Social Capital

The concept of social capital, as described by ^{Bourdieu (1985} , ¹⁹⁸⁰ and ¹⁹⁷⁹⁾ , sheds light on the different trajectories followed by immigrants as they incorporate into their host societies. When the conditions in the host society are hostile (due to ethnic discrimination, legal obstacles, or labor segregation, among others), the social capital of immigrants creates opportunities within a limited, although dynamic, structure of individual and collective development.

Bourdieu defines social capital as the aggregate of current and potential resources associated with a network of stable and institutionalized social relations that provides each of the members of the network with the capital possessed by the group as a whole (1985 and 1980). The volume of an individual's social capital depends on the size of the networks that they can mobilize and the volume of economic and cultural capital that they possess by virtue of their connection to these networks. The subjects invest their efforts, consciously and unconsciously, in the generation and maintenance of networks of relationships that allow them to access certain resources. This investment is translated into exercises of interchange. On the one hand, the accumulated benefits of the group represent the basis of its solidarity, which at the same time makes it possible to accumulate benefits and exchange them among its members. On the other hand, the exchange takes place within certain borders and codes, which exclude the possibility of interaction under other parameters or outside the borders.

In the field of migration, and drawing from Bourdieu's work (1985, 1980, 1979), Alejandro Portes ( ^{Portes 2000} , ¹⁹⁹⁸ , ¹⁹⁹⁵ ; ^{Portes & Sensenbrenner, 1993} ) defines social capital as the ability of individuals to manage scarce resources by taking advantage of their position in networks or wider social structures. Social capital is not an inherent property of an individual: it exists and is anchored in the individual's network of relationships and contacts. Social capital can provide access to resources or restrict individual liberties by controlling behavior (through social norms). Social capital is preserved by engaging in different activities, such as meeting with people, maintaining periodic communication, participating in events, joining associations, etc. ( ^{Vertovec, 2003} ); as expressed by ^{Bourdieu (1980)} time and resources are invested to reproduce social capital.

^{Portes (2000} , 1998) proposed a triple definition of social capital: a) social capital as a source of social control, b) social capital as a source of benefits mediated by the family, and c) social capital as a source of resources mediated by non-family networks; the last definition is the closest to Bourdieu’s concept (1985, 1980). In his work, Portes applies the concept to social science ( ^{Cachón, 2012} ) to describe the unintended and negative consequences of social capital, classified by the author into four types ( ^{Portes, 2000} , ¹⁹⁹⁸ & ¹⁹⁹⁵ ; ^{Portes & Sensenbrenner, 1993} ): a) restricted access to opportunities because the bonds that unite the members of the group exclude non-members from access to such opportunities; b) individual freedom is restricted because participation or membership in a group entails an unavoidable loss of individual autonomy; c) excessive pressure over group members in connection with the “closure” of social relationships in the group, which provides benefits but, under certain circumstances, can limit the development of its members; and d) downward leveling norms, which in certain groups are expressed as social pressure that maintains individuals within the group and prevents them from seeking opportunities outside the group.

The field of migration studies has understood migration networks as a form of social capital defined as a series of interpersonal ties connecting immigrants, former immigrants, and non-migrants with relatives, friends, and their communities of origin, both in their native countries and in their host countries ( ^{Massey, Arango, Hugo, Kouaouci, Pellegrino & Taylor, 2006, p. 42} ). Migration networks are sources of shared information and provide migrants with economic aid and different kinds of support, increasing their probability to achieve social mobility inasmuch as they reduce its associated risks and costs and increase the individual’s expected returns. Migratory networks are a form of social capital. They provide immigrants with access to fundamental resources, such as employment, higher salaries, options to send remittances, etc. They cushion the economic and emotional weight of migrating and represent spaces of certainty where migrants can obtain information and security ( ^{Gurak & Caces, 1992} ). They are, therefore, the main responsible mechanism for the expansion and endurance of migratory flows ( ^{Imilan, Garcés & Margarit, 2014} ).

Research in the field has shown that even if the original apparent causes of migration are overcome (wage differential, lack of formal recruitment, etc.), migratory flows become consolidated and reproduce thanks to existing migratory networks. Networks are reproduced and extended inasmuch as immigrants use them and acquire the possibility and obligation of consolidating the system by providing more information to potential migrants ( ^{Tilly, 2007} and ¹⁹⁹⁰ ). This process increases the complexity and extension of social networks and explains their role as migration microstructures ( ^{Portes & Rumbaut, 2010} ).

Occupational Mobility

The issue of occupational mobility among immigrants has been a focus of attention since ^{Barry Chiswick’s (1979} , ¹⁹⁷⁸ , ¹⁹⁷⁷⁾ pioneering work, in which the author notes that immigrants in the US were initially integrated into the labor market by providing them with low-level jobs and salaries below the average for Americans with similar characteristics. After a while, and as an effect of “Americanization,” these differences became nuanced and even nullified, resulting in upward occupational mobility. A Ushaped mobility pattern, characterized by an initial decline and a subsequent recovery of occupational status, was proposed by Chiswick to explain the adjustment processes of immigrants in labor markets (1979, 1978, and 1977). Chiswick identified two determinants of the economic progress of immigrants: transferability and self-selection. Transferability has to do with the fact that human capital, work experience, and skills acquired by immigrants in their countries of origin are not fully transferable to the labor market in the destination country: transferability depends on the immigrant’s language skills and possession of professional licenses, certifications, or credentials to support such capital ( ^{Redstone, 2008} ; ^{Chiswick, Yew & Miller, 2005} ; ^{Duleep & Regets, 1997} ; ^{Chiswick, 1997} , ¹⁹⁷⁸ ). The second factor points out that people who migrate due to economic reasons are more capable and motivated than people who do not migrate ( ^{Chiswick, 2008} , ¹⁹⁷⁹ , ¹⁹⁷⁸ ), which implies that, given similar socio-demographic characteristics, immigrants have more innate ability or motivation relevant to the labor market than native people ( ^{Chiswick, 1978, p. 901} ).

^{Chiswick (2008} , ¹⁹⁹⁷ , ¹⁹⁷⁹ , ¹⁹⁷⁸ , ¹⁹⁷⁷⁾ and other economists who have supported his approach base their research on the theory of social hierarchy ( ^{Erikson & Goldthorpe, 1992} ); that is, they analyze mobility as the movement of individuals between social groups organized hierarchically on the basis of their access to certain resources, such as prestige, wealth, or status ( ^{Goldthorpe, 2013} ; ^{Erikson & Goldthorpe, 2002} , ¹⁹⁹² ). When considering occupational or wage achievement from an attributive perspective, that is, as a result of having or not having income and wealth, economists treat income correlations as continuous variables and disregard variables related to the social structure, for example, production relationships in the labor market ( ^{Goldthorpe, 2013} ; ^{Erikson & Goldthorpe, 2002} , ¹⁹⁹² ). As a result, they tend to overestimate upward mobility and make economic immobility invisible—individual mobility trajectories are disassembled from the network of economic and social relations through which people acquire more or less advantageous positions in the social structure ( ^{Goldthorpe, 2013} ).

^{Erikson and Goldthorpe (1992)} regard social mobility as an element of the class structure and analyze relationships between different positions. Class structure is built as a result of the relationships established by individuals with production units and the labor market; the authors propose a basic classification of class positions: employers, selfemployed workers without employees, and employees ( ^{Erikson & Goldthorpe, 1992} ). The main argument of this perspective challenges the idea of social hierarchy and states that there are limited groups of positions whose central characteristic is their closure. Based on this theoretical approach, ^{Erikson and Goldthorpe (1992)} point out that absolute mobility—the rate of intergenerational movements movement between class positions—is the result of exogenous structural factors, that is, of the evolution of class structure over time, and that relative mobility—social mobility rates independently of structural influences—shows stable invariance patterns ( ^{Goldthorpe, 2013} ; ^{Erikson & Goldthorpe, 2002} , ¹⁹⁹² ). This relational analysis, focused on social fluidity, uses odds ratios to measure the individual’s probabilities to attain certain social positions considering their initial positions ( ^{Goldthorpe, 2013} ; ^{Erikson & Goldthorpe, 2002} , ¹⁹⁹² ).

Using the strategy proposed by ^{Aysa-Lastra and Cachón (2013a} , ^2013b) in their analysis of occupational mobility among immigrants in Spain, the present article harmonizes ^{Chiswick's (2008} , ¹⁹⁹⁷ , ¹⁹⁷⁹ , ¹⁹⁷⁸ , ¹⁹⁷⁷⁾ perspective on occupational mobility and ^{Erikson and Goldthorpe's (2002} , ¹⁹⁹²⁾ sociological analysis of social mobility with the theory of labor market segmentation ( ^{Doeringer & Piore, 1985} ; ^{Piore, 1979} ). The main argument of this theory is that the labor market consists of two segments: a primary segment, characterized by high wages, good working conditions, associated prestige, and stability, and a secondary segment, characterized by low wages, precariousness, poor working conditions, and instability. Consistent with the approach to social mobility as an element of class structure, the theory of labor market segmentation indicates that both segments have the characteristics of inward openness and outward closure. Thus, our analysis of absolute and relative occupational mobility within each labor segment considers each segment as an occupation cluster ( ^{Aysa-Lastra & Cachón, 2013a} , ^2013b ) per the International Standard Classification of Occupations (ISCO-88).

Two research hypotheses guide the present study:

Methodology and Data Sources

The present study used data from three different sources. Firstly, data on immigrant population from the last valid Population Census (2002) and administrative migratory records from the 2005-2014 period (visas, permits, nationalization procedures, refugee applications, and deportations) were obtained from the Department of Immigration and Foreign Status (DEM), ^{Chilean Ministry of the Interior (2016)} . Secondly, data from the 2015 National Socioeconomic Characterization Survey (Casen) ( ^{Ministry of Social Development, 2015} ), whose analysis unit is the household and its residents. Thirdly, data obtained using a survey developed by this author in the framework of the doctoral thesis entitled "Encuesta de Inmigrantes Sudamericanos en Santiago de Chile 2015” (survey of South American immigrants in Santiago), carried out between the months of April and June 2015.

In our survey, the sample consisted of immigrants of both sexes between 18 and 65 years of age who had lived in the city for at least six months; participants included people from the five quantitatively most important Spanish-speaking South American countries (Peru, Argentina, Bolivia, Ecuador, and Colombia), which represent 65% of total immigration in the RMS. The present study used non-probabilistic quota sampling. The sample was set at 700 cases, 140 per national group separated by sex: men (70) and women (70).

To avoid the overrepresentation of individuals ( ^{Corbetta, 2007} ) presenting certain profiles that are easier to locate due to their occupations (in commerce and services), we combined active recruitment in public spaces with the snowball technique, and the study included only the participants that were contacted via the snowball technique with two limitations. 1) the number of participants recruited by initial contact was limited to a maximum of two, excluding further contacts provided by these two; and 2) participants with family ties with the initial contact were excluded. The final sample, slightly different from the projected sample due to difficulties to find participants and non-response errors ( ^{Corbetta, 2007} ), is shown in Table 2.

When the analysis considers immigrants from the five nationalities included in the survey as an aggregate, the sample is weighted according to the relative weights and sex-based distributions of the groups with the purpose of matching the characteristics of Santiago's population. These cases refer to South American immigrants. When the observations are focused on national groups, the sample is not weighted so that enough cases are available for an adequate comparative analysis.

Absolute and relative occupational mobility analyses use the International Standard Classification of Occupations (ISCO-88), in its one-digit version (without considering category 0, Armed Forces):

This classification of occupations is in line with the theory of labor market segmentation, and it includes the two segments: primary, occupations 1 to 4, and secondary, occupations 5 to 9. The processes of absolute occupational mobility of South American immigrants between their first transition (last occupation in their country of origin and first in Santiago) and their second transition (first occupation in Santiago and occupation at the time of the survey) were analyzed using weighted joint distributions, shown as input (origin) and output (destination) tables for the two transitions. The analysis of relative occupational mobility, or social fluidity, used odd ratios² by labor market segments (primary, occupations 1-4, and secondary, occupations 6-9), and considered category 5 as a reference occupational category, a buffer or limit zone between the two segments ( ^{Parkin, 1978} ).

Both types of analysis used a subsample of respondents composed of immigrants who were employed in their original countries, obtained a first job in Santiago and, at the time of the survey, were employed in the city. These three conditions were applied in a summative manner. The variable of time was not controlled in the labor mobility analysis since 55% of this subsample had lived in Santiago for three years or less, and 70% for five years or less. That is, these migrants arrived in the city recently.

Migratory Networks

Interviewed South American immigrants reported that, in their first six months in the city, immigrant relatives and friends were their primary support (Table 3). Different elements could be identified depending on specific national groups: Ecuadorian immigrants rely on closed networks, predominantly composed of relatives and very close associates ( ^{Granovetter, 1983} , ¹⁹⁷³ ); Peruvian, Bolivian, and Colombian immigrants prefer more extended networks where, in addition to relatives, immigrant friends, acquaintances, and contacts are also important; Argentinian immigrants have developed the most open network, which includes the types of relations stated for the other national groups in addition to Chilean friends and businessmen, that is, a more significant presence of “weak ties” ( ^{Granovetter, 1983} , ¹⁹⁷³ ). A remarkable aspect of this context is the scant importance given to immigrants by public administration institutions (at the municipal and federal levels), even less than the interest shown by churches and non-governmental organizations (NGOs), reflected by the marginal relevance of migratory policies in the country ( ^{Stang, 2016} ; ^{Thayer, 2015} ; ^{Stefoni, 2011} ).

Concerning the role of networks as sources of job opportunities, Table 4 shows that the relative incidence of certain actors is accentuated to the detriment of others: it decreases the importance of the family in obtaining a job in favor of immigrant friends and contacts or acquaintances; the latter actors were found to be a significant factor to increase the individual's probabilities to enter the labor market. In other words, contacts and acquaintances are more effective than the immigrant's family and friends when it comes to helping them to obtain a job in Chile, reflecting ^{Granovetter’s (1983} , ¹⁹⁷³⁾  reflections concerning the higher potential of weak ties over strong ties in this regard.

One-fifth of the interviewed immigrants declared not having received support to find a job in Santiago; this circumstance was especially frequent among Argentinians (36.8%) and Bolivians (30.7%). However, both immigrant groups have different reasons for managing their incorporation into the labor market autonomously; these reasons can be placed in opposite ends of a spectrum. Argentinians, whose educational levels are the highest across immigrant groups and whose presence in the city is the longest, can enter the labor market by presenting their credentials ( ^{Chiswick, 1997} ), but Bolivians, who began arriving in the city much more recently and have lower educational levels than Argentinians, must use other strategies to obtain a job, such as approaching other immigrant groups. Peruvian immigrants were the group in which fewer people reported not having had help to find a job, which can be due to their quantitative importance in the city and, consequently, the availability of large networks of contacts and businesses run by members of the community, among other benefits.

Absolute Occupational Mobility

The processes of absolute occupational mobility experienced by South American immigrants in their first and second job transitions are shown, respectively, in Tables 5 and 6. These results show that the prevailing pattern is immobility, and upward and downward trajectories occur mostly within segments, which expresses the closure of the groups and their relative internal mobility. In their first occupational transition (Table 5) more than half of the immigrants maintained the occupational category that they had in their last job in their countries of origin (56.6%), one third of them experienced downward mobility (33.2%), and only 10.2% experienced upward mobility. Mobility between labor market segments is also shown in the table. Among the people who experienced upward mobility (10.2%), 8.3% did so within the secondary segment, 0.9% within the primary segment, and only 0.9% moved from the secondary segment to the primary segment. As for downward mobility (33.2%), 19.2% of surveyed immigrants remained in the secondary segment, 2.3% in the primary segment, and 11.7% descended from the primary to the secondary segment.

Concerning the second occupational transition (Table 6), 78.4% of surveyed immigrants remained in the same occupational group, 4.8% moved downward, and 16.8% experienced upward mobility. The dominant pattern of absolute mobility is invariance; among those who experienced mobility, whether upward or downward, 90% did so within their own segment, which confirms the hypothesis of discontinuity between social positions in the social structure ( ^{Erikson & Goldthorpe, 2002} , ¹⁹⁹² ), expressed as labor segments ( ^{Doeringer & Piore, 1985} ; ^{Piore, 1979} ).

A coarse analysis of the primary and secondary labor markets shows that 27% of immigrants were employed in the primary market and 73% in the secondary market in their original countries; in Santiago, these percentages changed to 16.8% and 83.2%, respectively. That is, the occupational status of immigrants moved downward from the first to the second segment (Table 5). The causes of this structural decline are related to exogenous factors ( ^{Goldthorpe, 2010} ): the predominance of low-wage jobs in the labor market (commerce, construction, domestic services), and an entry context marked by discrimination and racism ( ^{Tijoux, 2016} ; ^{Mora & Undurraga, 2013} ). Table 6 shows that the percentages of immigrants employed in the primary and secondary segments at the time of responding to the survey were 21% and 79%, respectively; this result suggests a condition of counter-mobility, although not to the levels of the original countries.

When comparing both transitions in aggregate terms, South American immigrants were found to experience a partial recovery from the downward movement associated with the first transition. Both transitions indicate the existence of a u-shaped occupational mobility pattern ( ^{Chiswick et al., 2005} ; ^{Chiswick, 1978} ): downward mobility in the first transition and upward counter-mobility in the second transition ( ^{Aysa-Lastra & Cachón, 2013a} , ^2013b ).

Relative Occupational Mobility

The natural logarithms of the odds ratios obtained for the two occupational transitions shown in tables 5 and 6 were used to analyze the relative occupational mobility patterns, which were found to be irregular in both cases. Firstly, occupational mobility occurs within each segment and is marginal outside the segment. Fluidity can be observed in both transitions within segments, as well as social closure. Secondly, mobility is downward in the first transition and upward in the second, and it occurs mostly within individual labor segments.

Both phenomena reflect the U-shaped occupational mobility pattern, although its character is segmented, that is, it presents downward mobility in the first occupational transition and upward counter-mobility in the second occupational transition within each labor segment; this is an example of social fluidity ( ^{Erikson & Goldthorpe, 2002} , ¹⁹⁹² ), which occurs within labor segments and scarcely between segments. This finding reflects similar results in other studies ( ^{Simón, Ramos & Sanromá, 2014} ; ^{Aysa-Lastra & Cachón, 2013a} , ^2013b ) and echoes the theory of segmented assimilation ( ^{Zhou, 1997} ; ^{Portes & Zhou, 1993} ).