
0.602*** (0.075) 0.578*** (0.129) 0.518*** (0.076) 0.047 (0.053) 0.261*** (0.051) 0.162*** (0.027) 0.023 (0.071)
0.075 (0.053) 0.121 (0.064) 0.077 (0.059)
0.072 (0.053) 0.115 (0.064) 0.062 (0.059)
2.643*** (0.046)
0.038* (0.017) 0 (0) 0.085 (0.051) 2.621*** (0.044)
0.495 32.39 0.036 82.83 1404 27
0.493 32.58 0.031 85.69 1404 27
Employment family
Refugee Other (reference) Last entrance to the US was documented At least some post-secondary education English proficiency Union membership Region East Midwest South (reference) west Level-2 Group concentration in specific labor markets (odds ratios) Group size Mean level of unemployment at local labor market Intercept Variance components rij % of level-1 variance explained u0j % of level-2 variance explained Ni Nj
Sources: Level-1: NIS and level-2: 5% 2000 US Census PUMS. Note: *p < 0.05; **p < 0.01; ***p < 0.001 (two-tailed test; standard errors in parentheses). All variables except for help from relative in finding a job and employer is a relative are grand-mean centered.
A. Levanon / Social Science Research 40 (2011) 77–86
85
the local labor market. This finding supports a view that focuses on the effect of ethnic resources on the incentives to invest in general human capital that would be appreciated in the non-ethnic labor market (Alba and Nee, 2003; Borjas, 1999; Esser, 2004). It is also in line with prior research documenting a negative effect of concentration of ethnic groups in specific occupational niches on economic attainment (Logan et al., 2003; Wilson, 1999). This paper provides two contributions to the immigration literature. First, prior research on the effects of ethnic social capital on earnings attainment among salaried workers generally relied on information from detailed case studies (Nee and Sanders, 2001; Logan et al., 2003). By contrast, I use in this study a nationally representative sample of legal immigrants to the US, thereby providing much stronger evidence on the role of ethnic social capital than prior studies. Second, prior studies tended to focus primarily on the role that aggregate group resources play in the process of economic attainment by immigrants. In contrast, I combined detailed information on individual-level attributes with information on group characteristics. This allowed me to focus on the competing implications of ethnic social capital for individual and group mobility. Combining information on individual-level attributes with information on group characteristics also allowed me to control for group differences in actual utilization of familial ties and for differences between groups in their demographic makeup, prior legal status, and human capital attributes. This provides stronger evidence on the role of ethnic social capital than prior research that utilized Census data (e.g., Logan et al., 2003; Wilson, 1999). Before concluding this discussion I note several important limitations of this study. First, while NIS will eventually provide panel data on the experiences of immigrant, only the first wave of the survey is currently available. Hence, this study is based on information on the experiences of recently arrived immigrants, which has important empirical implications. Prior research has shown that immigrants tend to rely more heavily on ethnic and familial support during the period of their initial adjustment to their new country. Reliance on group resources diminishes with time as workers acquire financial, human, and cultural capital that allows them to participate in the general labor market (Nee et al., 1994). Reliance on employment within the coethnic community, moreover, might not only provide temporary relief but also training opportunities that are unavailable for immigrants otherwise (Bailey and Waldinger, 1991). If this is the case, then the negative impact of reliance on ethnic and familial resources documented in this paper might be restricted to the period of initial adjustment. Analysis of future waves of NIS will allow examining this issue. Hence, it is important to restrict the scope of the implications of the current research to the period of initial adjustment. Second, the public use version of NIS does not have information on the composition and characteristics of the immediate ethnic community of each respondent. This prevented me from addressing the immigrant enclave debate (e.g., Wilson and Portes, 1980; Portes and Jensen, 1989; Sanders and Nee, 1987; Zhou and Logan, 1989). This serves to further limit the scope of the current research, since it is possible that enclave employment affects the economic prospects of immigrants differently than other forms of ethnic concentration. Third, NIS provides information only on legal immigrants, further restricting the scope of the current study. A final limitation of this study arises from my conceptualization of group economic success. Specifically, the current paper focuses exclusively on earnings of salaried workers while ignoring the role that ethnic social capital plays in shaping the economic success of immigrant entrepreneurs. The study of factors affecting entrepreneurial success occupies a central role in immigration literature. By and large, the literature documents a positive impact of reliance on ethnic social capital on entrepreneurial rates and success (for reviews, see Light and Gold (2000) and Zhou (2004)). Rather than disconfirming these findings, this paper underscores the need for a more nuanced understanding of factors affecting group attainment. The findings of this study specifically call for a framework that examines both dimensions of economic incorporation, while retaining the advantage of employing a nationally representative sample of the population (for a case study applying such a framework see Nee et al. (1994)). Acknowledgments The author wishes to thank Paula England, David B. Grusky, Laura Lopez-Sanders, Emily Ryo, C. Matthew Snipp and two anonymous reviewers for their valuable comments. References Alba, Richard D., Nee, Victor, 2003. Remaking the American Mainstream: Assimilation and Contemporary Immigration. Harvard University Press, Cambridge, MA. Bailey, Thomas, Waldinger, Roger, 1991. Primary, secondary, and enclave labor markets: a training systems approach. American Sociological Review 56, 432–445. Borjas, George J., 1999. Heaven’s Door: Immigration Policy and the American Economy. Princeton University Press, Princeton, NJ. Esser, Hartmut, 2004. Does the ‘‘new” immigration require a ‘‘new” theory of intergenerational integration? International Migration Review 38, 1126–1159. Gibson, Campbell, Jung, Kay, 2006. Historical Census Statistics on the Foreign-Born Population of the United States: 1850–2000. Population Division: Working Paper No. 81. US Census Bureau, Washington, DC. Jasso, G., Massey, D.S., Rosenzweig, M.R., Smith, J.P., in press. The US new immigrant survey: overview and preliminary results based on the new-immigrant cohorts of 1996 and 2003. In: Morgan, Beverly, Nicholson, Ben (Eds.), Immigration Research and Statistics Service Workshop on Longitudinal Surveys and Cross-Cultural Survey Design: Workshop Proceedings. Crown Publishing, London, UK, pp. 29–46. Light, Ivan H., 1972. Ethnic Enterprise in America: Business and Welfare Among Chinese, Japanese, and Blacks Berkeley. University of California Press. Light, Ivan, Bonacich, Edna, 1988. Immigrant Entrepreneurs. University of California Press, Berkeley. Light, Ivan H., Carolyn N. Rosenstein. 1995. Race, Ethnicity, and Entrepreneurship in Urban America. Aldine de Gruyter, New York. Light, Ivan Hubert, Gold, Steven J., 2000. Ethnic Economies. Academic Press, San Diego.
86
A. Levanon / Social Science Research 40 (2011) 77–86
Light, Ivan, Sabagh, Georges, Bozorgmehr, Mehdi, Der-Martirosian, Claudia, 1994. Beyond the ethnic enclave economy. Social Problems 41, 65–80. Lin, Nan, 2001. Social Capital: A Theory of Social Structure and Action. Cambridge University Press, New York. Logan, John R., Alba, Richard D., McNulty, Thomas L., 1994. Ethnic economies in metropolitan regions: Miami and beyond. Social Forces 72, 691–724. Logan, John R., Alba, Richard D., Stults, Brian J., 2003. Enclaves and entrepreneurs: assessing the payoff for immigrants and minorities. International Migration Review 37, 344–388. Martin, Philip L., 2004. The United States: the continuing immigration debate. In: Cornelius, Wayne A., Tsuda, Takeyuki, Martin, Philip L., Hollifiled, James F. (Eds.), Controlling Immigration: A Global Perspective, second ed. Stanford University Press, Stanford, CA, pp. 51–85. Massey, Douglas S., 2008. New Faces in New Places: The Changing Geography of American Immigration. Russell Sage Foundation, New York. Massey, Douglas S., Arango, Joaquin, Hugo, Graeme, Kouaouci, Ali, Pellegrino, Adela, Taylor, J. Edward, 1998. Worlds in Motion: Understanding International Migration at the End of the Millennium. Oxford University Press, New York. Massey, Douglas S., Espinosa, Kristin E., 1997. What’s driving Mexico-US bigration? A theoretical, empirical, and policy analysis. American Journal of Sociology 102, 939–999. Model, Suzanne, Lapido, David, 1996. Context and opportunity: minorities in London and New York. Social Forces 75, 485–510. Nee, Victor, Nee, Brett de Bary, 1973. Longtime California: A Documentary Study of an American Chinatown. Pantheon, New York. Nee, Victor, Sanders, Jimy, 2001. Understanding the diversity of immigrant incorporation: a forms-of-capital model. Ethnic and Racial Studies 24, 386–411. Nee, Victor, Sanders, Jimy M., Sernau, Scott, 1994. Job transitions in an immigrant metropolis: ethnic boundaries and the mixed economy. American Sociological Review 59, 849–872. Portes, Alejandro, 1995. Economic sociology and the sociology of immigration: a conceptual overview. In: Portes, Alejandro (Ed.), The Economic Sociology of Immigration: Essays on Networks, Ethnicity, and Entrepreneurship. Russell Sage Foundation, New York, pp. 1–41. Portes, Alejandro, 1998. Social capital: its origins and applications in modern sociology. Annual Review of Sociology 24, 1–24. Portes, Alejandro, Bach, Robert L., 1985. Latin Journey: Cuban and Mexican Immigrants in the United States. University of California Press, Berkeley. Portes, Alejandro, Jensen, Leif, 1989. The enclave and the entrants: patterns of ethnic enterprise in Miami before and after the Mariel. American Sociological Review 54, 929–949. Portes, Alejandro, Rumbaut, Rubén G., 2006. Immigrant America: A Portrait, third ed. University of California Press, Berkeley, CA. Portes, Alejandro, Sensenbrenner, Julia, 1993. Embeddedness and immigration: notes on the social determinates of collective action. American Journal of Sociology 98, 1320–1350. Portes, Alejandro, Shafer, Steven, 2007. Revisiting the enclave hypothesis: Miami twenty five years later. Research in the Sociology of Organizations 25, 157– 190. Portes, Alejandro, Zhou, Min, 1993. The new second generation: segmented assimilation and its variants. The Annals of the American Academy of Political and Social Science 530, 74–96. Quillian, Lincoln, 1995. Prejudice as a response to perceived group threat: population composition and anti-immigrant and racial prejudice in Europe. American Sociological Review 60, 586–611. Raudenbush, Stephen W., Bryk, Anthony S., 2002. Hierarchical Linear Models: Applications and Data analysis Methods. Sage Publications, Thousand Oaks, CA. Read, Jen’nan G., 2004. Cultural influences on immigrant women’s labor force participation: the Arab-American case. International Migration Review 38, 52–77. Read, Jen’nan G., Cohen, Philip N., 2007. One size fits all? Explaining US-born and immigrant women’s employment across 12 ethnic groups. Social Forces 85, 1714–1734. Reitz, Jeffrey G., 1998. Warmth of the Welcome: The Social Causes of Economic Success for Immigrants in Different Nations and Cities. Westview Press, Boulder, CO. Sanders, Jimy M., 2007. Nativity, human capital, and government employment. Social Science Research 36, 404–420. Sanders, Jimy M., Nee, Victor., 1987. Limits of ethnic solidarity in the enclave economy. American Sociological Review 52, 745–773. Sanders, Jimy M., Nee, Victor., 1992. Problems in resolving the enclave economy debate. American Sociological Review 57, 415–418. Smith, James P., Edmonston, Barry., 1997. The New Americans: Economic, Demographic and Fiscal Effects of Immigration. National Academy Press, Washington, DC. van Tubergen, Frank., Maas, Ineke., Flap, Henk., 2004. The economic incorporation of immigrants in 18 Western societies: origin, destination, and community effects. American Sociological Review 69, 704–727. US Census Bureau, 2003. Foreign-Born Population of the United States Current Population Survey – Detailed Tables (PPL-174). US Census Bureau, Population Division, Washington, DC (retrieved 23.10.09).
. US Census Bureau, 2006. Statistical Abstract of the United States (125 Edition). National Technical Information Service, Springfield, VA. Waldinger, Roger D., 1996. Still the Promised City? African Americans and New Immigrants in Postindustrial New York. Harvard University Press, Cambridge, MA. Waldinger, Roger D., 1999. Network, bureaucracy, and exclusion: recruitment and selection in an immigrant metropolis. In: Bean, Frank D., Bell-Rose, Stephanie (Eds.), Immigration and Opportunity: Race, Ethnicity, and Employment in the United States. Russell Sage Foundation, pp. 228–260. Waters, Mary C., 1999. Black Identities: West Indian Immigrant Dreams and American Realities. Russell Sage Foundation, New York. Waters, Mary C., Jiménez, Tomás, 2005. Assessing Immigrant Assimilation: New Empirical and Theoretical Challenges. Annual Review of Sociology 31, 105– 125. Wilson, Franklin D., 1999. Ethnic concentration and labor market opportunities. In: Bean, Frank D., Bell-Rose, Stephanie (Eds.), Immigration and Opportunity: Race, Ethnicity, and Employment in the United States. Russell Sage Foundation, pp. 106–140. Wilson, Franklin D., 2003. Ethnic niching and metropolitan labor markets. Social Science Research 32, 429–466. Wilson, Kenneth L., Portes, Alejandro, 1980. Immigrant enclaves: an analysis of the labor market experience of Cubans in Miami. American Journal of Sociology 86, 295–319. Yoon, In-Jin, 1991. The changing significance of ethnic and class resources in immigrant business: the case of Korean immigrant businesses in Chicago. International Migration Review 25, 303–332. Zhou, Min, 1992. Mew York’s Chinatown: The Socioeconomic Potential of an Urban Enclave. Temple University Press, Philadelphia. Zhou, Min, 2003. Assimilation, the Asian Way. In: Jacoby, Tamar (Ed.), Reinventing the Melting Pot: The New Immigrants and What it Means to be American. Basic Books, New York, pp. 139–153. Zhou, Min., 2004. Revisiting ethnic entrepreneurship: convergencies, controversies, and conceptual advancements. International Migration Review 38, 1040–1074. Zhou, Min, Logan, John R., 1989. Returns on human capital in ethnic enclaves: New York City’s Chinatown. American Sociological Review 54, 809–820.