PREPRINT. Measuring Social Capital in Ethnic Enclaves: Indians in Lloret de Mar, and Pakistanis in Barcelona city

Transcription

1 1 PREPRINT Contribution to the edited book by Li, Yaojun (2013). The Handbook of Research Methods and Applications on Social Capital. Edward Elgar Publishers. Measuring Social Capital in Ethnic Enclaves: Indians in Lloret de Mar, and Pakistanis in Barcelona city José Luis Molina, Hugo Valenzuela-García, Alejandro García-Macías, Miranda Lubbers and Judith Pampalona egolab-grafo, Departament d'antropologia social i cultural Universitat Autònoma de Barcelona Introduction The authors, when introducing themselves in international meetings, normally get this kind remark: from Barcelona? Ah! I visited the city some time ago. I would like to visit it again. And often they note that the city center has preserved its culture so well. These visitors are usually unaware that meanwhile something changed in this latter respect. If the returning visitors walk down the Ramblas, from Catalonia Square to the harbor, turning to any street on their right will take them to a bazaar-like landscape of cellular phone shops and halal butcheries mostly run by Pakistanis, which give commercial life to the urban transformation of this part of the city center triggered by the Olympic Games in 1992 (Marshall 2000). In the case that visitors want to travel around looking for sunny beaches and fun, then visiting Lloret de Mar in the Costa Brava will give them the impression that there is a significant number of people from India in Catalonia. This is because they run virtually all the many local souvenir shops. But in fact they just represent a tinny 0.5% of the foreign population settled in Spain.

2 2 Migration in Spain is a relatively new phenomenon. After the Olympic Games the proportion of immigrants in Spain was 1,5% while in 2011 the proportion was 14% including European Union residents (National Institute of Statistics, INE 2012). Two major causes explain this dramatic growth. The first one is the strong demand for unskilled labor force in the construction sector, fuelled by cheap loans and financial speculative purposes (Carballo-Cruz 2011); the second are the six extraordinary regulation processes that have taken place in Spain since 1986, affecting at least two million migrants (INE 2010). The result of these processes is that migrants who came with a tourist visa could regularize their situation in a relatively short period of time, and enjoy a legal status afterwards. Nevertheless today the proportion of immigrants in Spain starts to show a negative tendency and it is decreasing slightly due to the current economic crisis. Many of the migrants that came from Pakistan and India respectively, who were originally from the split Punjab, worked hard to get ahead. Those with more years of residence typically acquired the Spanish nationality and nowadays some of them even own one or, in some cases, several shops. As a matter of fact, the majority started in the business as employee of their kin or other co-ethnics owners; learned how to run the business and decided to establish their own business afterwards. Most of them stated that their main goal from the very beginning was to start their own business. This striking entrepreneurial orientation has been documented by many authors as a cultural trait of some ethnic groups, especially from Asia (see Valenzuela 2010 for the Pakistani case in Barcelona). What is true as well is that they do not really have other choices because of their low levels of education, the precarious salaries caused by the dual structure of the labor market, and the high rates of unemployment and informal economy that characterize South European countries (Baycan-Levent & Nijkamp 2009). In addition, these two places are singular because of the extraordinary spatial concentration of Pakistanis and Indians it holds. Concentration is in fact a distinctive feature of the ethnic enclave in the literature (Portes & Manning 1986); that is, economic and cultural complexes that provide people from a given ethnic group in a specific location with economic advantages compared to the mainstream economy.

3 3 Although some authors claim that there is no such thing as ethnic enclaves in Spain (Haller 2004, Solé & Parella 2005, Arjona & Checa 2006), we contend that Pakistanis in the Raval and Indians in Lloret de Mar (a mix of Hindu, Sikh and Sindhi origins with a majority of Sikhs from Punjab, see below) have developed ethnic enclaves which are currently coping reasonably well with the economic crisis. The strategies are different in each case, though they do share a similar situation: increasing self competition both in a specific economic niche and in a bounded space, with low human capital requirements, small initial investment, low margins, and intensive use of co-ethnic workforce. The interactive model (Aldrich &Waldinger 1990) explains quite well the emergence of these two enclaves: taking into account the structure of opportunities and the characteristics of the group, a given strategy emerges, (self-employment in this case), mostly in the same economic niche and replicating the model of business. In this vein, ethnic enclaves have often been presented as a reasonable alternative for migrants, providing better returns than the insertion in the secondary labor market (Wilson & Portes 1980, Zhou 1998, Salaff et al. 2003, Clark & Drinkwater 2010). Conversely, some authors (Fong & Ooka 2002, Logan et al. 2003) pointed out the social costs that ethnic economies represent for migrants, preventing them from being incorporated into the mainstream economy. Indeed, empirical evidence is mixed, and it is highly debated whether (or under which circumstances) immigrants employed in ethnic economies have higher, lower or equal returns to human capital as immigrants employed in the mainstream economy (e.g., Xie & Gough 2011). These contradictory evidences can be in part overcome if we pay attention to the internal inequality of ethnic enclaves. Apart from the invisible role played by women, who in most cases take care of children and do unpaid work in the family business (Menzies et al. 2003), the differences between owners and employees are important. Sanders and Nee (1987) observed that ethnic enclaves mainly benefit enclave entrepreneurs, but not employees. The latter find at best no net differences between enclave and mainstream economies, and at worst net disadvantages. We could confirm the latter observation for ourselves in Lloret de Mar during our ethnographic fieldwork (see below for details): Established owners seemed to be in

4 4 the best situation, adapting their business successfully to the new situation. Newcomers and employees, especially those having a pattern of circular migration adapted to the touristic season, seemed to be in a worse position. In order to better understand for whom ethnic enclaves may be a way of upward mobility, researchers have underlined the importance of considering the resources that immigrants bring with them and that further shape their trajectories in enclave and mainstream economies (Nee & Sanders 2001). Immigrants with high levels of human capital are more likely to find work in the mainstream economy, whereas immigrants with lower levels of human capital are more likely to become self-employed (especially in the case of high levels of social and/or financial capital) or employed by co-ethnics (e.g., Roth et al. 2012). Apart from human and financial capital, social capital is crucial. The mixed-embeddedness theory poses that for ethnic entrepreneurs, a mix of national and co-ethnic relationships along with involvement in professional and civic associations are is necessary for managing successfully (ethnic) businesses in Europe, where regulations and controls for economic activities are larger than in other areas (Kloosterman & Rath 2001). Employees in co-ethnic businesses, on the other hand, are thought to have less and primarily co-ethnic social capital. In our case, the immigrants in the two enclaves, both owners and employees, have relatively low human capital in terms of education, although over time they gain experience in the enclave and learn the new language. We expect social capital theory (Bourdieu 1980, 1986, Portes 1988, see below) to helps us understand the differences between owners and employees such as variations in the levels of social capital enjoyed by owners and employees, generated by their role in the business and by the length of residence in the country of settlement. The number of contacts with natives, the closeness, and the type of relationships maintained with them, for instance, could make the difference. In order to test the mixed-embeddedness hypothesis we applied a social capital survey in both places expecting to find variation in the levels of social capital of both Pakistanis and Indians respectively depending on their role or hierarchical position in the ethnic businesses and the time of residence. As far as we know, there are no previous attempts in the literature to test the mixed-embeddedness hypothesis in ethnic enclaves applying a social capital perspective.

5 5 The remainder of this chapter is organized in the following manner: In the next section, we briefly describe the two ethnic enclaves; in the sections that follow we present the literature about the measurement of social capital, and the details of the survey we conducted in the two sites. Finally, we present the results and the concluding remarks. Ethnic enclaves from South Asia: El Raval and Lloret de Mar In Spain, Asians represent only a small proportion of the migrant population. The main groups are Chinese (2.48% of the migrants), Pakistanis (1.17%), Indians (0.50%), and Filipinos (0.47%; INE 2011). These migrants, apart from their geographic origin, share other central tenets: first, these communities have not had any remarkable historical, economic, political or cultural bond with Spain before. Second, these populations show a persistent diasporic pattern of migration - which could partly explain the first point. Thirdly, they share high rates of self-employment and economic achievement thanks to their tendency to conform enclave economies. In this article, we focus on Pakistanis in the El Raval neighborhood in the city of Barcelona, and on Indians in the Catalan seaside resort of Lloret de Mar. Pakistanis in El Raval, Barcelona Pakistanis in Spain are characterized by a clear pattern of geographic residential and commercial concentration. According to the last official population census (INE 2011) there is a population of almost in Spain, of which 64% live in Catalonia (over ). Almost 90% of them dwell in the region of Barcelona, and of these 60% are settled in the city of Barcelona. In the neighborhood of El Raval we find around the 20% of this last figure (> 4.000), and they constitute the largest foreign community in number. El Raval, in the old inner city, with its narrow and dark alleys, has a long history of deprivation and poverty. In fact, apart from the economic significance, the Pakistanis settlement has also meant a commercial revitalization of the area (i.e., Werbner, 2002) along with the projection and expansion of cultural references as mosques, cultural and political associations, and Urdu newspapers, among others. Urban concentration has been constant since Pakistanis started to arrive in the 1970s to the UK, southern Europe and the Gulf countries. While highly educated and

6 6 professional Pakistanis chose to migrate to the USA, Australia, Canada or the UK, unskilled workers diversified their destinations. Some landed in Spain, for example in the city of Barcelona or in mining areas like Logroño, Linares or Teruel. In 2001 Pakistanis led protest movements resulting in almost eight thousand Pakistanis being granted legal Spanish citizenship, attracting more countrymen from all around the world. Since then the Pakistani population has grown sustainably and today it represents 8% of the foreign population in the city of Barcelona (see Güell 2012). --Figure 1 about here-- These newcomers generally have low levels of formal education (68% has just attained primary education), they are mostly Sunni Muslims, and the majority are young males (average of 30 years old). Nevertheless, they do not belong to the lowest socioeconomic strata: the cost to reach Spain illegally is very high for an average Pakistani and many of them are small landowners in origin. As we already mentioned, their scarce human capital s profile funnel them to the lower jobs of the mainstream labor market: unskilled workers (laborer, factory worker, taxi driver...), hawker or other informal economic activities, sometimes combining several of them. The lack of professional skills of the first generation of Pakistani migrants from Pakistan partly explains why they show a high remarkable tendency toward self-employment and why they have specialized in small businesses. But other important reasons respond to some other cultural causes: The lack of professional skills of Pakistani immigrants of first generation partly explains why they show a remarkable tendency to self-employment and a high rate of small entrepreneurship. However, other important reasons for selfemployment respond to cultural causes: historically, certain regions from the Asian subcontinent (Punjab, Sind, Gujarat) shown a strong tendency toward migration and trade: ancient Indian traders and Indian Diaspora are well documented (see, for instance, Vertovek, 1991, 2000; Oonk, 2007; Markovitz, 2008). Furthermore, all respondents without exception expressed a deep desire to become masters of their own work. On the one hand, in a society strongly governed by principles of honor and respect for manual work, self-employment is not highly appreciated in the culture. On the other hand, selfemployment is seen as the most direct way to increase their standard of living (ie, upward social mobility), to obtain economic autonomy and freedom of decision-making

7 7 and to increase both the level of income and the volume of remittances sent - which allows them to capitalize their income sources at homeland. Today we find more than 250 family businesses ran by Pakistanis in El Raval, with a clear specialized niche (70%) in phone shops and accessories, Internet and telephone booths ( locutorios ), supermarkets, groceries and restaurants (Döner kebabs and other fast foods), hairdressers, halal butcheries, greengrocers, travel agencies and some other minor businesses (tailors, bazaars, jewelry shops, see Figure 2). Most Pakistani working in El Raval are engaged in small and medium businesses, and approximately 20% of the shopkeepers own more than one business. --Figure 2 about here-- The strategic location of the Pakistani community in the urban core has given them access to a diverse clientele (co-ethnic and native locals as well as tourists). Although some of the businesses are specifically directed at co-ethnics, most are focused on the general public. A favorable opportunity structure (lax migratory laws, and a dynamic real-estate market) in the past, coupled with an efficient use of ethnic resources (informal credit, information, mutual support, group cohesion, and a Pakistani transnational field that provides a wide co-ethnic manpower pool) explain both the pull factors and their outstanding competitiveness. It is important to remember here the paradoxical nature of the ethnic enclave economy: if on the one hand it generates a protected economic niche for a vulnerable migrant population, and on the other hand it might hinder integration in the broader social and labor context (Ram & Hillin 1994, Fong & Ooka 2002). In this context, group solidarity, patron-client relationships, loyalty, and reciprocity play a relevant role within the enclave. More importantly, the enclave offers both the expertise (mostly informal entrepreneurial know-how) and the means (contacts with suppliers) to for employees to become owners in the future, thus reproducing the particular enclave economic synergy. In fact, the current economic crisis that hits particularly vulnerable sectors of the population is not having such a dramatic effect in these economic enclaves: though some stores are closing down and many employees lost their jobs, the number of Asian immigrants seems to be increasing and the economic enclaves are

8 8 expanding because renting in a commercial spot is now much cheaper than some years ago. Indians in Lloret de Mar, Girona During the last two decades, the Indian community has dramatically increased its number in the seaside resort of Lloret de Mar and other coastal areas in Girona. Although it was an ancient inshore fishing town, Lloret has become a very popular touristic destination for Europeans and it has rapidly increased its population and facilities. Today it is the touristic destination in Catalonia that attracts the highest numbers of visitors and the annual revenues reach 430 million Euros a year (Plan Estratégico de Turismo de Lloret de Mar ). Its almost inhabitants quadruple during the summer and Lloret becomes a chaotic, crowded mass tourism setting thirsty of 3S tourism ( sun, sea and sand ) that offers a stereotyped and commoditised pack of (supposedly) Spanish culture: parties, paella, sangria, flamenco, bull fighting and souvenirs (Jiménez & Prats 2006). Night parties, alcohol excess and discotheques make up the rest of the picture. Historically, ancient Indian traders and merchants settled in free ports according to their ethnic specialty. That explains their initial presence in Gibraltar, Andorra or the Canary Islands (López-Sala 2007). During the 1970s they were active merchants in the textile sector, jewelry, electronics and souvenirs in Barcelona. After the Olympic Games, some of them moved to other places in search of new opportunities and less saturated markets. Understandably, Pakistanis and Indians present some similarities in their respective migratory pattern, their diasporas being spread all around the world through consecutive waves (Jayaram & Atal 2004, Oonk 2007, Vuddamalay 2010, Markovits 2008, López Sala 2007). In fact there are no major differences between Pakistani and Indian enclaves, apart from the specific economic niche in which they operate: a more formal entrepreneurial orientation in the case of the Indian owners (Moodod 1998), and particular factors (religion and the cast system) that might determine some economic practices and choices e.g. halal butcheries or the avoidance of dealing with pork meat or spirits in the case of Pakistanis.

9 9 There are registered Indians in Spain, and approximately reside in Lloret de Mar (Census 2011, Idescat), forming already a 9% of the total foreign population of the municipality. In Lloret de Mar, Indians run more than 100 shops and their growing settlement is already known in town as Little India. Indians in Lloret de Mar generally practice Hinduism and Sikhism and are ethnically Sindh, Punjabis and Sikhs. Like the Pakistani counterpart, most of them are young males from suburban and rural areas. They also present a remarkable spatial concentration of shops and stores, but they do not reside in the commercial area in which they operate because housing is not common downtown, and cheap apartment renting is easy to find in the periphery in a touristic area like Lloret particularly when the summer season is over. Indians quickly replaced other migrants (particularly Moroccan settlers) and native competitors, who had to close down or transfer the business. But open rivalry in the souvenir traditional sector has fueled animosity and racism against them from native shopkeepers. The souvenir sector is a seasonal activity. Shops are opened for just nine months and during the rest of the year Indian employees tend to go back to India, thereby establishing a circular migration pattern. A souvenir is in fact a very unspecific commodity (one might find under this category from soccer T-shirts to postcards, castanets or figurines) and its sale does not require any specific expertise or even experience. Innovation is difficult because tourists do not demand very sophisticated items, but when some successful innovation is introduced in this sector, it is quickly plagiarized. Since the rental of the commercial locations are expensive and the souvenir is by definition a trinket, a cheap item, profit is highly dependent on the volume sold. Here the relationship with the supplier makes the difference in terms of discounts or deferred payment between small shopkeepers and those who own or run several shops at the same time. And this difference is not trivial at all, since the latter usually are socially well integrated in the locality, and their personal networks show a high degree of embeddedness with both local society and clusters of co-ethnics. Employees, on the other hand, have similar traits as the Pakistani workers

10 10 described above: they are hardly integrated within the local society because their working-time schedules are highly demanding, thus they show a very poor command of the local language. Sometimes these young workers sleep, eat and spend their spare time together in at the back of the shop or on the top floor of commercial premises. Measuring social capital in ethnic enclaves The concept of social capital The concept of social capital, originally formulated by Bourdieu (1980, 1986) as a dimension of social class stratification, has been applied to the comprehension of different realities, and levels of analysis (Portes 1988), especially since its adoption by the World Bank as a tool for planning economic development (Woolcock 1998). Brunie (2009) identifies at least three conceptualizations of social capital, i.e., as a collective, an individual, and a relational asset. The first focuses on the values and norms, institutions, and organizations shared by individuals that enable the coordination and cooperation in a given territory (Putnam 1993, 2000). The second focuses on social capital as an individual property, based on individuals predisposition to trust, cooperate and empathize (Newton 1997, Uslaner 1998). Finally, the third conceptualization focuses on the resources embedded in social structures which are accessed through social relationships (Lin 2001; Burt 2000). Although all three approaches to social capital rely on social networks, the latter is considered the network or structural approach to social capital, which is adopted here. Measuring social capital Three methods have been used for measuring relational social capital: the name generator, the resource generator, and the position generator (van der Gaag et al. 2008). The name generator method (Lin 1999) poses one or more questions about the contacts of the focal individual ("ego"). The list of persons enacted ("alters") is normally small (5-10). Once a list of names has been obtained, the method typically collects information about the relationship between ego and alter (e.g. closeness, frequency of interaction, contents of the relationship), as well as information about the attributes of each alter (e.g., gender, location, education, occupation), and whether each pair of alters

11 11 knows each other. From this information it is possible to obtain a proxy of the social capital accessed by ego, although several problems arise, such as the influence of the specific name generator used and the bias to recall strong ties or geographically close ties (Campbell & Lee 1991). The resource generator method (Snijders 1999, Van der Gaag & Snijders 2005) measures social capital through a roster of resources along with the closeness to the person who has the capability to provide them. This approach is realistic in the sense that it measures directly the individual's capability to obtain resources embedded in social networks what social capital is supposed to be. For specific resources, like getting a job, the resource generator is especially useful (Verhaeghe et al. 2012). Nevertheless, for generic purposes, the list of resources can vary at on a greater extent than the list of professions, which made us to prefer the position generator approach for this case. The method that is most widely used to measure social capital is the position generator (Lin & Dumin 1986, Lin 2001). This method uses a sample of the professions, occupations or authorities of a given society as a proxy of the hierarchical distribution of resources. These positions can be ranked on the basis of the socioeconomic status attached to them following national surveys or internationally standardized scales of occupations like SIOPS or ISEI, and these prestige scores are taken as proxies of the resources accessed by them. Respondents are presented with a list of occupations (typically between 18 and 30, where some of them are extracted from the local context), and they are asked whether they know anyone in each of these occupations. In the case of a positive response, respondents are typically also asked whether this person is a family member, a friend, or an acquaintance. Usually, if the respondents know multiple persons in an occupation, only the person with whom he or she has the strongest relationship is recorded. Other attributive variables of the alters can be collected as well. With this information, various measures of social capital can be obtained: 1) The number of different positions accessed; 2) The total prestige accessed, summing the prestige scores of the occupations to which the respondents gave positive answers; 3) The average prestige accessed, as the total prestige divided by the number of different positions; 4)

12 12 The span in prestige between the highest and the lowest positions accessed (range), and 5) The maximum prestige, i.e., the value of the highest position accessed (Lin, Fu & Hsung 2001). We adopted this latter method for several reasons. First, this method is widely used at the international level, which enables cross-cultural comparisons. Second, the structure of occupations can be objectively drawn from the extant literature, which makes the method easy to apply and interpret, compared to the resource generator. Third, the questionnaire can be administered in less than 10 minutes and makes sense for the interviewees, who tend to answer the questions provided without hesitation. Last, but not least, this method is intended to measure at the individual level the social resources that can be accessed and could be mobilized within a given social structure, which allows us precisely to understand the role of ethnic enclaves in the mainstream economy. The survey The survey was designed following the model by Chen and Wellman (2009) applied to ethnic entrepreneurs in Canada and China, which included a roster of 18 occupations as position generators. We adapted the method to allow nominations of transnational contacts (Chen & Tan 2009). In order to select the occupations for the Spanish context, we relied on the Spanish occupational prestige scale named PRESCA2C (Carabaña & Gómez 1996). Based on previous ethnographic fieldwork in Lloret de Mar, we built a wide roster of occupations already mentioned by informants and selected 18 occupations that were well distributed along the prestige scale (see Figure 5). Some occupations were specifically selected in order to take into account the local context, e.g. for Lloret de Mar fishermen (employed by others), travel agents, taxi drivers, waiters, street vendors, and hotel cleaning personnel (see Table 1 in the Appendix). The occupational prestige based on the PRESCA2C scale for these 18 occupations correlated 0,90 with the SIOPS prestige scores for the same occupations and 0,84 with the ISEI prestige scores. In the survey, the occupations were presented in a random order. --Figure 5 about here

13 13 For each occupation, respondents were asked whether they knew someone in that position. In case of a positive response, several attributes of the relationship were measured: the type of relationship (1- acquaintance, 2- friend, 3-kin), gender, place of residence (1-same neighbourhood, 2-in Lloret de Mar, 3- in Catalonia, 4- in Spain, 5- Other country, specifying town and country), duration of the relationship (1-less than 1 year, 2-from 1 to less than 5 years, 3- from 5 to less than 10 years, 4-10 to less than 20 years, 5-20 or more years), and emotional closeness (on a scale from 0 to 4, with higher values representing more intimate contacts). We also determined the country of origin (Spain or elsewhere), although this variable has 24 missing values. With this design we conducted a pilot test, and after small corrections we administered the survey to the selected cases. For the case of El Raval, we applied a 30-position instrument, which maintained the original 18 occupations of Lloret de Mar for comparative purposes, and it included 12 other occupations that took into account the more extended occupational diversity in Barcelona city, including architects, medical doctors, lawyers, care service workers, and construction workers (see Table 2 in the Appendix). The same attributes were measured as well. Again, a pilot test was conducted, which led to minor corrections. For the present chapter we only used the data based on the 18 occupations that coincided with the measurement in Lloret de Mar to maximize comparability. Although this decision may imply that the occupations are less representative of the social structure in El Raval, results for El Raval turned out highly similar when the calculations of social capital were based on all 30 occupations. Participants The survey for Lloret de Mar was administered to all shops in the main and major commercial street of the town center (N=50, of which 34 owned by Indians). The vast majority of the shops are souvenir shops. The results presented in this chapter are based on the 34 interviews with Indian shop owners (14) and employees (20). Only two of the respondents were women (which account for the 30% of the total Indian population in this place, IDESCAT 2012). Age varied between 18 and 64, with the

14 14 average being 33. Owners were on average 8 years older (M = 37,6; SD = 5,9) than employees (M = 29,3; SD = 10,1), but the variation in age within each group is large. In El Raval, the survey was administered to all businesses run by Pakistani owners in the main commercial axes of El Raval (N= 50 interviews, 8 owners and 42 employees). Of all respondents, only one was female (men are four times more frequent than women in this collective, IDESCAT 2012). Age varied between 17 and 56, and on average respondents were 32 years old. Owners were on average about 11 years older (M = 30,0; SD = 8,0) than employees (M = 40,6; SD = 11,1), but again the variation in age within each group is large. The businesses included 12 mobile phone shops, 10 bars and restaurants, 7 supermarkets and groceries, 6 hairdressers, 5 internet and telephone booths, 2 souvenir shops, and various other stores (a music store, a travel agency, a video shop, a refrigerator store, a launderette, a newsstand, a company specialized in sending remittances, and a franchise courier company). Results Descriptive results The results of comparing the social capital based on the 18 position generators in both places show that Pakistanis in El Raval enjoy higher levels of social capital than Indians in Lloret de Mar (see Table 3). In particular, they know people in more positions, the average prestige is slightly higher 1, and the maximum prestige is higher. That might be due to the location of the ethnic enclave, which is placed just in the old city centre of Barcelona. --Table 3 about here Locality Total prestige total accessed Number of positions accessed Mean prestige accessed Max. prestige accessed Range prestige accessed Lloret de Mean 645,6 5,8 107,1 171,5 104,8 Mar SD 405,5 3,1 16,8 58,2 65,3 El Raval Mean 895,9 7,6 115,2 186,6 119,3 1 As the samples were not random, and cases were not independent, we do not present statistics for testing the differences among proportions or averages. Nevertheless, the statistics were calculated in each case to guide our understanding of the size of differences.

15 15 SD 535,4 4,0 21,8 55,7 59,6 Table 3. Social capital in the ethnic enclaves in Lloret de Mar and El Raval. Another aspect that differed across the two sites is the geographical distribution of the alters known to the respondents (See Table 4). It is interesting that social capital in the case of Indians in Lloret de Mar is basically local (81,3%), i.e. based in the same neighbourhood, with only 6,5% of prestige accessed through transnational contacts. Conversely, Pakistanis in El Raval have local contacts in a wider area Barcelona city, and a remarkable proportion of transnational contacts in the positions studied (22,9%). In other words, while the Pakistanis in El Raval named more positions than the Indians in Lloret de Mar, a larger part of them were transnational. Of all the local ties that were mentioned in both locations, it is also noteworthy that the respondents in Lloret de Mar nominated a higher proportion of Spaniards (38%) than the people in El Raval (26%). --Table 4 about here-- Alter s residence Same neighbourhood as ego Lloret de Mar El Raval N % Prestige N % Prestige % % Alters accessed Alters accessed ,3 83, ,1 37,3 Same town as ego 23 11,6 9, ,3 27,5 Catalonia 3 1,5 1,2 27 7,1 6,9 Spain 0 0,0 0,0 7 1,8 1,4 Ego s country of 8 4,0 4, ,5 22,6 origin Other country 3 1,5 1,6 13 3,4 3,5 Missing 0 0,0 0,0 3 0,8 0,8 Total ,0 100, ,0 100,0 Table 4. Residence of alters, and prestige accessed for both places. Another attribute of alters that differed across the two sites is the proportion of male alters. Alters who were mentioned in response to the position generators were mostly male in both sites, but this tendency was even stronger in El Raval (88%) than in Lloret de Mar (80%). Other alters characteristics did not differ greatly across the two sites. In both sites, the vast majority of the alters were acquaintances (42%) or friends (45%) - family members (13%) were mentioned less often. The average time in terms of knowing the alters was similar for the two sites (corresponding to the category "between 5 and 10 years"), with around 45% of the alters in both sites known between 1 and 5

16 16 years (M = 2,8, SD = 1,0 for Lloret de Mar; M = 2,8, SD = 1,1 for El Raval). The average emotional closeness was also positioned in the middle of the scale for both sites, although the dispersion of the responses was somewhat higher in El Raval than in Lloret de Mar (M = 2,9, SD = 1,3 for Lloret de Mar; M = 3,1, SD = 1,5 for El Raval). Differences in social capital between owners and employees For our hypothesis that social capital is unevenly distributed among owners and employees despite their common origin and socioeconomic level, we found weak support (see Table 5). More concretely, the table shows that in both sites owners know people in more occupational positions than employees, stressing that they have a more diverse pool of contacts. Nevertheless, the average occupational prestige of the people they know is more or less equal as that of employees. Similarly, owners and employees do not differ much in the maximum prestige of their contacts, nor in the range of prestige accessed, although the latter is somewhat higher in El Raval. As a consequence of the higher number of positions accessed by owners, the total prestige accessed by owners is larger than that of employees. --Table 5 about here-- Locality Position Total prestige accessed Number of positions accessed Mean prestige accessed Max. prestige accessed Range of prestige accessed Lloret de Mar Employee 576,3 5,4 106,3 170,4 101,7 Owner 744,6 6,5 108,1 173,1 109,2 Total 645,6 5,8 107,1 171,5 104,8 El Raval Employee 838,1 7,2 114,3 183,4 116,0 Owner 1199,8 9,9 120,3 203,6 136,6 Total 895,9 7,6 115,2 186,7 119,3 Table 5. Social capital of owners and employees in both enclaves. In order to explore the role of the length of residence, we defined four categories instead of two: (1) recent immigrants or employees with a pattern of circular migration, (2) employees with more than 5 years of residence, (3) new owners (with less than 3 years in this status), and (4) established owners, with more than 3 years in this status

17 17 (see Table 6). The differences between the categories are now clearer. For El Raval, the table shows that the number of positions accessed increases as expected along these four classes. In general, recent migrants have least social capital, and established owners have most social capital. Interestingly, new owners deviate a bit from the overall pattern, because the mean prestige, the maximum prestige and the range of prestige accessed is more or less equal to that of employees. In Lloret de Mar, this deviation for new owners is even more pronounced. Here, new owners do not only have a lower mean prestige, maximum prestige and range of prestige than the other categories, but they also access less positions. --Table 6 about here

18 18 Locality Occupational classification Total prestige accessed Number of positions accessed Mean prestige accessed Max. prestige accessed Range of prestige accessed N Lloret de Mar Employees, recent or circular migrants Employees, > 5 years of residence 504,3 4,9 102,3 148,8 76, ,1 5,6 108,5 182,0 115,2 13 New owners 412,7 4,3 97,1 122,6 51,1 4 El Raval Established owners Employees, recent or circular migrants Employees, > 5 years of residence 877,3 7,4 112,6 193,3 132, ,5 7,0 114,1 178,9 111, ,1 7,3 114,5 189,0 121,2 19 New owners 935,0 8,3 113,8 183,2 117,2 4 Established owners 1464,5 11,5 126,9 224,0 156,0 4 Table 6. Average social capital among employees who recently migrated or who have circular migration, employees with more than 5 years of residence, new owners, and established owners. NOTE: cells depicted in grey are based on 5 or less observations. This classification also makes sense because we argued before that one of the main dynamics of an enclave is the tendency of experienced employees to open new businesses in the same economic niche. This gradual access to the social capital needed for running a business is more or less found in El Raval, but not in Lloret de Mar for the case of new owners. It is important to note that the number of new owners in the two

19 19 samples is small, so we need to have some caution with the interpretation of the results. In Lloret de Mar, the new owners were in fact newcomers to the enclave, with short previous experience in the enclave or no experience in it at all. In one case, a new owner worked several years in factories in order to accumulate the needed capital, which explains his lack of contacts. In two other cases, the capital was provided by people in India, without strong ties with the enclave. It seems unlikely in the present situation that the latter will be able to pay back the money they borrowed for starting a business. Except for these cases, the pattern seems to apply reasonably well. To determine whether owners showed a higher mixed embeddedness in terms of social capital than employees, we considered the country of origin and country of residence of the alters in our analysis. On the basis of these two variables, we distinguished three types of alters: Spanish (both the country of origin and the country of residence is Spain), co-national alters (the origin is Pakistan / India, the country of residence is Spain), and transnational alters (those for whom the country of residence is not Spain) and explored the different roles these alters had in the social capital of the respondents. In both sites and in all occupational categories, the majority of all positions were accessed through co-nationals in Spain. However, we found that owners access more positions through Spaniards than employees, and their mixed embeddedness is therefore greater. Employees maintained resourceful relationships with both Spaniards and with co-nationals. The only exceptions are new owners in El Raval, who access very few positions through Spaniards. Transnational contacts are more often mentioned by recent employees, perhaps because they do not know local people with these occupations, and especially (as we had already mentioned) in El Raval. Furthermore we explored the average emotional closeness the respondent felt toward the three groups of alters. In general, respondents felt closest to their transnational ties (which appeared to have the largest proportion of family ties, i.e. 34% compared to 11% among co-ethnics residing in Spain), and least close to the Spaniards. Again, in the latter respect, we found a small advantage of established owners, who had a greater affinity with the Spaniards. Lastly, we explored the average prestige scores of the three groups of alters. In general, transnational ties had the highest prestige scores (it is likely that transnational

20 20 ties are mentioned more often when local ties to specific positions aren't available), followed by Spaniards; the occupations with the lowest prestige tended to be accessed through co-nationals. Again, we see that established owners are more embedded within the Spanish society, since they name more Spaniards with respect to occupations with a higher average prestige as compared to employees and new owners. Some cautions need to be taken with these average results though. First, as the table indicates, some cells were based on only a few observations, in which cases the averages might be affected by one or two cases with somewhat extreme values. Second, Figure 6 shows the distribution of the average prestige of Spaniards, and even though the median scores differ quite a bit among the categories, the distributions also show great variation and much overlap. In other words, whereas on average the respondents tended to access positions with higher prestige through Spaniards and with somewhat lower prestige through co-nationals, and whereas in Lloret de Mar three of the four highest occupations were accessed exclusively through Spaniards, most positions - both low and high - were accessed through all types of alters, although in different degrees. --Table 7 about here

21 21 Locality Occupational classification Average proportion of positions accessed through... a Average closeness with alters who are... b Average prestige scores of alters who are... b Spaniards co-ethnics in Spain transnational contacts Spaniards co-ethnics transnational contacts Spaniards co-ethnics transnational contacts Lloret de Mar Employees, recent or circular migrants Employees, > 5 years of residence 12,4 75,7 11,9 2,8 3,2 4,0 144,2 98,6 110,0 26,3 68,8 4,9 2,0 3,0 4,2 114,0 106,0 133,7 New owners 37,2 62,8 0,0 2,1 3,0-102,5 93,4 - Established owners 40,5 55,6 3,9 2,6 3,2 4,0 126,1 101,5 229,6 El Raval Employees, recent or circular migrants Employees, > 5 years of residence 20,5 52,7 26,9 2,2 2,7 3,2 98,8 102,7 153,2 14,0 66,0 20,0 2,9 3,2 3,9 116,2 104,1 163,8 New owners 12,5 78,1 9,4 2,0 3,1 2,0 131,8 107,3 151,3 Established owners 30,6 52,1 17,4 3,4 3,8 4,7 148,7 94,3 154,3 Table 7. Differences between the four occupational categories in Lloret de Mar and El Raval in the role of Spaniards, co-ethnics in Spain, and transnational contacts (including contacts in the country of origin) in their social capital. NOTE: cells depicted in grey are based on 5 or less observations. a Data were first aggregated to the individual level. a Results are reported at the alter level.

22 22 Conclusions and discussion We found weak support for the mixed embeddedness hypothesis in our study. However, in our ethnographic fieldwork, based on semi-structured interviews and participant observation, this hypothesis seemed to be supported. Established owners seemed to be embedded in both the Spanish context and among their co-ethnics: Their personal networks exhibited a mix of national and co-ethnic contacts, and they participated in civic and professional associations. Employees, especially those having a pattern of circular migration adapted to the touristic season, seemed to be in a worse position. Nevertheless, when the hypothesis was tested with the position generator methodology, we only found weak support to it. Especially, new owners (being an owner for less than three years) showed a low level of mixed embeddedness. The lack of strong variation in the data could be interpreted in another way: the ethnic enclaves that we are measuring are stagnation markets at the bottom of the economic system, with low levels of overall social capital (i.e., with low access to the resources distributed all over the social structure). The differences observed by ethnographers and the actors themselves should be interpreted in this context. Thus, the results are twofold. On the one hand, the position generator does not capture neatly the differences between owners and employees observed during the ethnographic fieldwork. On the other hand, the position generator methodology reveals its capacity to effectively measure social capital taking into account the global social structure. Recent theory about ethnic economies (Kloosterman 2010) helps us to interpret the results as the confinement of these ethnic enclaves in stagnating markets with low levels of human capital. In addition, it is worth mentioning that low levels of social capital at the individual level do not seem to be a problem for coping with the economic crisis better than other sectors better capitalized in all regards. Enclaves have economies of scale, and synergies that still need to be well understood. We are well aware of the weaknesses of the sample, which was limited to selfemployed immigrants and employees of co-ethnics in the main commercial streets in both locations. Among others, a comparison with co-ethnics who found work outside of the enclaves is not possible. A more extensive survey is needed in order to make stronger conclusions about the two enclaves. Also, to understand the causal mechanisms

23 23 behind the observed relations, it would be of interest to follow employment trajectories and social capital of the respondents longitudinally. Finally, the position generator tool does not provide information about the social links among the people nominated by the occupations, information that could add new dimensions to the measures in terms for instance of bonding or bridging ties (Putnam 2000). On the other hand, as far as we know, our study represents the first attempt to measure social capital in ethnic enclaves, and the results are of interest. Acknowledgments This research is part of the project Profiles of Ethnic Entrepreneurship ( ). Perfiles del Empresariado Étnico en España, MICINN (CSO ). A preliminary version of this paper was presented in the Seminario Internacional. Sociedad y Empresariado. Desafíos locales en una economía globalizada, August 21-22, 2012, Universidad Autónoma de Aguascalientes. We would also like to thank Berta Güell for sharing her knowledge about El Raval and for her support administering the survey. Gabriel Blinder collected reliable data in Lloret de Mar. Finally, we want to thank our informants in both enclaves for their patience and kindness.

24 24 References Aldrich, H. E. and R. Waldinger (1990), Ethnicity and entrepreneurship, Annual Review of Sociology, 16, Arjona, A. and J.C. Checa (2006), Economía étnica. Teorías, conceptos y nuevos avances, Revista Internacional de Sociología, 64 (45), Baycan-Levent, T. and P. Nijkamp (2009), Characteristics of migrant entrepreneurship in Europe, Entrepreneurship and Regional Development, 21(4), Bourdieu, P. (1980), Le capital social: notes provisoires, Actes de la Recherche en Sciences Sociales, 3, 2-3. Bourdieu, P. (1986), The forms of capital, in J.G. Richardson (ed.), Handbook of Theory and Research for the Sociology of Education, New York. Greenwood Press, Brunie, A. (2009), Meaningful distinctions within a concept: Relational, collective, and generalized social capital, Social science research, 38 (2), Burt, R.S. (2000), The network structure of social capital, Research in Organizational Behavior, 22, Campbell, K. E. and A. L. Barrett (1991), Name generators in surveys of personal networks?, Social Networks, 13 (3), Carabaña, J. and C. Gómez (1996), Escalas de prestigio, Cuadernos Metodológicos, 19, Madrid. CIS. Carballo-Cruz, F. (2011), Causes and Consequences of the Spanish Economic Crisis: Why the Recovery is Taken so Long? Panoeconomicus, 3, Chen, W. and B. Wellman (2009), Net and Jet: The Internet Use, Travel and Social Networks of Chinese Canadian Entrepreneurs, Information, Communication and Society, 12 (4). Chen, W. and J. Tan (2009), Understanding Transnational Entrepreneurship through a Network Lens: Theoretical and Methodological Considerations, Entrepreneurship Theory and Practice, 33 (5),

25 25 Clark, K. and S. Drinkwater (2010), Patterns of ethnic self-employment in time and space: evidence from British Census microdata, Small Bussiness Economy, 34, Coleman, J. (1988), Social capital in the creation of human capital, American Journal of Sociology, 94, Fong, E. and E. Ooka (2002), The social consequences of participating in the ethnic economy, International Migration Review, 36 (1), Güell, B. (2012), Els negocis dels pakistanesos al barri del Raval a Barcelona. Motius i estratègies per a l autoocupació. Unpublished master thesis in Sociology. University of Barcelona. Haller, D. (2003), Place and Ethnicity in two Merchant diasporas: a comparison of Sindhis and Jews in Gibraltar, Global Networks, 3 (1), IDESCAT (2012a). Estadística sobre la població estrangera a Catalunya. [accessed: ]. IDESCAT (2012b). Estadística sobre la població estrangera a Catalunya [accessed: ]. Jayaram, N. and Y. Atal (eds) (2004), The Indian Diaspora: dynamics of migration, London. Sage. Jiménez, S. and LL. Prats (2006), El turismo en Cataluña: evolución histórica y retos de futuro, Pasos, Revista de Turismo y Parimonio Cultural, 4 (2), Kloosterman, R.C. and J. Rath (2001), Immigrant entrepreneurs in advanced economies: Mixed embeddedness further explored, Journal of Ethnic and Migration Studies, 27 (2), Kloosterman, R.C. (2010), Matching opportunities with resources: A framework for analysing (migrant) entrepreneurship from a mixed embeddedness perspective, Entrepreneurship & Regional Development: An International Journal, 22 (1), Lin, N. and M. Dumin (1986), Access to occupations Through Social Ties, Social Networks, 8 (4),

26 26 Lin, N. (1999), Social Networks and Status Attainment, Annual Review of Sociology, 25, Lin, N. (2000), Inequality in Social Capital, Contemporary Sociology, 29 (6), Lin, N. (2001), Social Capital: A Theory of Social Structure and Action. Cambridge. Cambridge University Press. Lin, N., Y. Fu and R. Hsung (2001), The position generator: Measurement techniques for investigations of social capital, in Social Capital: Theory and Research., New York. Aldine de Gruyter, Logan, J. R., R.D. Alba, and B.J. Stults (2003), Enclaves and Entrepreneurs: Assessing the Payoff for Immigrants and Minorities, International Migration Review, 37, López, A.M. (2007), Migración internacional, vínculos transnacionales y economía étnica. El caso de la comunidad indostánica en Canarias, Migración, transnacionalismo y empresariado asiático en España. Revista CIDOB d Afers Internacionals, 78, Markovitz, C. (2008), Merchants, traders, entrepreneurs. Indian Business in the colonial era, New Delhi. Longman. Marshall, T. (2000), Urban Planning and Governance: Is there a Barcelona Model?, International Planning Studies, 5 (3), Menzies, T.V., G.A. Brenner and L.J. Filion (2003), Social capital, networks and ethnic minority entrepreneurs: Transnational entrepreneurship and bootstrap capitalism, in Etemad, H and R. Wright (Eds.), Globalization and Entrepreneurship: Policy and Strategy Perspectives. Northampton, MA. Edward Elgar Publishing. Molina, J.L., H. Valenzuela, A. García-Macías, M. J. Lubbers, J. Pampalona & J. Lerner. "The ethnic enclave of Lloret de Mar: a personal network approach to investigate mixed embeddedness and transnational fields". Global Networks, submitted on September 4, 2012.

27 27 Moodod, T., H. Metcalf and S. Virdee (1998), British Asian entrepreneurs: cultural and Opportunity Structures, in Taylor-Gooby, P. (ed.) Choice and Public Policy. London. St. Martin s Press, Nee, V., and J. Sanders (2001), Understanding the Diversity of Immigrant Incorporation: A Forms-of-Capital Model, Ethnic and Racial Studies 24(3), Newton, K. (1997), Social capital and democracy, American Behavioral Scientist 40 (5): Oonk, G. (ed) (2007), Global Indian Diaspora. Exploring Trajectories of Migration and Theory, Amsterdam. Amsterdam University Press. Portes, A. and R. Manning (1986), The Immigrant Enclave: Theory and Examples in Nagel, J. and T. Olzak (eds.), Competitive Ethnic Relations, Orlando. Academic Press. Portes, A. and L. Jensen (1987), What's an ethnic enclave? The case for conceptual clarity, American Sociological Review, 52 (6), Portes, A. (1998), Social Capital: Its Origins and Applications in Modern Sociology, Annual Review of Sociology, 24, Putnam, R.D. (1993), Making Democracy Works. Civic Traditions in Modern Italy, Princenton. Princenton University Press. Putnam, R.D. (2000), Bowling Alone: The Collapse and Revival of American Community, New York. Simon & Schuster. Ram, M. and G. Hillin (1994), Achieving break-out : developing mainstream ethnic minority businesses, Journal of Small Business and Enterprise Development, 1 (2), Roth, W. D., M.-D. L. Seidel, D. Ma & E. Lo (2012), In and Out of the Ethnic Economy: A Longitudinal Analysis of Ethnic Networks and Pathways to Economic Success across Immigrant Categories, International Migration Review 46 (2), Salaff, J.W., A. Greve, S.L. Wong and L.X. Li Ping (2003), Ethnic Entrepreneurship, Social Networks, and the Enclave, in B. Yeoh, T.C. Kiong and M.W. Charney, Approaching Transnationalism: Transnational Societies, Multicultural Contacts, and Imaginings of Home, Boston. Kluwer Academic Publishers.

28 28 Sanders, J. M. and V. Nee (1987), Limits of Ethnic Solidarity in the Enclave Economy, American Sociological Review, 52 (6), Snijders, T.A.B. (1999), Prologue to the Measurement of Social Capital, La Revue Tocqueville, 20, Solé, C. and S. Parella (2005), Los negocios étnicos. Los comercios de los inmigrantes no comunitarios en Cataluña, Barcelona. Fundació CIDOB. Uslaner, E. (1998), Social capital, television, and the mean world : Trust, optimism, and civic participation, Journal of the International Society of Political Psychology, 19 (3): Valenzuela, H. (2010), Pecunia Ex Machina, el emprendedor pakistaní en la ciudad de Barcelona, Revista CIDOB d Afers Internacionals, 92, Van der Gaag, M.P.J. and T.A.B. Snijders (2004), Proposals for the measurement of individual social capital in Flap, H.D. and B. Volker (eds.), Creation and returns of Social Capital, London. Routledge, Verhaeghe, P.P., Y. Li and B. Van de Putte (2012), Socio-economic and Ethnic Inequalities in Social Capital from the Family among Labour Market Entrants, European Sociological Review, first published online March 15, 2012 doi: /esr/jcs047. Vertovec, S. (2000), The Hindu Diaspora. Comparative Patterns, London, New York: Routledge. Vertovec, S. (ed.) (1991), Aspects of South Asian Diaspora, Delhi: Oxford University Press, Papers on India, Vol. 2 Vuddamalay, V. (2010), Empresarios indios en Francia. Una aproximación inicial a sus actividades en un contexto en rápida evolución, Revista CIDOB d Afers Internacionals, Barcelona, 92, Werbner, Pnina (2002), The Migration Process: Capital, Gifts and Offerings among British Pakistanis (Explorations in Anthropology), London. Berg Publishers

29 29 Wilson, K.L. and A. Portes (1980), Immigrant Enclaves: An Analysis of the Labor Market Experiences of Cubans in Miami, American Journal of Sociology, 86 (2), Woolcock, M. (1998), Social Capital and Economic Development: Toward a Theoretical Synthesis and Policy Framework, Theory and Society, 27 (2), Xie, Y. and M. Gough (2011), Ethnic Enclaves and the Earnings of Immigrants, Demography, 48, Zhou, Y. (1998), Beyond Ethnic Enclaves: Location Strategies of Chinese Producer Service Firms in Los Angeles, Economic Geography, 74, Zhou, M. (2004), Revisiting Ethnic Entrepreneurship: Convergencies, Controversies, and Conceptual Advancements, International Migration Review, 38 (3), Figures Figure 1. Pakistanis shops in El Raval, Barcelona.

30 30 Figure 2. Small businesses ran by Pakistanis in El Raval, Barcelona (Source: Güell 2012, reprinted with permission).

31 Figure 3. Shop owned by an Indian in Lloret de Mar, province of Girona. 31

32 32 Figure 4. Shops in the main commercial street in Lloret de Mar (black ones run by Indians).

33 33 Figure 5. Distribution of prestige among occupations selected (in the survey they appear in a randomized order).

34 Figure 6. Distribution of the prestige of positions accessed through Spaniards. The box plots show among others, the lower quartile, the median, the upper quartile and extreme values. 34