Locked in inferiority? The positions of Estonian construction workers in the Finnish migrant labour regime

The aim of this article is to analyse how different policies and actors have structured the current migrant labour regime in the Finnish construction sector and to discuss the consequences for migrants. Our study shows that a strong industrial relations system such as in Finland is able to curb the posting of workers regime (the most disadvantageous for migrant workers). The position of labour migrants has become more diverse in the segmented labour market, although it remains inferior compared to that of the natives. Consideration of the policy development revolving around the changing migrant labour regimes constitutes the first part of the analysis and is based on government and trade union officials’ accounts. The more substantial part of the study draws upon biographical interviews with Estonian construction workers and analyses the division of migrant labour according to their employment in four ‘patterns of firm ownership’ that range from the most unfavourable to most favourable position: workers posted by Estonian firms; workers employed by firms registered in Finland but operated by Estonians; self-employed/small business owners; and workers employed by Finnish firms. The structuring of the regime according to the pattern of firm ownership can be interpreted as a manifestation of employers’ intentional strategies to adapt to or avoid national regulations and to some extent as also reflecting workers’ individual and collective agency.


Introduction
This study, which draws upon biographical interviews with Estonian construction workers working in Finland and expert interviews with union and government officials, analyses the emerging migrant labour regime in the context of the large inflow of Estonian construction workers to Finland. Firstly, it elaborates on the structuring of the regime and shows that labour market regulations and industrial relations are crucial in shaping it. Secondly, it aims to show the differing position of migrant workers within the regime, which is contingent upon their employment in different patterns of firm ownership. Although some studies have shed light on the internal division of segmented migrant labour regimes (e.g., Fan 2002;Colic-Peisker & Tilbury 2006;Friberg 2012;Wills et al. 2009), to our knowledge there is no previous scholarly attempt to grasp the internal variation of the migrant labour market regime on the basis of workers' employment in firms with different ownership types within one nation-state and taking a particular sector as an example. This study seeks to fill this void.
Based on biographical interviews with workers, we consider firm ownership as an important factor contributing to the segmentation of employment conditions.
Prior to the accession of 10 Central and Eastern European (CEE) countries to the European Union (EU) in 2004, the Central Organisation of Finnish Trade Unions (SAK) warned that there would be 400,000 Estonian workers coming to the Finnish labour market if no restrictions on free mobility were imposed. In Finnish 'corporatist' tradition trade unions, alongside the government and employer associations, play a major role in shaping national industrial relations (Sippola 2015). Significantly influenced by the opinion of the SAK, the Worker posting has proven to undermine strong industrial relations systems in Europe (Wagner & Lillie 2014). This is a European-wide problem, where the nexus of subcontracting, transnational mobility, legal insularity and employer anti-unionism complicate union and legal enforcement of labour market standards (Berntsen & Lillie 2014;Eldring, Fitzgerald & Arnholtz 2012). The 'posting of workers' regime remained widely used in the Finnish construction sector for many years after the lifting of the restrictions on free movement of labour in 2006. Posted work gave employers the means to 'opt out' of national regulation (Lillie & Sippola 2011). However, a change towards a greater variety of employment relationships has occurred in the 2010s. The use of posted workers from Estonia has lost its attractiveness for companies, as the 'beating of the system' has become more difficult. This has occurred mainly due to deliberate politics by the Finnish government, which have been influenced by social partners. Today, there are signs that the policies of the government aimed at regulating the construction sector have had a positive overall effect on the construction labour market. This suggests that even within the single EU labour market, certain (strong) national industrial relations systems are more resistant to the corroding effects of post-accession labour migration than others.
Based on our data we elaborate on the internal division of the current migrant labour regime and introduce the notion of 'pattern of firm ownership'. It refers both to the distinct division of labour between main contractors with Finnish ownership and subcontractors with Estonian ownership and to the migrant workers' relationship with and distance to the main contractor.
The pattern of 'firm ownership' is thus a conceptual short form that points to the fact that the Estonian migrants are incorporated into a system of production where staffing strategies and working conditions differ systematically in accordance with their employers' nationality, the relationship to the main contractor (and function in the production chain) and workers' contractual status.
Accordingly, a major part of the analysis concerns the division of migrant labour according to four main patterns of firm ownership, ranging from the most unfavourable to the most favourable position for the Estonian workers: posting of workers; employed by a firm registered in Finland but operated by Estonians; self-employed/small business owners; and employed by Finnish firms. The sequence of the presentation of these different positions is based on the observed ranking of precariousness (in terms of uncertainty) and the level of wages of the various positions. We also argue that even though it is possible for Estonian construction workers to end up in a rather secure employment situation, they mostly remain in positions materially inferior to and culturally detached from those of natives.
The article is structured as follows. First, in the theory section foundations for the different regimes for natives and migrants are presented, including the factors structuring the evolving migrant labour regime in Finland. The second section familiarises the reader with the Finnish construction sector. Then follows a methodology section, after which the making of the 'migrant labour regime' -a heuristic concept emphasising political and relational grounds for labour migration -is elaborated on; this elaboration is mainly based on government and union officials' accounts. Section five describes the structure of the current migrant labour regime prevailing between Estonia and Finland and is based largely on worker biographies.
The discussion and conclusion section sums up the findings and suggests issues for future research.

Different labour regimes for natives and migrants
In his seminal work Piore (2008Piore ( [1979) developed the idea that labour markets in the advanced industrial economies are divided into two tiers. The migrant workforce occupies positions mainly in the lower one which is characterised by lower wages, inferior working conditions, less stability, lack of advancement opportunities and less autonomy than those in the first tier that are mainly occupied by natives. The uncertain and temporary nature of work in the lower tier is best captured by the concept of precariousness. According to Kalleberg (2009: 2), precarious work refers to 'employment that is uncertain, unpredictable, and risky from the point of view of the worker'. In a similar vein, Friberg et al. (2013: 112) conclude that rather than a lack of income, 'precariousness is related to lack of stability, security and control in one's employment situation'. Although precarious working conditions have become more widespread throughout the Western world during recent decades and are no longer only characteristic of migrant jobs, migrants tend to experience a higher degree of precariousness.
One cannot ignore the role of governmental and intergovernmental regulations in shaping the segmentation of the labour market. Here, the EU's principle of free movement of labour and services serves as an incentive for institutional changes at the national level; local state apparatuses, in their turn, serve as active progenitors of neoliberalisation 'from above' as they call for greater wage differences and the use of different standards for foreign subcontractors (Sippola 2012). The regulation by the nation state can also function in the opposite direction by curbing the effects of labour market flexibilisation through legislation. However, national and multinational employers as well as migrants themselves can still reinforce the flexibilisation tendency. As Meardi, Martín and Riera (2012: 7) argue: Cross-border worker mobility, especially but not only within the EU, offers a possible solution to the problem of combining flexibility and security: the creation of a hyper-flexible buffer of migrant workers who, being disposable in case of downturn, can carry most of the uncertainty burden without causing political problems.
We argue that migrants more often than not indeed find themselves in less advantageous positions than natives, but the position of migrants is becoming diversified according to their employment status within the segmented labour market. This might especially be the case in Finland and other Nordic countries where the labour market is still rather highly regulated and employers try to find ways to gain competitive advantage by circumventing national regulations (Herrmann 2008;Lillie & Sippola 2011) via, for example, employing migrant workers through foreign subcontractors. To fully capture the variation of employment statuses (in the labour market), and to incorporate the role of the state in the contemporary shaping of the labour market, we use the concept of labour market regime. The notion of 'regime' itself is based upon Elvander's (2002: 118) use of the term that embraces: …the totality of a country's labour market relations, from the workplace to the central political level where the political conditions for the development of labour relations are decided, such as general economic policy, labour market and incomes policies, labour law, etc. The scope and main direction of policies directed to the labour market will be an important element in comparisons between regimes or groups of regimes.
The regime concept has some parallels with concepts such as 'industrial relations (IR) systems' or 'IR models'. A difference between the IR concepts and the concept of 'labour market regime' resides in the fact that the latter attributes a more active role to the state.
Furthermore, the labour market regime presupposes some degree of permanence and pathdependency. Historical trajectories have been essential in shaping the Nordic labour market regimes (Elvander 2002). Potentially, there is one important feature in the Finnish regime which might prevent some employers from circumventing national regulations: the legal extension of collective agreements. In this respect, Finland differs from the other Nordic countries, in particular Denmark and Sweden (e.g. Ebbinghaus 2004). The Finnish wage bargaining system is based on sectoral collective agreements which set industry minimums and in most sectors, including construction, are extended over the whole sector by government decrees. Thus, also workers employed in companies without collective agreements are covered by the minimum provisions set in the legally extended agreements.
The conditions of sectoral collective agreements are also extended to posted workers; posted workers can also authorise trade unions to take legal action on their behalf against their home-country employer for not complying with the host-country's collective agreement. Our notion of migrant labour regime, however, cuts across the original concept as Elvander (2002) sees it, which makes our term more distinctive. We argue that the emerging Estonian-Finnish migrant labour regime has dissimilarities with the Finnish labour regime for natives in two respects. Firstly, as those coming from outside into the Finnish labour market will often remain outsiders (culturally and economically), it is reasonable to speak in terms of a distinctive cross-border regime reserved for migrant labour. Secondly, as our definition of 'regime' also involves supranational (EU) regulation, it makes sense to differentiate between the labour regime for locals -which is mostly national by nature -and the labour regime for migrants -which entails quite a high deal of European, regulation.
The concept of a migrant labour regime provides the state and social partners with an active role. One proposition of this study is that in the host society the national regulations that are The financial crisis of 2007-2009 gave a major boost to individual emigration from all the Baltic countries, which was connected with the deteriorated social and labour conditions in these countries (Sippola 2013). Based on the interplay of regulations, employer and worker strategies, we argue that the regime has become more diversified and segmented according to the pattern of firm ownership, as it is no longer predominantly based on the posting of workers.
Also, Wills et al.'s (2009) study on the division of migrant labour in London provides an important contribution to understanding the shaping of a local labour market regime in terms of migrant citizenship status, employer preferences and government regulation. An important observation by Wills et al. (2009) -which is also acknowledged in the present study -is that the regime undergoes constant changes due to intersecting decisions by the government, employers and workers. Our approach introduces the internal differentiation of the migrant labour market regime on the basis of one's employment in different patterns of firm ownership, which has not been elaborated on before, but according to our data is an important factor contributing to the segmentation of the regime.

The construction sector in Finland
In the following we will give a brief overview of structural features of the Finnish construction sector. There are around 170,000 employees in the sector, of whom 100,000 are in manual jobs; 20-25% of the latter are foreigners, around a third of whom are working in the Helsinki metropolitan area (Alho 2013). Estonians form the biggest foreigners' group in the construction sector (Statistics Finland 2013). The unionisation rate of migrant workers is relatively modest. Alho (2013) has calculated that in 2012 the unionisation rate among migrant construction workers was 12-14%, which is far from the 70% rate prevailing in construction in general (Union official #3).
International subcontracting and cross-border labour mobility, especially the posting of workers, have characterised the labour process of the construction industry in recent years (Lillie and Sippola 2011). This development has contributed to obscuring the main contractor's liability for the wage and working conditions at the lower end of the subcontracting chain. Thus, the main contractors are capable of evading the norms of the Finnish labour code by using foreign (mostly CEE, especially Estonian) subcontractors, who bring their own low-paid labour onto construction sites (Lillie and Sippola 2011;Sippola 2012). Foreign labour and especially workers posted from abroad as 'service providers' bring forth such circumstances on the sites that workers executing the same jobs can have different labour conditions and their social security contributions are made to different countries -or they are not made at all. Cases in point are as follows: the commuting of Estonian workers from Estonia to Finland, which has created a 'translocal' labour market in the construction sector (Alho 2013) and the use of Polish posted workers at Olkiluoto 3 construction site, which has turned the site into a 'space of exception' in regard to the Finnish industrial relations system (Lillie and Sippola 2011).
In recent years the Finnish government -first and foremost pushed by the construction union Rakennusliitto -has taken deliberate steps to reregulate the construction sector. Since January 2013, all construction workers are obliged to obtain a tax number, which has enhanced the controllability of construction sites. In July 2014, the so-called 'construction site register' and main contractor liability were introduced to get fuller access, for example, to hours worked and tax information of all individuals involved in a construction project. A change in the subcontracting chain management of the main contractors is visible: the posting of workers is no longer the most favoured form of operating in the Finnish construction sector. Instead, a new practice of prodding Estonian firms to register themselves in Finland can be seen. Earlier it was easier for both domestic companies and transnational firms to find alternative ways of circumventing institutional constraints at the national level (Herrmann 2008;Lillie & Sippola 2011). We assume that these developments have had an effect on the restructuring of the migrant labour regime as well.

Methodology
This study applies a migrant-centred focus on studying the structure of the Finnish migrant labour regime. The analysis is mainly based on 18 biographical narrative interviews (Schütze, 2005) conducted during 2013-15 with Estonian men (as the sector is highly male-dominated) working in the Finnish construction sector (see Table 1). Although a couple of them were small business owners, they also worked in their own company as construction workers. In addition, we made 7 follow-up interviews 1.5-2 years later. Respondents were found through the personal contacts of the research group, snowballing and also social media.
Using workers' employment stories to examine the position migrants occupy in Finland is especially valuable as statistics about different forms of employment are either not available or not very reliable, given that companies often use illegal employment practices that are not reflected in the register data. The biographical approach also enables us to track the changes in the employment situation, analyse the meanings people give to them and examine what kind of barriers and opportunities migrant workers experience in the Finnish labour market.
Importantly, these interviews provided us insight into the situation related to different kinds of employment arrangements and enabled us to construct the notion of 'pattern of firm ownership' that helps to understand the internal division of migrant labour regime. * Also follow-ups ** Often the workers themselves do not see a difference as to whether they are posted, agency workers or in a normal employment relationship; the blurred borders between 'posted' and 'normal' statuses have also been evidenced elsewhere (Eldring 2011).
Although the main body of the data builds on the migrants' views, we also wanted to include information on policy developments influencing the migrant labour regime. Therefore we These statistics do not show the full picture of the phenomenon, as not all companies apply for these forms, and some of the postings fall into the grey area, but they do indicate the major trends in the construction labour market.
Union officials' interviews provide background information about the change in the regime.
After the introduction of tax numbers in the construction sector in 2013, there appeared 10,000 -11,000 more workers in the sector who turned their 'black' status into official employment (Union official #1a). At the time of the interview, there were around 80,000 Estonians on the pension insurance register in Finland plus around 15,000 posted workers; these numbers included, however, workers from all sectors. However, the union official noted that the number of posted workers was decreasing in the construction sector, while Estonians seemed to have been given more permanent positions in enterprises (Union official #1b).
Rakennusliitto's cooperation with the employers' association in the construction sector, the Confederation of Finnish Construction Industries RT, has been fruitful in initiating regulatory changes in the sector. This started with joint efforts in lobbying for the law on main contractor liability, followed by the reverse-charge VAT system and culminating in the introduction of mandatory tax numbers in the construction industry in 2013 (Union official #2). Finally, the construction site register was introduced in summer 2014. Under the new system, the main contractor is liable to provide information for the tax authority on the employees, subcontractors' employees, sums of contracts and wage sums of the entire site.
The union official (#2) insisted that in this way the Finnish contractors, who do their work in accordance with the rules, will gain back the competitive advantage they have lost due to grey economy actors. In the official's (#2) opinion, the tax numbers and main contractor's liability have worked well. What is more, the tax authority got more resources for have to make sure that nobody should be able to work illegally in Finland. consists of somewhat more privileged workers in terms of income and autonomy, but they are still precarious in terms of job security (there is no guarantee that work will continue; the responsibility for employment is completely on them). The fourth group, consisting of those workers employed in a Finnish company, is the one in which the workers are in the most secure position in terms of income, employment and rights. It would have been possible to distinguish those employed by Finnish temporary work agencies as a sub-category, since such workers have a distinctive position in the contracting chain, and they risk being employed for shorter time periods. However, we decided to include agency workers in the 'employed by Finnish companies' category, because migrants did not perceive it as considerably unfavourable to being employed by the Finnish construction company. It was argued that wages are good and paid according to the rules, regular working hours offered and employment contracts extended.

Posted workers
Although only one of the construction workers we interviewed was posted at the time of the interview, several of them had experiences of being posted during their earlier careers in Finland. The stories about being a posted worker indicated that these were difficult times characterised by uncertainty, poor working conditions and disputes with employees over unpaid wages. if need be.

Employed by a firm registered in Finland but operated by Estonians
Some accounts of the interviewed workers let us understand that there is a tendency that Estonian firms move from posting (of Estonian workers) to registering a company in Finland and then to employing Estonian workers. The reasons for this shift can be derived from both the concern for the reputation of Estonian companies in Finland and the stricter regulations in the Finnish construction sector. For example, Tom reported that the Finnish main contractors were not willing to subcontract purely Estonian firms any more. That is why the Estonian owner of the firm he was working for registered the firm in Finland.
Although the recent legislative changes had affected the situation, the Estonian firms were still able to circumvent the labour regime aimed at native workers. For instance, Peter, an interviewed Estonian worker who later became a Rakennusliitto official, regarded the tax number system as ineffective since the tax authority was not authorised to completely inspect working hours. Peter also maintained that it is easier for a Finnish contractor to deal with Estonian workers, since the former has no obligation to arrange fixed working hours or take financial -for example tax payment -responsibilities for the latter. The influx of cheap labour from Estonia has meant that the Finnish construction workers have had to accept lower hourly wages (which however are according to the construction collective bargaining agreement) instead of having higher contract-based wages. This is due to the pricing of the contracts below the Finnish standards by the Estonian firms. In Peter's view, Finnish companies were equally guilty for this development as main contractors.
The stories the migrants working for Estonian companies registered in Finland told us indicated that although through this kind of arrangement working conditions and wages might have been better than when posted, there were also several factors that remained 'Estonialike'. For example, long working hours resembled those of posted workers. Although Timo was locally employed in Finland, his work schedule hinted at an arrangement typical for posted workers: 3 weeks work in Finland, 1 week free time in Estonia. Another resemblance to the Estonian labour regime is that the Estonian employers did not generally allow their workers to belong to a union.
Similar patterns of irregularity and insecurity as in the case with posted workers were seen in Anti' story. He was generally unsatisfied with his position and considered working under Estonians as not being very beneficial for workers, as 'under Estonian management […] the system is still like you are working in an Estonian company.' Anti was, however, unmotivated to change this arrangement due to his current flexible work schedule and to his ultimate plan to move back to Estonia. In the meantime he had to be flexible, as he did not have a permanent contract and there were breaks without any work between successive temporary contracts given by his employer. He had complementary job spells as a painter, a job he needed for the breaks when he was not offered any work. The introduction of the tax numbers had had some effect on the conduct of the Estonian employer, but there was still 'some kind of tricking' by the employer. The payment of the salary was often irregular. Anti felt himself to be bullied by the employer, when he found out that he was the last one to whom the employer paid the salary. This is a telling example of how the workers can be at the mercy of a single employer: I knew that I am going to be paid, but it really got to me that I was the last one /.../ I do not believe that he [the manager] kind of, or directly thinks of me badly, but, but yes, he just, like he knows, that I will manage.
Valdo once worked for a company that was registered in Finland but that had Estonian owners and similarly to Anti he was very sceptical about Estonian employers in Finland: '…well, let's say 99% of the time, if you are working for Estonians then you either don't get paid or they don't pay you that … you know, all the extras and things that you need for construction, you usually don't get that.' Valdo belonged to Rakennusliitto. He regarded the union as powerful enough to curb the illegal activities of Estonian companies: But yeah, the Rakennusliitto, they can go to the workplace and stop a company from working, for example. They ban you and you can't work, for example. It happened to me once at a company. We went to work in the morning, the big boss from the main contractor came and said … the guys from XXX [A Finnish company with Estonian ownership] … you are not allowed to work in Finland. So you can't work. Go home.
Valdo told, however, that the boycott imposed by the Rakennusliitto and the ban on business operations imposed by the authorities did not prohibit the company owners from establishing a new company with a different name and starting operations again in Finland.
Evert's story illustrates how the Estonian construction business arranges its human resource policy in a way which is beneficial for Finnish contractors. He was employed by a middlesized Estonian construction company (with around 100 employees), in which almost all of the workers were Estonians. The company did subcontracting for major Finnish construction companies in the biggest Finnish towns. He mentioned that he had applied for work for a Finnish main contractor (for whom they did subcontracting) a couple of times. The fact that the main contractor refused to hire the Estonian (subcontracting/agency) firm's workers bears resemblance to Lembit's story (see 'Posted workers' section). When applying for a job the first time, Evert did not even receive a reply because the company -as he was told -did not need any workers; however, that company had a new Finnish worker hired the following week.

Self-employed and small business owners
On the basis our analysis, we regard that being a self-employed service provider or a small business owner might provide better labour conditions for Estonians in the Finnish migrant labour regime than the statuses of being a posted worker or employed at Estonian-owned companies registered in Finland. First and foremost, it is manifested in better remuneration and a higher degree of autonomy, although in regard to employment security this kind of position is less secure and predictable than being employed by a 'real' Finnish company. It is also to be noted that certain Estonian owners of construction companies are in comparatively advantageous positions (they sometimes exploit their co-nationals), but those are beyond the scope of this article as they represent the position of managers or employers rather than that of workers.
Veiko had established a small firm of his own. Through his company, he provided his services in the construction sector in both Estonia and Finland. The idea of establishing the company derived from the seasonality of his earlier job spells in Finland; typically he worked from March-April until November-December, and then he was sent on unpaid holiday. So far, however, the emphasis of his business activity had been in Estonia, and there had been only a few contracts in Finland. He acknowledged that working in Finland is project-based and seasonal; however, he did not seek more secure, long-term commitments. As an Estonian employer in the Finnish market, Veiko saw the pros of the well-regulated Finnish system, although foreign contractors were undermining it. Veiko's account, therefore, reflects awareness of the positive side of regulation on the one hand, but also acknowledgement of the fact that a foreign employer can circumvent institutional constraints on the other: In many respects, the system like this is good. When you [the employee] achieve a certain level, you see that you are supposed to have such a wage. But the market puts everything in place […] one told me that their firm does not get contracts anymore, because Estonians and other foreigners have put the market price at a low level. There is, indeed, quite tight competition in that respect.
Arno, who was currently working in a Finnish firm, also pondered the pros and cons of entrepreneurship. Some of the colleagues with whom he used to work in Estonia were self-employed in Finland. They appeared to have a better income than Arno. While his net earnings per hour were 13 Euros, he argued that as a self-employed person he could get 22 Euros, minus taxes and other costs. However, Arno maintained, the main contractor had no responsibility for a self-employed person. Were he 10 years younger, he would consider becoming self-employed. He was also aware of the downside of being employed by himself: 'there wouldn't be any work in the wintertime.' The difference between the statuses of being self-employed and a wage-earner is seen in the fate of three fellow workers, Jaan, Rein and Edgar. Jaan had worked in Finland for 7 years,

Employed by a Finnish firm
Working for a 'real' Finnish company, even if only a temporary agency firm, was seen as something desirable but hard to achieve by several of our interviewees. Despite his employment as an agency worker, Vello was quite satisfied with his position as carpenter at one of the biggest Finnish labour agency firms. Although he did not have a carpenter's degree in Estonia, he was given a salary corresponding to a carpenter's qualification. However, the work was hard, and he was interested in educating himself to do a less-demanding occupation, such as that of a crane driver. Vello realised there would be better paid positions in other firms, but he was afraid of changing the job because there would be a 4-month probation period during which the employer could dismiss him at any time. He had tried to get onto courses to get educated as a crane driver, but his employer (the Finnish labour agency firm) had refused to pay for that. He considered that this was odd: the company would benefit from his higher qualification.
After having been employed by the same Finnish agency firm as Vello, Arno had managed to get a permanent job at a 'real' Finnish construction firm. He worked as an excavator driver.
The difference between these two workers seemed to have been that Arno had the required qualifications already when coming to work in Finland, whereas Vello had just started to pursue a higher qualification. Arno clearly saw the pros of being employed by a Finnish employer: 'Well, he pays the salaries, and… working hours are at 8 hours a day, you are at home on Saturdays and Sundays and nobody forces you to work more than that.' In the current firm, Arno worked without a written employment contract, or at least he had not seen the contract. However, 'the salary has always been put onto the account.' The All this seems to be evidence that despite the Estonian workers' ability to climb up the career ladder up to the position of 'employed by Finns', they still seem to be locked in a less favourable position than the native workers. There seems to be, however, an exception to the rule, which is Dagmar's story. Dagmar's story of getting a job in Finland is exceptional, since he got it through a trade union contact in Finland while being a trade union activist in Estonia at the beginning of the 2000s. He called the employer and asked for a job and got an affirmative response. In his words, the employer appeared to be benevolent especially to those foreign workers who had been fired by other (dishonest) employers in Finland. The employer had also a positive attitude towards the trade union membership of the workers.

Concluding discussion
The aim of this analysis has been firstly to throw light on the drivers of regulation that lie behind the emergent migrant labour regime in the construction sector in Finland. Although employer and labour strategies are important in shaping the regime, the roles of EU (e.g., through different directives) and government regulation, which are influenced by social partner institutions, remain crucial. There are clear signs indicating that the policies of the Finnish government alongside trade union efforts and lobbying activities (sometimes in cooperation with the employers' federation) have had a positive overall effect on the evolvement of the construction labour market. Thus, a strong industrial relation system such as that in place in Finland seems to be more resistant to the corroding effects of postaccession labour migration than other systems; even within the single EU labour market. The Finnish government, in contrast to the effects seen elsewhere, has been able to deter the undermining of the strong industrial relations system in construction (Wagner & Lillie 2014).
Further research is needed in different sectors and national contexts to analyse whether also alternative ways of circumventing institutional constraints at the national level (Herrmann 2008;Lillie & Sippola 2011) can be prevented by interventionist regulation and cooperation between authorities at the national and international level.
Secondly, we have explored the current migrant labour regime apparent in the Finnish construction sector. We have identified four differing positions within the regime: posted workers; those working for a company registered in Finland but owned by Estonians; selfemployed or small business owners; and those working for a 'real' Finnish company. We have seen that the pattern of firm ownership -reflecting both the nationality-based division between the main contractors and subcontractors and the workers' relationship with and distance from the main contractor -is a major factor shaping the structure of the emerging migrant labour regime. The translocal labour market setting that has emerged between Estonia and Finland has generated different worker positions based on their employers' nationality, the relationship to the main contractor and workers' contractual status that reinforces the segmentation of labour.
This study further reveals that irrespective of employer strategies, the Estonian workers had managed to improve their position through hard work and lucky coincidences, partially rendered possible by the recent development of labour market regulations. There is indeed a variety of forms of the use of labour, and the position of migrant workers might have improved compared to that of posted workers. However, the migrants still do not regard themselves as 'full' members of the labour collective. Although the analysis showed marked differences between the four positions in the construction labour market -the first position (the posted workers) being most and the fourth position (those working for a Finnish company) being the least precarious and low-paid -the whole migrant labour regime seems to be based on the inferiority of the Estonians. It is remarkable that the Estonians were still in a more disadvantageous position in terms of career advancement, cultural acceptance by locals, dependency on the employer and flexibility demanded from the side of the employer.
Moreover, the expectation of the employers that Estonians tend to be more flexible workers than locals often persisted even though the employer was not any more an Estonian. Whether this is due to the overall deterioration of the position of the workers in the Finnish construction sector or to a systematic tendency of segregation in the labour market remains a question to be considered in future research.
By means of the biographical lens, this study suggests that migrant workers' agency also plays a role in shaping the migrant labour regime. Seemingly, these Estonian workers desired to leave their precarious positions behind and enter the ranks of fully respected and valued worker-citizens (see e.g. Anderson 2013) by seeking employment in 'real' Finnish firms.
However, the Estonians appeared to seek improvement in their positions via becoming 'good workers' (MacKenzie & Forde 2009) -referring to qualities attributed to workers by employers -rather than demanding their rights as workers per se. Thereby, migrant workers were on the one hand able to gain a 'competitive advantage' over local labour and to secure employment, but on the other hand, they simultaneously consented to inferior employment conditions in comparison to natives. Analysis of different forms of workers' individual and collective agency was beyond the scope of this study but is an important aspect that ought to be scrutinised in further research.