Introduction
An increasing use of seasonal migrant workers is a typical feature of many high-wage industrialised countries, not at least in the agricultural sector (OECD 2018). A wide amount of research has dealt with the exploitative conditions of migrant workers in this sector (Eriksson & Tollefsen 2018; Preibisch 2010; Rye & Scott 2018; Tollefsen et al. 2020), showing among others the high dependency of migrant workers on their employers, low or absent wages, precarity in relation to the legal status of migrants and generally hard and exploitative labour conditions. Since migrant workers are ‘open to exploitation’, they serve a purpose for firms in neo-liberal economies to maximise their profits: they become ‘the quintessential incarnation of precarity’ (Schierup & Jørgensen 2016: 949). Workplace regimes in agriculture are changing, as employers are engaging seasonal migrant workers under conditions not tolerated by native-born workers (Preibisch 2010). In order to justify the use of migrant workers, they are constructed as hierarchically different from native-born workers, based on ethnic, racial and gendered stereotypes (Acker 2006; Rye & Scott 2018).
In this article, we build on this knowledge to outline how seasonal migrant workers in the Swedish green industries of agriculture, forestry and wild-berry picking are constructed as the most suitable labour. Labour migration to Sweden has long been considerably lower than in other OECD countries, but during the 2000s, the state has actively searched to increase labour migration in all sectors of the economy (Emilsson 2014a). A major part of labour migration is seasonal work within the green industries (Swedish Migration Agency 2021). We outline how discourses on migrant workers in these industries are justifying and constructing migrant labour markets while constructing symbolic boundaries among groups. We do this while combining analyses of both state, employer and union discourses, and extending the focus to include three green sectors; forestry, agriculture and wild berries. As such, our study contributes with deepened knowledge about the ‘employability’ of specific migrant workers in forestry (Wikström & Sténs 2019).
Previous research on Swedish green industries has only sporadically dealt with the symbolic construction of migrant workers, which we argue still is an under-reported issue with a few exceptions. No extensive analysis, scrutinising the discourses on migrant workers from wide range of sources and in all green industries, has to our knowledge yet been done. Our research on Thai wild-berry pickers (Eriksson, Tollefsen & Lundgren 2019; Hedberg 2013; Tollefsen et al. 2020) has indicated how ethnic stereotypes are being used to justify the import of this group of workers. One exceptional statement, from the employer side, even stated that the physical appearance of Thai migrant workers, being ‘short and slender’ and ‘used to walking like a fishhook’, made them particularly suited to employ as workers in the wild-berry industry (Hedberg 2013: 68). Another example is Wikström and Stén’s (2019) work on employer and state policy discourses in Swedish media regarding forestry as a suitable arena for the integration of refugees. This analysis found that employer and policy discourses were favouring settled refugees before seasonal migrant workers as a solution to labour shortages in the green industries.
Our paper, based on discourses by employers, state and unions, points in a different direction. Supported by international research on the dependency of migrant workers to sustain agricultural production (Krifors 2021; Strauss & McGrath 2017), we found in our analysis that employer discourses and legal texts stressed the importance of continued access to temporary and flexible workers, whereas union discourses questioned the seriousness of short-term and under-financed show-case ‘integration projects’ to recruit newly settled refugees to the forestry sector. Accordingly, we emphasise in this paper that the group of migrant workers, which is of interest to the green industries, primarily refers to seasonal migrant workers, that is, those migrants who are imported for the solemn purpose of work, rather than newly arrived migrants, who have been the focus in the Swedish media discourse (Wikström & Sténs 2019). The import of seasonal migrant workers is still the main interest of state and industry while the union side is working for the inclusion of workers already present in Sweden.
Our analysis shows that the state is actively involved in sustaining this discourse, while supporting the picture that the green industries are un-avoidably dependent on specific groups of seasonal migrant workers in order to be competitive. Hence, the green industries in Sweden are no exception to most research on migrant work (Castles 2006; Krifors 2017; Hansen 2021; Munck, Schierup & Wise 2011), where both state and employers, in the current neo-liberal migration regime, are treating migrants as just labour rather than as citizens or workers with citizenship rights. We argue that this relates both to seasonal migrant workers but also to the dis-regard of employing newly arrived migrants or refugees, who are seen as being more ‘rights oriented’ (Friberg & Midtbøen 2019), demanding their rights as residents in Sweden, compared to migrant workers who are imported seasonally to work for low wages. An underlying issue, which largely goes un-spoken, is that employers want continuous access to inexpensive and flexible labour, where migrant workers are part of the ‘logistics’ of making industries profitable (Krifors 2021). This picture remains despite the case that Sweden has a relatively high un-employment rate (EuroStat 2021; Statistics Sweden 2021a), and especially among first generation migrants (Statistics Sweden 2021b), why the Swedish case of importing migrant workers cannot be explained with a lack of domestic workers. Thus, we link to the neo-liberal structure of global economies and its role in producing symbolic boundaries, which is built around an ‘in-situ spatial fix’ (Harvey 2001; Scott 2013), that is, the import of migrant workers to high-wage countries. The Swedish context contrasts with other Nordic welfare states due to its active policies of privatisation in migration policies (Hedberg & Olofsson 2022).
While concentrations of migrant workers to green industries are common in many countries, the processes of how they are forming into ‘immigrant niches’ differ between national and legal contexts (Friberg & Midtbøen 2019; Waldinger 1994). In this paper, we point at how hierarchies are constructed both among seasonal migrant workers, and in relation to native- and foreign-born populations in Sweden, in particular, newly settled refugees. Migrants’ legal status relates to these constructions (Allen & Axelsson 2019) and constructions of hierarchy, that further links to which groups of migrants are likely to be inexpensive and flexible labour.
Against this theoretical framework, analysing the neo-liberal re-structuring of migrant workers in the green industries, and the ‘spatial fix’ of the natural resources it is dealing with, this paper hence advances on the inter-linked role of the state and employers in constructing symbolic boundaries between various groups of workers, divided into segments of hierarchies, between native and migrant workers, and between migrant groups. Drawing on interviews with unions and employer organisations, together with secondary text-sources and legal texts, the aim of this paper is to analyse the processes of how symbolic boundaries are constructed, based on the demand of private actors, and co-produced with trade unions and the state, in order to identify both concurring and conflicting attempts of boundary making. The analysis is structured around discourses on (1) the industries’ ‘dependency’ on seasonal migrant workers, (2) the construction of seasonal migrant workers in relation to Swedish workers, (3) the construction of symbolic hierarchies between groups of migrant workers, based on geographical origin and legal status (particularly membership in the European Union) and (4) the construction of stereotypes of newly arrived migrants/refugees versus seasonal migrant workers.
Constructing and Regulating Migrant Workers
Symbolic Constructions of Migrant Workers
We draw on theories on symbolic constructions of national, ethnic and racial boundaries as a mean of social actors to categorise people (Brubaker 2009; Lamont & Molnár 2002). Symbolic boundaries are inter-linked with the social boundaries that are constructed by this categorisation of people. It manifests and reinforces the social differences among categories and has material consequences, such as un-equal access to opportunities and resources (Lamont & Molnár 2002: 168). In connection to that, we also draw theories on the functions of low-wage labour markets (Harvey 2007) and the active recruitment of migrant workers to specific industries (Friberg & Midtbøen 2019; Waldinger 1994). Scholars have theorised how employers and managers construct migrant workers as hierarchically different from domestic workers (Rye & Scott 2018), especially in agriculture and forestry as low-wage sectors in high-wage industrialised countries (Rye & Scott 2018; Preibisch 2010). Employers re-produce stereotypes of migrant workers based on gender, ethnicity, race and nationality (Acker 2006; Tsing 2009), and categorical characteristics become a proxy for skill and desirability (Friberg & Midtbøen 2018). Workers embody certain assumptions, perform them and, thereby, give them legitimacy (McDowell, Batnitzky & Dyer 2007). Migrant workers are portrayed as ‘better’ by being hard-working (McDowell, Batnitzky & Dyer 2007; Rye & Scott 2018), willing to do tough manual labour (Bonanno & Cavalcanti 2014; Preibisch & Binford 2007) and being particularly skilled for certain tasks (Hedberg 2016). In this way, migrant workers fit ‘perfectly’ as inexpensive and flexible labour in contemporary capitalism (Smith & Winder 2008), whereas native workers are constructed as un-willing to work hard (McDowell, Batnitzky & Dyer 2007; Rye & Scott 2018). Boundaries are drawn within the migrant group as well, where certain ethnicities and nationalities signify suitability for certain jobs (Friberg & Midtbøen 2018).
Our previous research, focussing on Thai wild-berry pickers in Sweden as an example of low-wage migrant workers shows how symbolic constructions of boundaries also justifies temporary migration. Within the Swedish wild-berry industry, there is a hierarchal order of preferred workers based on nationality (Hedberg 2013). These workers are not only seen as having the right characteristics, but shows compatibility between seasons in sending and receiving countries (Krifors 2017; Hedberg 2021), leaving wages, working and living conditions un-contested (Krifors 2021). Thai berry pickers’ exploitative employment relations appear less problematic when discursively placed outside regular Swedish labour market rights (Krifors 2017), creating precarious work conditions that are based on the workers transnational points of reference to their home villages in Thailand (Hedberg 2021).
The Intersection of Labour Market Dynamics and State Policy in Creating Precarity
Harvey (1981, 2001) discusses how firms use spatial strategies to mediate the effects of periodic crises. Green industries in high-wage industrialised countries cannot re-locate spatially, but capital may resolve with an ‘in-situ spatial fix’ (Harvey 1981, 2001) by recruiting low-wage migrant workers. Labour migration from the periphery to the capitalist core works as a temporary mediator of crises for capital as it lowers the costs for labour (Kasimis, Papadopoulos & Zografakis 2015; Scott 2013), and shifts power from labour to capital through the regulation of labour (Scott 2013). Having a migrant labour force can facilitate circumvention of national legislation, that is, payroll taxes (Axelsson & Hedberg 2018; Lillie & Wagner 2015). Seasonal migration of workers externalises the re-production of the workforce to a place separate from the place of production (Buraway 1976), permitting employers to avoid responsibility (Smith & Winder 2008).
The state plays a significant role in the regulation of labour, and especially of labour migration (Allen & Axelsson 2019; Scott 2013). The neo-liberal state increasingly assumes the role of providing beneficial conditions for capital and a territory for investments (Krifors 2017), with labour migration policies favouring receiving country companies and interacting with migration industries (Castles 2006; Hedberg & Olofsson 2022; Rye & Scott 2018). Migrant workers find themselves being subject to a nexus of regulations that is connected to structures in labour markets, migration policy and control (Allen & Axelsson 2019; Lewis et al. 2015). The nation state has a role in constructing difference along national, ethnic and racial lines (Lamont & Molnar 2002). Swedish labour migration policy is constructed as a neo-liberal migration regime in negotiation with the requirements of Swedish industries and firms (Hedberg & Olofsson 2022), which has also affected the migration regime in the sending country (Axelsson & Hedberg 2018). Migrant workers tend to face processes of subordinated inclusion, where certain groups de facto are not included on equal terms, despite their formal legislative rights (Tollefsen et al. 2020; for Norway, see Rye & Scott 2018; for Canada, see Preibisch 2010). This further contributes to the making of boundaries between migrant workers and native workers, with their differences in bargaining power depending on how ‘exploitable’ they are (Friberg & Midtbøen 2018). Following Friberg and Midtbøen (2019), we see the demand-side as crucial for the formation of ‘immigrant niches’ to specific sectors in Sweden, together with the developments in the broader structural and legal national context facilitating an ‘in situ spatial fix’ (Harvey 1981, 2001) of the green industries.
Migrant Workers in Swedish Green Industries
Swedish green industries recruit seasonal labour migrants both from non-EU countries and from the EU. The legal framework for labour migration from non-EU countries to Sweden changed in 2008 to become one of the most liberal in the OECD (OECD 2011). Previously, Sweden had a quota-based system based on labour shortages controlled by a state agency, the trade unions and employer federations. After 2008, employers define their own needs for foreign labour, skilled or non-skilled, while the state and the trade unions have limited influence (Emilsson 2014a). Since then, non-EU labour migration has steadily increased (Emilsson 2014b), with Thai berry pickers being the largest group, around 3,000–6,000 workers per year (Hedberg, Axelsson & Abella 2019). Recruitment has also taken place to other industries with low wages and no shortages (Swedish Migration Agency 2021; The Swedish Trade Union Confederation/LO 2013) and Calleman and Herzfeld Olsson (2015) identify a division of labour between sectors and nationalities. After 2008, some amendments were made to the law to prevent excess exploitation of seasonal workers (Axelsson & Hedberg 2018; Hedberg & Olofsson 2022). In 2018, the law was adapted to the new EU-Directive 2014/36/EU on seasonal workers from non-EU countries.
Migrant workers from the EU, on the other hand, move freely and have the same rights as Swedish citizens to reside and work in Sweden. They need not to register with the authorities for visits of up to three months, which explains the lack of official statistics on EU seasonal migrants to the green industries (Swedish Migration Agency 2020a). Nevertheless, wild-berry companies estimate that migrants from the EU comprise about one-third of the total work force (Hedberg & Olofsson 2022). Within agriculture, a government report estimates that around 2,000 migrants work in Sweden annually (Svensson et al. 2012). Eight years later, the Federation of Swedish Farmers claimed that agriculture and forestry would be missing out of 8,000 migrant workers due to closed borders of the COVID-19 pandemic (LRF 2020a; LRF 2020b). An interview study with the 10 largest forestry companies in Sweden showed that a vast majority of recruited workers were migrants, mainly from Poland, Lithuania and Ukraine (Forsmark & Johannesson 2020).
A work permit to Sweden requires an offer of employment and generally the working conditions and wages should be in pair with collective agreements (Swedish Migration Agency 2020b). In sum, seasonal migrant workers compose a major and increasing share of the labour force in Swedish agriculture, forestry and wild-berry picking, and the 2008 legal framework has facilitated the process of formation of ‘immigrant niches’ in particularly low-wage sectors, such as the green industries.
Data and Methods
In this paper, we use discourses analysis to get an understanding on how actors within the green industries of agriculture, forestry and wild-berry picking in Sweden construct different groups of workers as well as the roles that seasonal labour migration have within the industries. We reject that language is neutral but rather a key feature to understand society. In our approach, we utilise an understanding of discourse as enmeshed with the material reality and geography, which produces it, and those discourses are part of the re-production of society (Gill 2000; Mitchell 2002). This entails the importance to regard language and discourse as constructive and as a social practise, where discourse is used to instil action (Gill 2000).
The materials analysed are interview data from six recorded and transcribed interviews with national representatives from trade unions and employer federations, as well as written materials collected from employer and industry federations and the legislative body. The first author has been responsible for the data collection while the analysis has been a joint effort. All interveiwees were contacted via phone or e-mail and then followed up with information about the research via e-mail, the same written information was given to them at the time of the interview. The choice of interviewing the trade unions and the employer federations is related to the regulative power they have in the Swedish labour market. The people interviewed for this article are seen as representatives for the organisation rather than speaking for themselves as individuals. We have chosen to exclude their names, but the organisation they represent are named. This was also disclosed to them during the time of the interview. All interviews were made during 2019 and early 2020, before the start of the COVID-19 pandemic in Sweden. The written material from the industry consists of two different booklets: This is Our Land (Det här är vårt land) from 2014 and Berry Power with Social Responsibility (Bärkraft med socialt ansvar) from 2019, where the first is produced by the employers’ organisation and the second is commissioned by the temporarily organised association of berry companies. Both of these materials share that they are written to inform policy-makers, and to some degree also the public, about how these industries function from their own perspective, and that migration to Sweden in different forms is an important part of that. From the state, we draw on legislative documents from the Ministry Publications Series, covering the changes to the labour migration policy in 2008 (Ds 2007: 27) and the Government Bill, on the Swedish implementation of the EU-directive on seasonal labour migration (Prop 2017/18: 108). While being important due to their legislative power, the documents also contain the motivation from the government to the proposed and implemented changes. All material is orginally in Swedish and all of the authors are native Swedish speakers.
A cyclical approach was used in the analysis of the material. The starting point for the analysis was how migrant workers and their role in the industries were described in the different materials, but descriptions of Swedish workers were also early added to the first round of analysis. Through the analysis process, we also added a focus on how different nationalities of migrant workers were described along with differences in how workers in Sweden with migrant background were described. While all actors carry a societal influential they are not so equally and they all represent different interests on the Swedish labour market, this was an important notion during the analysis to distungish inbetween different discourses. This was also used as the main differentiation when analysing the material rather than written or interview. As we cycled through the material, we also found that some discourse overlapped with each other from the different actors.
Conflicting Discourses on Divergent Groups of Migrant Workers
Industry ‘Dependency’ on Seasonal Migrant Workers
In order to motivate the interest for discourses on symbolic boundaries between different groups of workers in the green industries, we first turn to the discourse on the industries’ acclaimed ‘dependency’ on migrant workers. This is central because it explains why the differences are made in the first place. This discourse, primarily brought forward by the employers, hence claims that the green industries are fundamentally and crucially dependent on seasonal migrant workers.
Both the employer federation and the trade unions are expressing difficultly to acquire enough national labour power, but given the high un-employment rate in Sweden, they do not point to an actual lack of workers. Instead, the employers’ side mentions the un-willingness of Swedes to perform this kind of work, especially since it is seasonal. ‘It certainly won’t become any easier to find the labour power in Sweden. The jobs are more demanding and pay less than in construction’ (This is Our Land 2014: 50). In texts and interviews, the employers’ side in all three industries are explicit in their conviction that production would seize without migrant workers.
The trade unions acknowledge the same dependency, but in contrast to the employers, they describe it as ‘an artificial situation’, created by the employers’ un-willingness to pay fair wages and provide fair working conditions.
The state side rather supports the arguments of the industry. The government proposition highlights the need for a system that is efficient and flexible. The state’s requirement of a guarantee for fair wages and working conditions, however, is only done in relation to Swedish workers:
The regulations regarding labour migration should be flexible, facilitate quick appointments for labour needs and also ensure that wages, insurances and other employment/working conditions that are offered to people that migrate to work are matching the ones already being offered to employees in the country (Ds 2007: 27).
The employers’ discourse of migrant dependency integrates an idea that the current situation, based on the availability of migrant workers, should be preserved. This includes both the description by employers of the wage levels in the industries, and also their aversion for seemingly all types of state regulation.
The employers in green industries see the recruitment of temporary migrant workers as a win–win situation for them and the migrants, because of the wide gap in living costs between the migrants’ home countries and Sweden. This argument could, according to both employers’ organisations and unions, explain why Polish workers are no longer common to work in Sweden. As the wages in the home country has increased, the wages in Sweden are no longer attractive enough; along with competition from other countries, such as Germany. At the same time, relating to wages in agriculture, this is not seen as a reason for increased wages: ‘No […] it is the same wage development as for everyone in Sweden’. Similarly, in This is Our Land, wage levels are described as being beyond the control of employers, and it is taken for granted that they are, and will remain, low. Instead, employers shift to other groups of workers, who are willing to work for low wages.
In Berry Power with Social Responsibility, the berry companies point to differences in wage levels as a motive for migrant workers, which would justify the low wage in Sweden. They argue that Thai berry pickers are earning five to six times more than what they would earn during the same time at home.1 They further argue that if the cost for workers increases, it will be carried by the workers and ‘make it un-profitable for the pickers to come to Sweden’ (2019: 5). This, in turn, would resolve in a major cut in production which would have ‘dire consequences’ to the companies. The argument about the necessity of low wages is shared among employers in the three industries; there is no capacity for increased labour costs. The employers almost consistently describe all state regulations and their implementations as too though and a hindrance for perform their business properly: ‘You could say that all of it has become more difficult and more obscure, for example with the implementation of the Seasonal Workers Directive […] it’s incomprehensible how Sweden decided to do that’ (Employer federation, forestry).
Taken together, the green industries portray themselves as un-avoidably dependent on low-wage migrant workers in order to be competitive. However, employers and unions produce opposite discourses on the process of its formation in this particular national and legal context (Friberg & Midtbøen 2019).
Oppositional Relationships between Migrant Workers and Swedish Workers
In line with many other studies (Friberg & Midtbøen 2018; McDowell, Batnitzky & Dyer 2007; Rye & Scott 2018), hegemonic discourses that contrast domestic workers with migrant workers were salient in our material. Migrant workers are described as everything that the domestic workers are not, and vice-versa. This appears most strikingly in the accounts from the employer, where the migrant workers foremost are seen as working bodies (McDowell, Batnitzky & Dyer 2007; Smith & Winders 2008), whereas in legislative texts, the migrant workers foremost represent just labour without being people (Castles 2006).
In our analysis, a dividing line goes between employers and state, who are constructing seasonal migrant workers as favourable for employment, and the unions, who are supporting the pools of workers who are available in Sweden and their paying members. As illustrated previously, the employer side in agriculture and forestry expresses a strong dependency on seasonal migrant workers, in order to stay competitive in their respective markets. This, they justify with constructions of migrant workers versus native workers, which can be seen both in the interviews and text material. In order to justify the use of migrant workers, employers are constructing symbolic boundaries between migrant workers and native workers, portraying migrant workers with qualities that make them particularly skilled to perform manual and hard work (McDowell, Batnitzky & Dyer 2007; Preibisch & Binford 2007; Smith & Winders 2008). In line with McDowell, Batnitzky and Dyer (2007), the discourse on migrant workers focuses on them as dedicated to hard work and as being content with the job: ‘Most people I’ve met have been very happy to come here to work’ (Employer federation, forestry). The focus is on the migrants’ bodily performance of work and on how strong and resilient they are. This also comes through in relation to Swedish labour regulation of working hours:
Our problem is that they can’t work as much as they would like to. They want to work, get their salary and then go home, and it’s like stop, there are regulations about rest and there are work environment aspects (Employer federation, forestry)
Employers highlight that many workers are returning annually, which is taken as a sign that they are treated fairly and come to almost familial contexts. In consequence, foreign workers are constructed not only as hard-working and content, but also as loyal and eager to return on a seasonal basis.
Sure they make more money here than at home. The pay is maybe 20–30 percent higher but that’s not the main reason for people from Lithuania and Rumania to come here to work. It is the working conditions. They have better accommodations and don’t have to work as hard as they would have to in the forest at home (This is Our Land 2014: 191).
In contrast, Swedish workers are described as being picky or lazy about doing this kind of manual, gruelling work. This type of work, then, is not desired by Swedish workers. The employer organisation’s text states that: ‘To recruit Swedish labour is not a viable option’. Swedes, in a historical context, ‘used to be’ both physically and mentally equipped for this kind of work. Thus, employers do not point to an actual lack of Swedish workers, similarly as described by Friberg and Midtbøen (2018, 2019) for the Norwegian context. Instead, Swedish workers are constructed as both un-willing and also un-fit to do the work and therefore un-wanted by the employers.
When he has advertised for work in Sweden he gets heaps of applications from people who are in long-term unemployment and who lack both the willingness to work in the forest and also all the skills that are required […]. To meet the contract you need a strong mind and a solid physique (This is Our Land 2014: 43).
According to the employer side, the right characteristics instead fit with migrant workers. This view of migrant and Swedish workers becomes incorporated in the explanation of why the individual employer hires migrant workers. The migrant worker becomes foremost a hard-working labouring body (McDowell, Batnitzky & Dyer 2007; Smith & Winders 2008).
Constructing Borders and Hierarchies among Migrant Workers
In the material, we also found symbolic boundary construction between various groups of migrant workers. Nationality plays a central role in creating symbolic boundaries, matching migrant origins to certain tasks. These symbolic boundaries connect and interact with the differentiation created by the state (Lamont & Molnar 2002), with different legislative conditions directed towards different nationalities, creating major difference between migrant workers from EU countries and from non-EU countries. A lack of regulations on EU berry pickers, for instance, makes them particularly vulnerable in the labour market (Vogiazides & Hedberg 2013).
From the agricultural employer’s side, EU and non-EU workers are viewed as different in terms of their deportability, in particular if you may ‘send them home’ or fire them (Green Employer, agriculture). When discussing wages for strawberry picking, the employer representative described how workers need to fulfil 80% of the ‘guaranteed’ wage through the piece-rate, otherwise they are fired. ‘This is no problem if the pickers are from the EU, but non-EU pickers cannot be fired so easily […] you need to know beforehand that the person is made for picking […]. You cannot send home a person from third countries […]. It is a huge dilemma’ (Green Employers, agriculture). This exemplifies both a hierarchy amongst which type of workers are preferred related to their bargaining power (Friberg & Midtbøen 2019), but also how migrant workers gets caught between employment conditions and migration legislation (Lewis et al. 2015).
The union side, too, recognises a division of labour between EU and non-EU migrant workers. In forestry, the GS representative draws a line between planters, who are migrant workers from non-EU countries (propelled by the 2008 legislation), and thinners/clearers who continue to come from EU countries. In the interview with The Swedish Municipal Workers’ Union’s representative for agriculture attached the differentiation based on nationality to the geographical distance of the migrants’ countries of origin:
There is only one aspect in this, and that is from how far away you come. The further away, the lower you are on the ladder and the lower wages you are paid. So it is always about the wages.
In both forestry and agriculture, differences are made between Polish workers and ‘the others’. The union representative for agriculture portrays the Poles as ‘more organised’. In Sölvesborg, during 2006–2007, ‘they have even gone on a strike’ at a time when ‘the others were not here yet’.
Similarly, as to what Scott (2013) relates to the ‘in-situ’ spatial fix, as the migrant workers have been returning to Sweden over several years they start to demand and want similar things as the Swedish workers would. As workers become more organised, employers are prompted to go further away to recruit a less demanding workforce (Harvey 1981, 2001) that will accept the current wages and working conditions.
In the wild-berry industry, on the other hand, Thai workers are portrayed as particularly skilled, compared to European workers. The industry has created their own model to enable Swedish companies to import Thai workers with favourable taxation rules (Axelsson & Hedberg 2018). In Berry Power with Social Responsibility, the berry industry describes how the industry is threatened because the model built around Thai workers could be replaced by the non-preferred East European group: ‘The empty space runs the risk of being filled with free pickers from Eastern Europe without collective agreement, education or arranged accommodation’. While Thai workers are ‘experienced and effective’, a complete shift to East European workers would consist of un-organised ‘tent-camps’. The discourse on nationality in the wild-berry industry thus comes from pictures of variegated skills that are attached to nationality. The Swedish Municipal Workers’ Union describes how these ideas originate from an established notion in the industry, which views Thai as ‘defenceless’, ‘unprotected’ and ‘friendly’, based on the tradition that Swedes travel to Thailand for vacation:
It is cynical when one talks to berry buyers about which countries they bring pickers from. It feels like 1910 when one talks to them, which nationalities that are good, who is industrious, who is honest, who is dishonest. We try to mark that we don’t think like that. It is not nationality that decides if one is honest. (Trade Union, Berry Industry)
Symbolic Boundaries of Migrants Living in Sweden
Next to the hierarchies between groups of seasonal migrant workers, there are also processes of boundary-making between migrants living in Sweden, such as newly arrived refugees and asylum seekers, in relation to ‘Swedish’ workers.
Newly arrived refugees are on several occasions mentioned as an alternative to seasonal migrant workers in the green industries (Wikström & Sténs 2019). In our material, however, both employers and state are dismissing them as employable.
In the interviews with the employer and union representatives, the recruitment of the ‘newly arrived’ refugees living in Sweden became part of the discourse on the difficulty to recruit workers to the green industries. The employer side constructed these groups as problematic on similar ground as ‘Swedish’ workers, who were un-willing to perform this kind of work. The union side, instead, expressed frustration over the employers’ ‘lack of interest’ in promoting the work both to the ‘newly arrived’ and to ‘the Swedish youth’.
Shifting delineations appear in the ways the labour market parties respectively construct newly arrived migrants as potential workers in the green industries. The Green Employer representative presents a sceptical picture of ‘newly arrived’, first by describing the variety of crops within agriculture and how it is not an easy job: ‘About the newly arrived, things have to work out in terms of them being able to travel to the sites, because that is the problem out there […] then it totally depends on the contacts you have with the Employment Office’. The construction of the newly arrived then shifts, and they are mentioned as similar to ‘Swedish youth’, un-willing to take a job in the green industries: ‘You cannot find a Swede who wants to do this’. The union representative (GS) constructs a different picture of the ‘newly arrived’, who also by them are described in similar terms as the ‘Swedish youth’, but also as potential workers in the green industries, were they should be given fair work conditions and wages; with the added benefit of ‘integration’. This was expressed in frustration about the lack of interest from politicians to work constructively with recruitment among un-employed youth in Sweden, including the ‘newly arrived’.
In other words, persons with migrant or refugee backgrounds, who are already living in Sweden, can to some degree similarly to ‘Swedes’ demand higher wages and fair work conditions compared to seasonal migrant workers. Nevertheless, some employers highlight that there are some ‘so-called Swedes’ who accept to work in the green sector. Similarly, the employer representative for agriculture describes that ‘Swedes who come from Asia’ does a good job while working in the fields, ‘On tomatoes and vegetables we have a couple of Swedes who come from Asia and who are very good’ (employer agriculture). Greenhouse workers are exemplified in the employer organisation’s text This is Our Land as being recruited among persons with migrant or refugee backgrounds:
We have tried with Swedes, but they don’t want the job. The work tasks are not seen as ‘modern’. This work has low status in the eyes of the Swedes (2014: 158).
While being Swedish citizens, as Bosnian refugees coming to Sweden in the early 1990s or born in Sweden to migrant parents, these workers are nevertheless here constructed as Other to ‘the Swedes’. The greenhouse industry has mostly full-year employees, all with migrant backgrounds. ‘In fact, we are now starting to employ the children of those who came as refugees during the 1990s. There are many sunshine stories among them’ (2014: 150). Thus, there are shifting constructions of what a ‘foreign worker’ is, and who is apt for manual work in the green industries. Sometimes, being ‘migrants’ include people, who are living in Sweden but with a different ethnic background while at other times, ‘migrants’ are non-citizens/people who are not living permanently in Sweden.
Conclusion
This paper has examined the process of how the Swedish employer side is constructing symbolic boundaries between various groups of workers in the Swedish labour market, and how this has been supported by state but partly contested by unions. In a time when green industries are facing intensified global competition, the industries are experiencing increased pressure on wages and deteriorating working conditions, which increases migrant precarity.
As in other OECD countries, the most common trend for labour migration to Sweden is seasonal migrant work to agricultural and rural industries, facilitated by a liberal and de-regulated policy of labour migration. Although many industries are using a ‘spatial fix’ (Harvey 2001), re-locating production to low-wage countries to mediate crisis and increased competition, the green industries have instead developed a strategy of reliance on inexpensive labour imported from other countries, an ‘in-situ fix’ (ibid.). Located in a high-wage industrialised country, our analysis shows how this creates dilemmas for employers in terms of how to justify the lower wages and sub-standard working conditions for certain employees. We have analysed how employer representatives discursively construct a dependency on seasonal migrant workers, supported by state but partly contested by unions, which places different groups of workers in hierarchical order.
In this paper, we agree with Friberg and Midtbøen (2019) on their emphasis on the demand-side in the construction of symbolic boundaries on migrant workers. In the Swedish case, our analysis shows that groups of migrant workers in dominant employer discourses are hierarchically represented as essentially hard-working and grateful, in contrast to the un-willing and low-performing Swedish workers. The industry portrays temporary migrant workers with qualities that make them particularly suited to perform manual and hard work, with ‘a strong mind and a solid physique’, and migrant workers are also constructed as essential to the flexibility and competitiveness of green industries. We also identify hierarchies within the group of migrant workers, which are based on nationalities. A special group concerns the migrants already living in Sweden, such as newly arrived refugees, who are seen as more similar to native-born workers because they possess the same rights in the Swedish welfare system. They are not dis-regarded as workers in the same manner as native-born workers, but still not described as the same perfect fit as seasonal migrant workers. Accordingly, whereas Friberg and Midtbøen (2019) see the workers as ‘negatively selected’, we rather argue that domestic workers, including newly settled refugees, are un-wanted because they are not ‘exploitable’ to the same extent as temporary migrant workers. We speculate that newly arrived migrants might be ‘rights orientated’ (Friberg & Midtbøen 2019), and as such less preferred by the employer side, who constructed both ‘Swedish youth’ and ‘newly arrived refugees’ as having similar expectations. This points to the discourse primarily being based on the workers’ institutionalised rights and legal status, before ethnic premises.
Also, in relation to Friberg and Midtbøen (2019), who discuss the role of national contexts but see similarities within Nordic welfare states, we here highlight important differences between the Nordic countries. One important distinction regards the role of the state, and its role in symbolic boundary-making, and in the construction of migration regimes. Although the Norwegian state in general is restrictive and almost exclusively source labour migrants from the European Economic Area (EEA), the Swedish state is explicit to support the import of also non-EU migrant workers to strengthen the conditions for the Swedish companies. This is explained by differing national contexts between Sweden and Norway, in un-employment rates, economic growth trends and effects of the 2008 financial crisis and not least liberalised laws and regulations on labour migration since 2008. All of these factors have contributed to the distinct import of migrant workers in Sweden.
Accordingly, we show in this paper how symbolic boundaries of migrant workers are not only performed by the employers’ side, but they are also co-constructed together with and sanctioned by the state; in some regards, the symbolic boundaries are also co-constructed with the unions. The system of labour migration in Sweden has been de-regulated in accordance with demands from the employers’ side, which in turn has placed increased risks upon migrant workers, which brings Sweden even closer to the same neo-liberal migration regime prevailing in Europe and the EU. Swedish state discourses, in recent national law and in the implementation of EU-directives, construct migrant low-wage workers as legitimate, as long as they are seasonal, not affecting domestic labour markets and non-users of welfare services. The unions, in turn, see the dependency on migrant workers as an artificial state, constructed by the employers’ un-willingness to increase wages.
This paper, thus, shows how symbolic boundaries are constructed between different groups of migrant workers, primarily by the employers’ side, but also by the state and by the unions. The discourse appears against the background of tightened structural conditions among green industries, who use an ‘in situ fix’ of importing migrant workers.
Notes
[1] Survey data in Hedberg et al. (2019) reports that earnings in Sweden are only three times those in Thailand.
Funding Information
2013-01457 Grapes of Wrath: Global labour mobility in the wild berry industry affecting rural development in Sweden and Thailand
2021-02226 Wild berry shifting: Sustainability transitions and diverse economies in a rural ‘sparseland’
Competing Interests
The authors have no competing interests to declare.