Chameleon brokers: A translocal take on migration industries in the Thai-Swedish wild berry business

Abstract

Migrant brokers constitute a substantial node in the industries that underpin contemporary global migration processes, including seasonal labour migrants in agri-food businesses. This article adds a translocal perspective to the role of migrant brokers, while emphasising the multi-sited embeddedness of brokers in sending and receiving countries, and their role in sustaining transnational migration flows. The example of the Swedish wild berry industry shows how two groups of translocal brokers operate in multi-sited space, first, Thai women brokers residing in rural Sweden, and second, local brokers, residing in rural Thailand. This article emphasises how translocal brokers are giving migration industries access to multi-sited embeddedness, both at the site of recruitment in Thai villages and at the site of work in Sweden. The translocal embeddedness is noticed in how moral economies and trust are at play in recruitment processes, and how moral economies are then transferred across space to the site of work. Also, it accentuates how translocal brokers are main subjects, in how their biographical histories are creating translocal relations across space. Lastly, we show how spatial divisions of labour are creating social hierarchies among workers, where the brokers themselves incorporate shifting, ‘chameleon’ roles in multi-sited space. The analysis brings the moral complexity of brokers to the surface, while showing how the social relations of their ‘moral economies’ are commodified within profit-seeking migration industries.

1. Introduction

The crisis following a pandemic such as the COVID-19 (Coronavirus disease 2019) virus highlights in a cruel way the dependency on foreign labour of the global food industry. Integral to the idea of a global countryside (Woods 2007), such international migration flows have during the last decades significantly entered and reshaped rural spaces, not at least in global agri-food businesses (Bonanno and Cavalcanti 2014; Rye and Scott 2018; Rye and O’Reilly 2020). The channels of seasonal workers that underpin contemporary food industries are crucially mediated by migration industries. The role of profit-making actors in channelling migration processes has until recently been much overlooked in contemporary migration studies (Gammeltoft-Hansen and Nyberg Sørensen 2013; Cranston, Schapendonk and Spaan 2018; Zhang and Axelsson 2021). This article deals with a central link in this chain of actors, namely, the role of migrant intermediaries and local brokers (Lindquist, Xiang and Yeoh 2012; Hernández-León 2013; Lin et al. 2017). In their position between actors in the migration industry and the workers ‘on the ground’, we argue that a translocal take, related to the moral economies of migration brokerage, adds crucial dimensions to the understanding of migration industries.

A translocal perspective on migration emphasises local-to-local relations and the embeddedness of migrant actors at multiple sites in the migration processes (Guarnizo and Smith 1998; Brickell and Datta 2011; Hedberg and Carmo 2012). In relation to brokerage, the translocal dimension is central, primarily since it highlights how translocal brokers are giving migration industries direct access to social embeddedness in multi-sited localities. Migrant brokers are operating ‘from below’, in the very villages of the migrant workers, where they are grounded in moral economies based on social networks (Scott 1976; Rigg, Bouahom and Douangsavanh 2004). The local embeddedness is anchored in the brokers’ personal biographies (Yeoh et al. 2017), such as their histories of migration and work.

Within translocal brokerage, what also travels across space, is the spatial divisions of labour (Massey 1995; Massey 2004). A translocal spatiality is visible in the divisions of labour, through the stretching of social relations across space and the multi-sited embeddedness of work, both in local/regional, national, and transnational senses. Crucial to this article, the translocal perspective gives that the brokers are not only socially embedded in the place where they recruit workers, but simultaneously also in other places, primarily the site of work for which they are brokering. Through their biographies of work and residency, they possess multi-sited knowledge about the nature of work and sometimes also social relations to employers and companies.

In the tension that spans translocally related sites of recruitment and work lies the complexity in translocal brokerage, and in how the brokers are shifting their identities and loyalties like chameleons shifting their colours. In their role as brokers, they are providing profit-seeking migration industries with their translocal skills, the moral economies based on social networks. Here, they are encountering a moral dilemma, while gambling with their social trust and transferring it into a commodity in a market. Being part of a broader structure of migration industries, translocal brokers are activating a high number of labour migrants, and are here part of sustaining precarious migration flows. However, in the chameleon argument lies that the brokers themselves in many cases are close to the workers, and in the spatial divisions of labour are shifting from actors in the migration industry into becoming workers. In this way, the brokers are, as part of the migration industry, at times even creating precarious arrangements for themselves.

Related to the literature of migration industries, this article aims to contribute to the understanding of intermediaries and brokerage as a translocal process, while paying attention to the multi-sited embeddedness and roles of local brokers. This article adds understanding to translocal embeddedness of brokers as being both a temporally embedded process in the biographies of brokers, and a spatially embedded, multi-sited process based on social networks and a moral economy of kin.

The analysis builds on the empirical example of the Swedish wild berry migration industry, while analysing two sets of brokers: on the one hand, ‘Thai women brokers residing in rural Sweden’, and on the other hand, ‘local brokers residing in rural Thailand’, who are residing next to the recruited staff. Although these actors are starting their recruitment activities from opposite ends of the migration industry, one at the site of work and the other at the site of the workers’ residence, they are both linked to the same type of companies in the migration industry, primarily Thai staffing agencies and Swedish berry companies. Both groups of brokers are anchored in moral economies in Thailand, transferring social relations in multi-sited space, and shifting roles in the spatial divisions of labour like a chameleon shifting their colours to adapt to different environments: from taking part of the arrangement of migration in Thailand, into performers of work, either as hosts or berry pickers, during the berry season in Sweden.

2. Translocality in migration brokerage

2.1 Migration industries, brokerage, and precarity

Processes of international migration are performed within neoliberal migration regimes, related to resource distributions and social inequalities on global and national scales (Delgado Wise 2015; Schierup et al. 2018). Central agents in contemporary migration are the commercial activities of migration industries, emphasising the for-profit motives of those involved in migration-as-business, as opposed to migration processes, which primarily are based on reciprocity and solidarity in social networks (Salt and Stein 1997; Gammeltoft-Hansen and Nyberg Sørensen 2013; Cranston, Schapendonk and Spaan 2018). Hernández-Leon’s (2013) often-quoted definition of a migration industry encompasses a wide range of actors and institutions that enable international mobility, settlement, as well as flows of information, communication, and resources, whose operations are driven by an entrepreneurial outlook.

Although migration brokers are part-and-parcel of the migration industry, a narrower approach via more ethnographic methodologies, details the brokers’ roles in organising transnational mobility and migration (Lindquist 2017; Shrestha and Yeoh 2018; Deshingkar 2019). Lindquist, Xiang and Yeoh (2012: 8) define a broker as ‘a party who mediates between other parties’, arguing that there are great variations in who becomes a broker, how they work, and their degree of professionalisation. Moreover, their observation that ‘a broker is not a fixed identity and must be considered in relation to location, time and power’ points to a need to understand brokers in relation to their particular spatio-temporal contexts. That is, as is quite central to the analysis of this article, the roles and social relations of brokers could be shifting and should be analysed as parts and outcomes of both their historical and their spatial contexts.

Migrant agency and precarity have been pointed out as core themes in labour geography (Buckley, McPhee and Rogaly 2017). Rogaly’s (2009) study of temporary migrant workers urges for attention to individual and collective agency at micro-scales, such as the workplace and incremental changes in working conditions. Migrant agency, according to Deshingkar (2019), should also incorporate multiple and repeated migratory movements, potentially improving precarious and exploitative situations over time.

The focus on agency also brings forward the central discussion on migrant precarity as sanctioned by the activities of the individual migrant. Precarious work is a multidimensional concept that includes uncertainty regarding the continuation of work and income, the degree of control over working conditions, and the extent of regulatory protection, particularly for migrants (Strauss 2018a,b). As pointed out by Tappe and Nguyen (2019), migrant precarity is characterised by ambiguity, and next to being part of exploitative situations, migrant workers also experience improved livelihoods and resilience, why they act to take part in migration themselves. In this article, we highlight the agency of migrant brokers, and their double and complex relation to migrant precarity. On the one hand, as part of migration industries, they are contributing to precarity, while being part of sustaining migration flows. The agency of brokers is contributing to produce migrant precarity by providing migrants’ access to mobility (Deshingkar 2019), inserting migrant workers into precarious positions in ethnically segmented labour markets, including their initial involvement in the selection of workers, based on employers’ preferences. On the other hand, they are also experiencing precarity themselves, when they take part in the work they are arranging.

2.2 Translocal migration brokerage and moral economies

This article adds a translocal perspective to the literature on brokerage in migration industries. We argue here that seeing the brokers as translocal agents adds complexity to the analysis, while focussing on how translocal brokers are giving the migration industry access to multi-sited embeddedness and how they are transferring social and economic relations across space within moral economies and based on their biographical experiences.

The translocal perspective emphasises the local-to-local and multi-sited relations that are constituting enduring migration processes (Carmo and Hedberg 2019). As such, it refers to the spatial embeddedness of migration processes, or ‘transnationalism from below’ (Brickell and Datta 2011), and the translocal structures and institutions that migrant practices are grounded in at multiple places (Peth, Sterly and Sakdapolrak 2018; Schröder and Stephan-Emmrich 2016). In other words, translocal structures constitute a spatial inertia, based on social relations and hierarchies, which nonetheless are in a process of continuous change.

Quite central for translocal moorings to be grounded in space is migrant agency, and the distinct activities that individual migrants perform (Naumann and Greiner 2017; Webster 2017). Translocal practices are incorporated migrant agency, which follows from the individual migrants’ mobilities, tasks, and performances (Daskalaki, Butler and Petrovic 2016). Velayutham and Wise (2005) link translocal villages to the concept of moral economies, and how the transnational relations of communities are based on moral economic obligations in rural translocal villages. According to Scott (1976), a moral agricultural economy is based on social relations among friends and relatives in the sending area. These relations are typically non-market driven, but they might nonetheless contribute to structure labour market relations (Rigg, Bouahom and Douangsavanh 2004). Relating to migration industries, moral economies also form the basis for the trust that exists between brokers and potential migrants (Alpes 2017).

In this article, we view moral economies as central for the activities of translocal brokers, since this concept encapsulates the social embeddedness of brokers in villages. However, we also see that the moral concept brings substantial tensions into the work of translocal brokers, since migration industries are profit-driven organisations. Delgado Wise (2013) highlights how a ‘moral distancing’ is present in transnational labour market relations, where actors in long supply chains, brokers included, are contributing to remove the employers’ responsibilities away from the worker. Our analysis complicates this, while showing that morality is not only part of distantiating employment relations, but is also the social basis that migration industries are leaning upon, through the last link in the chain, the brokers.

We also add a biographical perspective, which highlights the biographical embeddedness of the migrants in previous experiences and future plans (Halfacree and Boyle 1993; Yeoh et al. 2017). The biographical argument shows how temporal aspects are integrated into transnationally linked migration processes (Hedberg 2021). In our case, the adding of migrant biographies shows how the current agency of brokers in the migration industry is embedded in their past experiences as workers.

A challenge following from the brokers’ mutual activities in both Thailand and Sweden is how to conceptualise their roles and the transnational divisions of labour they are part of, since they are both analytically distinct and interrelated. To this end, we employ Massey’s (1995, 2004) concept of spatial divisions of labour, which emphasises a relational view on space where social classes are defined in relation to each other and in relation to their economic functions at work. The stretching out of these social relations across space leads to different spatial structures for sectors and industries. Furthermore, characteristic of the social relations that are constitutive of spatial structures are unequal power relations, resulting in hierarchies with positions of subordination and dominance, creating ‘sub-hierarchies in migrant secondary labour markets’ (Rye and Scott 2018).

The translocal perspective takes this argument one step further, by adding the importance of locally established and spatio-temporally interlinked relations that migrant brokers possess at multiple ends of the migration process. Accordingly, the translocal perspective points specifically at the multi-sited embeddedness and richness of social networks of brokers. We highlight here how translocal networks are based in moral economies in villages of recruitment, but also, how they are stretched to the site of work. The mixing of social relations, built on trust, within the profit-seeking arrangements of migration industries, adds complexity to the analysis.

3. Methodology

This article is based on long-term, multi-sited fieldwork, and information from multiple sources. Periods of fieldwork, enduring for about a week, have been performed in Thailand at repeated times (2012, 2013, 2016, and 2019), in villages in the Kaeng Khro district in north-eastern Thailand, the origin of the recruitment chain of Thai berry pickers to Sweden, and in major cities where Thai staffing agencies are seated (Bangkok and Khon Kaen). Similar periods of fieldwork were also performed at the site of work in the north Swedish inland (2011 and 2012). The multi-sited fieldwork provided an overall understanding of the migration industry, based on interviews and participant observations at multiple sites of work and residence of the migrant workers in rural Sweden and in Thai villages. Interviews were performed with a wide range of actors in the migration industry, including Thai and Swedish companies, Thai berry pickers, and policy makers from both Sweden and Thailand.

Within the framework of this study, this article draws particularly on interviews with four Thai women brokers in Sweden and ten migrant brokers performed in separate villages in rural Thailand. The Thai women brokers were approached 2011–12 in their homes in Sweden, and two of them also in their ‘second home’ in Thailand. The interviews with local brokers in Thailand were performed in 2013 (3), 2016 (2), and 2019 (5). In 2012, the research team also followed a staffing agency when meeting in villages with local brokers and berry pickers. Whereas the first interviews were performed in a study designed to understand the migration process in general, the study in 2019 aimed specifically at deepening the understanding of local brokers. The article also adds research material from interviews performed in Sweden with one camp leader residing in Thailand and four Swedish berry companies.

The main body of interviews with brokers was arranged via snowballing technique, together with some that were arranged via contacts with staffing agencies. The local brokers represented five different staffing agencies, which included all major actors sending workers to Sweden. The interviews with brokers were performed within various constellations of research teams, which included interpreters and research assistants, who were familiar with rural Thailand and spoke the local dialect, and who were recruited based on social networks established with the researchers before fieldwork. Discussions with the research teams afterward, observations during the interviews, and impressions during the fieldwork contributed to contextualise the material. The interviews with Thai women brokers were performed in Swedish, and they were brokering the link between two main staffing agencies in Thailand and three of their Swedish partners. For reasons of confidentiality, the names of the brokers and their employers are omitted.

4. Labour migration between Sweden and Thailand

4.1 A Thai-Swedish migration industry

Turbulent transformations of policy-making and business involvements have resulted in a migration industry surrounding the berry industry (Hedberg and Olofsson 2020). This industry includes a profit-driven recruitment chain based on regulated labour migration, involving actors in Sweden and Thailand. They are acting within a globalised industry, which today is a globe-spanning, large-scale business, embedded in the globalised markets of production and consumption (Hedberg 2013; Krifors 2020). Annually, 25,000 tons of wild berries are exported from Sweden to a global market, mainly China, driven by increased demand and profitability of wild berries (Hedberg 2013; Carmo and Hedberg 2019). This process of globalised production has resulted in intensification and masculinisation of work, and downward pressures on employment conditions (Eriksson and Tollefsen 2013; Eriksson and Tollefsen 2015; Hedberg 2016; Tollefsen et al. 2020). Because of the right of public access (Allemansrätten in Swedish), commercial picking is enabled even in privately owned forests (Sténs and Sandström 2013).

This transformation of the industry into a large-scale business has involved a complete internationalisation of the labour force, based on migrant workers that are provided by a migration industry (Hedberg 2013). Over the last 30 years, the industry has mainly relied on a mix of workers from Thailand and from Eastern and Central Europe, travelling within the European Union framework of free movement. Annually, around 5,000 Thai workers travel to Sweden within an institutionalised Thai–Swedish mobility system (Carmo and Hedberg 2019). The vast majority of the workers are middle-aged men with low educational levels. In their home villages in North-Eastern Thailand they tend to work as self-employed farmers, growing rice for their own consumption and sugar cane as an export-oriented cash crop (Hedberg, Axelsson and Abella 2019).

The main actors in this system are transnational, based in a regulatory context that spans Sweden and Thailand. The most powerful actors in the migration industry we identify to be, on the Swedish side, ‘berry merchants’, and on the Thai side, ‘staffing agencies’ (Fig. 1). The berry merchants buy berries for global export from berry companies, who invite the Thai berry pickers via an offer of employment, and are responsible for dealing with day-to-day issues in Sweden. They are also in direct contact with the Thai staffing agencies, which are the formal employers of the workers (Axelsson and Hedberg 2018). This relationship between berry companies and staffing agencies, illustrated in Fig. 1, provides the main transnational connection in the Thai-Swedish migration industry. Cooperation in business operations has closely aligned these actors to each other (Carmo and Hedberg 2019), and has created a distantiated employment relation between workers and the ‘real’ employer in Sweden (Axelsson and Hedberg 2018). On the Swedish side, three berry companies have dominated the recruitment of workers. In 2017, these companies accounted for two-thirds of total visas for berry pickers (Wingborg 2018), and in 2019, they recruited half of the 6,000 workers that attended the Swedish labour market. In Thailand, the staffing agencies rely on networks of regional and local brokers to recruit workers (Hedberg 2014; van Eerbeek 2019). As this article shows, these brokers then act as chameleons, while transforming from their role as brokers to other positions in the work place in Sweden (camp leader, driver, team leader, and camp staff in Fig. 1).

Figure 1.

Hierarchy of the main actors in the multi-sited spatial divisions of labour for the berry industry.

This migration industry has evolved as a result of mainly two coinciding processes on the Swedish side during the first decade of the 2000s, which formalised the, until then, quite informal import of migrant workers. On the one hand, there was a process of neoliberalisation of Swedish migration policies, which made Sweden into the OECD’s most open labour migration models (OECD 2011), and lifting previous restrictions for employers (Herzfeld Olsson 2018). This implied that Thai migrant workers should only enter Sweden on work permits, rather than on tourist visas, which had previously been practice. On the other hand, the Swedish Tax Agency got its’ eyes on the business, which long had gone under the radar of the Swedish state due to its location in peripheral areas (Eriksson and Tollefsen 2018). In 2005, however, a sentence was passed in the highest court that required the Swedish berry companies to pay payroll taxes. To avoid this, the berry industry in Sweden shifted to a practice of ‘posted work’, stretching the employment relation across Sweden and Thailand to involve Thai staffing agencies as the formal employers (Axelsson and Hedberg 2018).

4.2 The Thai migration context and precarious livelihood strategies

The involvement of Thai staffing agencies, and Thai migrant workers, in this process should be seen in a wider context of transnational Thai labour migration. Overseas migration has been part of governmental development policies since the 1970s, with government promotion of labour migration following economic conjunctures (Chantavanich 1999). Migration and mobility are of paramount importance to Thailand’s social and economic development. Next to being an important host country to 4 million labour migrants, annually 100,000 Thai migrate abroad for work, and estimated remittances in 2018 were equivalent to one and a half percentage of the Gross Domestic Product (Harkins 2019).

Characteristic of this system of labour migration, based on research by mostly Thai scholars (Chantavanich 1999; Jones and Pardthaisong 1999a; Chantavanich et al. 2010; Ayuwat and Chamaratana 2014; Ayuwat and Chamaratana 2015), have been government licensed private recruitment agencies. Furthermore, the bulk of overseas employment takes place via some two hundred agencies, predominantly located in the Bangkok metropolitan area. These agencies rely on the networks of local brokers, who work part-time with migrant recruitment and preparation of documents. The aforementioned research also shows the fees that the migrants are required to pay to agencies and local brokers tend to be high, and despite government involvement, overcharging of fees that far exceed the legal maximum is common practice in the Thai migration industry. The staffing agencies of the berry industry, which denote around five, are smaller and specifically related to the berry industry.

The legal framework set by the Thai state is the Recruitment and Job Seekers Protection Act, specifying conditions and procedures for working abroad and the migration industry (Chantavanich et al. 2010). Official clearance from the Thailand Overseas Employment Administration (TOEA; part of Department of Employment) is required for working abroad. The TOEA provides pre-departure orientations to workers to prepare them for work abroad, and has two to three officials dealing with the Swedish wild berry industry, including monitoring wage slips from staffing agencies and working conditions in Sweden. The state-owned Bank for Agriculture and Agricultural Cooperatives (BAAC) supports farmers’ income diversification through working abroad by providing low-interest loans to agricultural households to pay recruitment fees charged by staffing agencies. Hedberg, Axelsson and Abella (2019) found that 85 per cent of the berry pickers in their survey had loaned from BAAC. Furthermore, BAAC also screens staffing agencies by verifying proper payment of wages in the past, and inquires into working conditions in Sweden.

Thai labour migration is especially associated with the north-eastern Isan region (Mills 2012), which has been explained by its’ ‘persistent condition of regional underdevelopment’ (Rigg and Salamanca 2011). Here, labour migration is seen as part of the income diversification strategies, which arose in response to both a declining capacity of farm land to deliver sustainable livelihoods, and to an export-oriented, agricultural restructuring process, which has marginalised small-scale farmers (Buch-Hansen 2001). A study on 36 Thai villages, however, showed that although migrant work abroad increased the earnings of migrant households, it was seldom able to alter radically their chances for a better life (Jones and Pardthaisong 1999b). Overseas Thai labour migration has primarily taken the form of contract workers for several years to countries in the Middle East and East and Northeast Asia (Chalamvong et al. 2011; Fielding 2016; Harkins 2019). Moreover, a period of enormous economic growth from the 1980s to the 1997 Asian financial crisis, resulted in increased regional inequalities and internal migration flows by farmers from peripheral rural areas in need of additional income to the booming industries in the greater Bangkok area (Jansen 2001; De Jong et al. 2012; Pholphirul 2012). Rambo (2017) argues that this transformed rural economy has altered the traditional sense of community in rural villages. Income diversification notwithstanding, for many households agricultural work remains a base and safety net since much non-farm work is precarious (Rigg et al. 2018; Rigg 2019). According to Rigg, Bouahom and Douangsavanh (2004: 983), there is still a ‘remnant role of the “moral” economy’ in rural Thailand, meaning that social relations remain a corner stone in some rural transactions.

Partaking in labour migration for the Swedish wild berry industry is precarious, while related to uncertainty about earnings, and debt-financed migration compounding the already precarious situation. Berry pickers are paying high fees to Thai staffing agencies, often financed by loans from BAAC, and are in Sweden subject to a daily cost for food, accommodation, and access to a car. Their earnings are determined by a combination of a traditional piece-rate system and a guaranteed wage. The latter was introduced by Swedish trade unions in 2009 to prevent berry pickers from returning indebted. Despite this formal entitlement, Hedberg, Axelsson and Abella (2019) found that one-third of the workers earned less than the guaranteed wage, connected to prevalence of the piece rate system in practice. Moreover, the kilo price of berries is volatile and creates uncertainty among berry pickers about their earnings (Wingborg 2019).

5. Translocal activities of migrant brokers in Thailand and Sweden

5.1 Biographical embeddedness of brokers

Translocal activities of migrant brokers are spanning the multi-sited space where the berry industry is engaged, that is, in a variety of sites in Thailand and Sweden, but with a particular focus on the practices of brokers in rural areas. The biographical approach (Halfacree and Boyle 1993) shows that the current agency in the migration industry is embedded in their past experiences in the migration industry, and how past and contemporary activities blend into each other. The migrant trajectories illustrate how translocality has existed as a long-term process, by repeated seasonal migration across time, and how brokerage tends to be a continuation of their previous work as berry pickers.

The brokerage system builds on complex and historically constructed relations within the berry industry, where we discern two separate but interrelated groups of brokers. The first group of brokers is the ‘Thai women brokers’, who already in the 1980s moved to Sweden to partner Swedish men, and while in Sweden realised the economic potential with berry picking for their relatives and wider communities in rural Thailand (Hedberg 2016). Recruitment started out as a small-scale activity with the closest family, which then escalated into an informal system of migrant brokerage in the Kaeng Khro district in Northeastern Thailand. One woman explains the snow-balling effect:

Since my sister - they are pretty poor in Thailand. I said that I’ll try to bring them here and see if they find their way in the woods and earn some money. And they earned money. And then it grew and grew for everybody else to earn money too. I have been successful with that. (Thai woman in Sweden, 2011)

With time, however, their role as independent brokers in an informal system changed. As showed in Hedberg (2016), the formalisation of the berry industry and deregulation of the Swedish migration policy deliberately outmanoeuvred the Thai women. During this shift, one of the interviewed women stopped recruiting workers, while the other three continued as brokers, mediating workers between Thai staffing agencies and Swedish berry companies. They travel to Thailand, where they recruit workers, and then function as ‘camp leaders’ in Sweden, while hosting the workers and scaling berries at night. At the time of the interview, this process was relatively new, and the women were bringing 50–85 workers, but two of them had the capacity to bring larger groups of 160–190 workers. Interviews with Swedish berry companies also give that there are examples of Thai women brokers, who are partnering with Swedish berry entrepreneurs. One interviewee, the owner of a Swedish berry company with a Thai wife, tells how they have tried to set up their own recruitment agency, based on his wife’s long experience with the recruitment of berry pickers, and his knowledge of the berry industry in Sweden.

The second group of brokers, the ‘local brokers’, consists of farmers residing in rural Thailand. The interviewees have a long-term involvement in the berry industry of 9–23 years. Central to their entry and upward mobility in the industry are their social and cultural capitals. Almost all local brokers started out as pickers themselves, after learning about the work through family members and friends. One local broker narrated how he heard about this opportunity through a family member with relatives in Sweden, while others decided to pursue this work by following friends and kin from villages in Thailand. The interpreter explains the situation of one interviewed broker, whose positive experience with earning money from berry picking made him start acting as a broker:

When he worked with the rice field and sugar fields he lived in a small house, but he still worked hard too. But then he worked with berries and then he built up [his house] new, new, new until now. […] But when he comes back he earned money, then everybody know about this around the village. Then he started to help people from the village to go.

The only female local broker interviewed did not start out by picking, but began her career in the berry industry directly with the recruitment of workers, gaining knowledge about the industry through her boyfriend whom worked as a picker. This gendered division of labour repeats the common pattern that Thai berry pickers tend to be men, along with the idea that berry picking is hard work (Hedberg 2016).

Opportunities for berry pickers to work as a broker appeared during the transformation of the berry industry towards large-scale operations and the formalisation of seasonal labour migration. Often, the interviewees had been recruited as berry pickers by one of the Thai women brokers. When the role of the Thai women brokers changed, an empty space for this activity was left, and some of the berry pickers were offered upward mobility within the recruitment system as local brokers. Although the Thai women brokers continued as brokers, their activity diminished, whereas the industry grew more large-scale. Through their extended contacts with friends and family in the villages they got access to participate in the new organisation of the industry. Their translocal embeddedness in rural Thailand, which also stretched to the Swedish side through their experience as workers, hence enabled them to start as brokers. Illustrative of this is the following broker’s story as told by the interpreter:

At first, the company would send workers to [the Thai woman], and the company saw that he was working well, so [after the shift in system] they asked if he wanted to work with them. The company will come here [to the Thai village] and visit […], we'll discuss which people we are gonna find…I still pick berries, but I also do this recruiting, then I get more money.

However, not only a strong local network, but also being a diligent berry picker was crucial to being offered to work as a broker for the staffing agency, as narrated by the interpreter:

They [the staffing agency] contacted him because they saw that he worked hard, they would look at…their files and information, and how much he had picked last year.

A common thread for both groups or brokers is that their brokerage is embedded in their biography, both as part of their migration histories, as for the Thai women brokers when partnering Swedish men, and also through their previous experiences as berry pickers. In their biographies, the role of their extended social networks in rural Thailand cannot be overestimated, a translocal embeddedness that had developed across time in multi-sited space.

5.2 Moral economies in translocal recruitment processes

Central to the understanding of a translocal perspective is the role of local-to-local relations (Brickell and Datta, 2011) and how they are embedded in local space (Hedberg and Carmo, 2012). In this section, we develop how this embeddedness is primarily anchored in the brokers’ translocal networks, where the recruitment process is based on the remnant role of the moral economy’ in rural Thailand (Rigg, Bouahom and Douangsavanh 2004: 983). Through residence in the same villages, past or present, and shared histories through a friend and kinship, the brokers often had close relationships with the individuals they recruited. As outlined by Scott (1976) a moral economy is affected by conduct and expectations, and not solely by maximising economic gains. This makes recruitment not merely being a market transaction, since the networks that the brokers rely on are also shaped by non-market relations between family and friends.

Our analysis shows that local recruitment in the wild berry industry contains a mix of socially based factors and market-based factors. To some extent, this mix is based on the degree of local embeddedness of the brokers, such as if they recruit only in their home village or if they recruit in larger geographical areas, which even cross-national borders. One of the larger local brokers in this study, recruiting about 300 workers a year, recruits both in his own village, and has sub-brokers in other villages, who are recruiting for him. This shows that local brokerage can consist of many layers of brokers. The same pattern of brokerage is discernible for some of the Thai women brokers. Members of both of these groups mention criteria and skills they want the workers to possess. As an example, one Thai woman mentions that ‘One must have been working hard before. […] If they can pick bamboo in the woods on the mountain [in Thailand] then it’ll be much easier for them, it’s the same way as berries almost’. Larger local brokers express similar skills, also stressing that age, health, and gender matters in recruitment (see also Hedberg 2016).

However, for many of the local brokers in this study, often smaller recruiting 8–50 workers, answering questions about whom they bring and why, the response was almost surprised. For them, it was obvious that they recruit family and friends living in the same village and district. As an example, one local broker emphasises how expectation from family members shaped selection: ‘I did not really choose them, it’s my relatives that go’. Furthermore, their accounts emphasised the importance of continuity. The majority of the workers that the brokers recruit have been going with them repeatedly, and if there is a need for new berry pickers they are recruited via the existing personal ties. In these cases, the ‘moral’ economy might be even stronger than among the larger brokers, and less characterised by market relations.

Underscoring the multi-sitedness in the recruitment process, the locally grounded recruitment process is connected to the spatial divisions of labour during the berry season in Sweden, since the composition of each team of berry pickers is related to the typical seven to eight person’s capacity of the vans used for transport in Sweden. As an example, one of the brokers mentions that he aims to bring eight vans, consisting of friends and neighbours. He prefers to recruit strong men that he thinks will not come back indebted. Another broker mentions that even if people do not pick enough in their first year, this does not constitute a barrier to go next year. The brokers are aware that workers without experience first need to learn the skills. This lenient attitude points at more than a cold economic calculus at work.

Whereas the social embeddedness in rural Thailand is common for all brokers in the study, some of the brokers are also translocally related to Sweden and to Swedish berry companies. This is obvious for the Thai women brokers, but also for some of the local brokers. Having hosted pickers for long, Thai women brokers know, for example, which pickers are familiar with certain areas in Sweden. Interviews with Swedish berry companies also show their relatedness with the Thai women brokers, which they can see both as positive and be ambivalent about. One berry entrepreneur primarily viewed Thai women brokers as competitors, but still used them both as brokers to reach into their social networks and to navigate bribes in Thailand:

You have to understand that you don’t understand, if I say (laughter). You must understand that you don’t understand the Thai. You can never even think how much ruffle and bow there is behind getting these poor workers here. (Swedish berry company, 2011)

One Swedish berry entrepreneur holds forward the particularly skilled berry pickers that one Thai women broker has brought to him:

[She has] the best super-pickers, and always had! […] If a picker in general takes 30 kilos, then they take 80 kilos. They work something extraordinary, and never wastes even 5 cents. (Swedish berry company, 2012)

However, not only the Thai women brokers, but also some of the local brokers have direct contacts with Swedish berry companies. One example is a local broker, who has travelled to the same Swedish berry company during the last 18 years. In the beginning, when he was a berry picker, he was recruited by a Thai woman broker, who was partnering with a Swedish man. Today, however, the couple has separated, and the Swedish man has continued in the berry industry on his own, with the help of the Thai berry picker as a local broker, helping him out with the recruitment process in Thailand. Even though the broker works for the Thai staffing agency, he has a personally established relationship with the Swedish berry buyer, who travels to the Thai village each year.

The analysis shows that Thai women brokers and local brokers are translocally related to rural Thailand, through their widespread social networks, and in some cases also to berry companies in rural Sweden. Another important relationship that the brokers possess is their relatedness to the Thai staffing agencies, who are in direct contact with the brokers regarding the recruitment of pickers, and who also give them tasks related to the collection of documents and fees. Relations to state authorities are scarce, as they are for regular berry pickers. A general exception is some of the brokers, those bringing large groups of workers, directly interacting with BAAC. Complementary to the contact between the BAAC and the staffing agencies, they verify that loan requests are from workers actually going to Sweden. One local broker also arranges meetings with the BAAC for berry pickers that want to take a loan.

Another aspect of the translocally grounded and moral economy is our observation that the competition between local brokers is low. None of the interviewed brokers seemed to be involved in recruitment for multiple staffing agencies at the same time, in marked contrast to Lindquist’s (2012) description of the recruitment practices of local (sub-) brokers in Indonesia, playing off different staffing agencies against each other to maximise their own gain. One explanation for this could be that the local brokers in Thailand depend on the same staffing agency themselves to perform their second, translocal role, as workers or camp leaders in Sweden. Adding to this is the absence of price-based competition between the different staffing agencies. In their relation towards the staffing agencies, the local brokers are closely akin to what Martin (2017) calls recruiters-as-partners, anticipating repeat transactions and following employer preferences, instead of recruiters-as-agents seeking to maximise their own gain from a single transaction. However, we also have examples of brokers in this study who change companies across the seasons, in their search for better conditions. Many of the Thai women brokers have changed company, in their trials to navigate the new, formal system.

Several local brokers portrayed their relations with other brokers as more cooperative rather than competitive, involving discussions about how many people each one can bring. While such cooperation might be logical for those brokers recruiting for the same staffing agency, some local brokers suggested that this also extends to those working for different staffing agencies. For instance, if one broker has too many workers that want to go, while another needs more to make their quota; they will try to sort this out. We also witnessed a similar activity during a meeting between competitive brokers during fieldwork in 2012, when the regional head of one staffing agency visited the local broker of another company. The basis for this collaborative mode is most probably that there is a surplus of workers in the area.

To summarise, our analysis gives that the translocal relatedness of the brokers to Thailand, and often also to Sweden, constitutes a basis for a moral economy of recruitment. Thus, brokerage is not only based on market transactions, but also on social relatedness, something that also might contribute to explain the low competition among brokers.

5.3 Spatial divisions of labour during the berry season

Previous research has addressed the ambiguity in the precarious and subordinated position of Thai berry pickers in the Swedish labour market (Hedberg 2013; Eriksson and Tollefsen 2015; Herzfeld Olsson 2018; Hedberg, Axelsson and Abella 2019; Tollefsen et al. 2020). However, less attention has been paid to the differences in employment and working conditions between categories of workers, where the brokers’ role in both recruiting and then participating at work would take on a special position. Spatial divisions of labour have developed (Massey 1995; Massey 2004), consisting of a complex and messy hierarchy of social relations, which are transferred across nation state borders between the site of recruitment and the site of work. We analyse here how sub-hierarchies have evolved translocally among the brokers, which partly depend on their activities as brokers in Thailand, partly on the idea of berry picking as a way to earn high incomes, together with the wide number of tasks that the brokers perform while working in Sweden.

In order to understand the divisions of labour in Sweden, it is necessary to first clarify details in the organisation of work. Considering the whole berry industry during the season of berry picking in Sweden, usually from mid-July to mid-September, labour is divided into a number of tasks. Whereas the vast majority of the Thai migrants in Sweden work as berry pickers, some also work as mechanics or cooks. There are also combinations of tasks, such as linking berry picking with being the driver of a van. Moreover, a small group of selected workers is exempt from picking and instead have tasks that are related to managing the camp, and the important task of weighing berries, around three to five non-picking staff on one hundred berry pickers. In other cases, however, hosting workers and weighing berries are tasks that are performed by Swedish berry companies.

These divisions of labour also relate to the interviewed brokers. While they all partake in seasonal labour migration, the tasks they perform in Sweden are ranging from being a camp leader to being a berry picker. Two of the Thai women brokers are hosting berry pickers on their private grounds in Sweden, for instance in small houses that are built to accommodate berry pickers or old housing that is re-used to host berry pickers. We have met them during evenings, when the women received the workers and organised the scaling of berries, where they had arranged some facilities for the berry pickers, such as a small house where they could make a fire to keep warm.

Among the local brokers, several of them work as a cook or manage a camp, whereas others partake in berry picking, sometimes combining this with managing a camp. The female broker residing in Thailand works as a cook in the camp for the workers she has recruited herself. These gendered divisions of labour are typical for the berry industry (Hedberg 2016). One male broker has picked berries for several years, but now functions as a camp leader, who is responsible for everything that goes on in a camp of 200 workers, including hygiene, accommodation, quality of food, and looking after the workers in general.

Several of the local brokers who work as berry pickers in Sweden take on a ‘mid-position’ in the spatial divisions of labour, as a sub-broker between the brokers described above and the regular berry pickers. They combine berry picking with being the driver of a van. This normally comes with the advantage of being able to pick a team that will come with them to Sweden, as part of the selection mechanism. In this way, the recruitment process in Thailand is translocally interlinked with the divisions of labour in Sweden. Drivers are also often the team leader, whose decisions influence the earnings.

However, we also find brokers, who hardly associate themselves with brokerage when not directly asked about it. In one case, the interviewee was approached and interviewed in his role as a picker, and it was only some days after the interview when we learned that he also functioned as a local broker. In another interview, with one of the larger brokers, recruiting 300 workers a year, the emphasis of the interview is on his role as a berry picker, rather than as a broker. He talks about his earnings in a similar way that interviewed berry pickers tend to respond:

Di mak. [Very good.] The last year 260 000 bhat [8500 USD; laughing]. He helped to collect people here and then he went to Sweden and picked berries. The same like berry pickers.

Accordingly, there is a discernable hierarchy between brokers acting as camp leaders, brokers picking berries, and recruited workers in the divisions of labour in Sweden, but it does not neatly follow formal job descriptions and tasks. The translocal link between tasks performed as brokers in Thailand, such as the size and location of recruitment, and the position in the hierarchy in Sweden is discernible, but not straightforward.

Additionally, our analysis shows that the spatial divisions of labour, and the social stratification it creates, are related to precarity and risk-taking among migrants (Strauss 2018a; Tappe and Nguyen 2019). We argue here that the brokers’ hierarchical position in the translocal divisions of labour strongly affects their precarity, since only the brokers that pick berries are paying fees to the staffing agency and run the risk of low earnings since they are paid per kilo berries picked. Thus, although there are variations within the research material, the brokers who are higher up in the social hierarchy tend to be less exposed to the risk of indebtedness, one example being the Thai women brokers, who do not work as berry pickers. Although the interviewees would not disclose the exact height of their earnings, one local broker suggested making more in his current role as a camp leader than he would have done as a successful berry picker, also adding that the work is more comfortable compared to the harsh and sometimes dangerous work performed in the forest. Another broker, however, who works as a camp leader in Sweden, emphasises that he earns less than the general berry picker, but is not exposed to the same risks. On the other hand, one of the major recruiters is satisfied with picking berries, since this enables him to take risks and earn more money. This shows that, although there is a hierarchy attached to the spatial divisions of labour it is not clear-cut. Rather, it follows how the individual broker manages to navigate the system, together with individual preferences of risk versus income.

6. Conclusion

This article has analysed the translocal embeddedness of migrant brokers in migration industries through the example of Thai workers in the Swedish wild berry industry. We recognise that translocal mobility systems are key to understand temporary labour migration processes, due to their emphasis on the local context, rather than the national system, and the actual activities that are partaken in space (Carmo and Hedberg 2019). As such, a translocal perspective on migrant brokerage incorporates migrant brokers and their agency within migration industries, as well as the relations that they, through their very performance, are transferring and creating across space. This double-edged role of the brokers, shifting roles and positions between the countries like chameleons adapting to their environment by changing colours, paves the way for a translocal analysis of the brokers.

The main contribution of adding a translocal perspective to that analysis is, according to this article, that it highlights how migration industries through translocal brokers are acquiring access to multi-sited embeddedness in their social networks and locally based moral economies. Our empirical case shows that the brokers can reside both in the villages of the migrant workers and at the site of work they are recruiting for. Although their social relations towards workers and companies vary, they are still multi-sited embedded in both the sites of recruitment and the sites of work. Both groups have established trust with the workers and are partaking in the moral economies of kin in rural Thailand. This means that the translocal grounding of the brokers does not end with their knowledge of recruitment relations in the sending villages, but their translocal skills are stretched to also include the receiving country. However, we notice how conflicts of loyalty can appear in the sometimes conflicting imperatives of moral economies and market logics within migration industries, and how the ‘distantiating moral’ in migration industries is funnelling down sensitive labour market transactions to the social relations in rural areas, where brokers are contributing to perpetuating precarious migration processes.

The translocal perspective also points at the agency of migrant brokers, and how they in their very bodies, are incorporating the migration process through their shifting roles, first, as facilitators in the recruitment process and, second, as performers of seasonal work while hosting or cooking for workers or picking berries. In order to understand the meaning of these shifting roles, and to unpack the meaning of translocal brokerage, we suggest to add the concept of spatial divisions of labour, which is related to the migrants’ biographical embeddedness across time and how the moral economies are not only part of the recruitment process but are also relocated in space to the site of work. Spatial divisions of labour enlighten how social positions of workers are varying and transferred across the borders of nation-states. In the Thai case, the ‘chameleon’ broker first takes part in enabling the spatial divisions of labour through their brokering activities, to thereafter join all sub-hierarchies in the spatial divisions of labour during work in Sweden. Our analysis shows a complex hierarchy of positions, which partly is related to their size in brokerage activities, and also to the location of their brokerage activities as either in their home village or to a wider geographical area. Moreover, the hierarchy is related to their preferences and willingness to take risks in order to reach high earnings. The concept of biographical embeddedness contributes to explain how the previous experiences of work in the sector makes it natural to continue picking berries even for a successful broker, with the built-in promises of a brighter future (see also Hedberg 2021). Additionally, due to the moral economy of recruitment, the brokers are closely related to the workers, which also might help to continue performing work in the lower end of the hierarchy.

Through this analysis, this article suggests that the analysis of migration industries can be complemented conceptually to include a translocal perspective on brokers. We see translocal brokers as key actors in the migration industry, who are providing the migration industry access to multi-sited local embeddedness and trust in social networks. These networks are based on moral economies of kin, and are closely intertwined with the broker’s personal biographies of migration and work histories. Accordingly, we suggest that a translocal take on migration industries and brokerage could be more attuned to the moral economies that brokers are incorporated in, how this intersects with their personal biographies in multi-sited space, and how they are part of a spatial division of labour. This would contribute to a deeper understanding of the importance and complexity of migrant brokerage, who are operating ‘on the ground’ in the increasingly important migration industries, not at least in agri-food industries.

Acknowledgement

We would like to thank Novalee Branting for her support as interpreter and assistant during fieldwork.

Funding

This study was supported by the Swedish Research Council FORMAS [2013–01457]; the Swedish Research Council [2017–01010]; and the Swedish Society for Anthropology and Geography (SSAG).

Conflict of interest statement

None declared.