Abstract
Introduction
European integration, and more specifically freedom of movement, has proceeded much more quickly in terms of economic freedoms than in terms of social and welfare policies. Indeed, whether social policy should be Europeanised at all is controversial (Daly, 2012). The transnational social protection literature describes migrant workers as planning their personal economic strategies across a variegated economic space, accessing state, market and personal network resources to compensate for their marginal position in any particular Member State (Levitt et al., 2017). In the EU, however, there is, in principle, a reciprocity-based social insurance system, which is supposed to ensure continued access for intra-EU mobile workers to prevent the risk of non-coverage becoming a barrier to free movement. On this basis, transnational social protection should be superfluous. National welfare states both redistribute wealth and socialise risk, and this is the standard by which we assess social protection. EU mechanisms focus more on socialising risk through mutual recognition and reciprocity than on redistribution. This means that EU systems provide rules on determining the access of mobile workers to Member States’ national welfare systems, with social insurance provided either by the sending state or the host state, or in certain situations provided by the host state, but paid for by the sending state.
In this regulatory context, first, we would expect sending and host state welfare mechanisms to dominate worker strategies in terms of ensuring security and minimising personal risk. Second, insofar as these fail to insure mobile workers, we would expect them to engage in transnational social protection strategies to compensate for and address failings in those systems.
This logic led us to design our research around three broad categories of workers: EU workers from wealthier states (Finland, Denmark, Austria, Italy), EU workers from weak welfare states (Eastern Europe, some southern EU countries), and workers who do not hold EU citizenship. The final category showed a clear lack of social insurance access, but for the others, the picture was more muddled. We found that the dominance of informal work arrangements means that workers’ rights must be enforced by the workers themselves. For those in a weaker position, whether due to skill, citizenship or lack of home-country social benefits, the gap between the rights available and benefits received is wide, linked primarily to employers cheating on benefits payments. We saw only occasional cases of workers using transnational social protection strategies, which we define as workers actively ensuring the availability of social insurance possibilities from private, public and market resources across a transnational space. Those from strong welfare states use (and sometimes abuse) home-country benefits systems. Those from weaker welfare states accept uninsured risk as an occupational hazard. Overall, this pan-EU context produces unmitigated social risks for migrants. These risks do not fall equally on all mobile workers, but rather are hierarchised both within and between labour markets.
All the intra-European mobile construction workers we interviewed were at one point or another ‘posted workers’, meaning they had been mobile as a dependent employee. This was part of our selection criteria. It became clear that, from their perspective, ‘posted workers’ do not constitute a distinct market segment, but rather one among several contractual forms under which similar forms of mobility exist. Although all interviewees had posting experiences, it is more accurate to refer to them as intra-EU mobile construction workers. Some were citizens of third countries resident in an EU Member State, but who work, or have worked, in an EU country other than their country of residence. Our study used a migrant-centric approach, like the transnational social protection literature, but bridges the EU social protection literature, through expert discussions, to show the gap between how the social protection system is designed, and how it operates in reality.
Social insurance, bounded welfare states and the pan-EU labour market
Nation states frame the provision of social insurance; the boundaries are sometimes geographical (services are provided to those on a given territory, but not outside it), sometimes based on citizenship or some state of intermediate residence (citizens are eligible and non-citizens are not, regardless of where they are at the moment), and sometimes contributory (those who have paid in for a sufficient time are eligible). However, transnational mobility creates exceptions in national social insurance systems because the basic design assumes a sedentary population. Whether specific benefits are based on citizenship, geography or contributions, welfare states as a whole and in general are designed to mitigate economic risk, and distribute income along the life course, and between social classes for a population that stays within state boundaries from cradle to grave (Faist, 1995).
Welfare states vary in their focus on insurance versus redistribution, and use different mechanisms to accomplish their goals, depending on the national welfare regime. Some systems rely on employment contributions and quasi-private funds, while others lean more on general taxation and residence-based eligibility, but most have a variety of arrangements, depending on the type of benefits (Esping-Andersen, 1990). Differences between these systems could well have implications in the pan-EU labour market, but that is not our concern in this article. For our purposes, it is sufficient that these differences between national welfare systems contribute to deflect the political project of building a pan-EU welfare state into a project of building instead a pan-European system based on reciprocity and mutual recognition. National welfare states normally have effective systems in place to ensure that workers and employers make required social insurance contributions. These are not perfect, but compliance by employers and workers is generally the norm, unlike, as we show, in the pan-EU reciprocity system for mobile construction workers.
The pan-European labour market is segmented into a variety of national welfare systems and labour markets. These national systems are linked via the economic and organisational ties of transnational actors through the infrastructure of EU welfare coordination, and through the careers and life experiences of mobile workers crossing national boundaries. Europe-wide nationality-based hierarchies interact with intersectional socio-economic and/or demographic labour market segmentation within countries (Arnholtz and Leschke, 2023). Welfare states, to a greater or lesser degree, can mitigate risk and reduce the effects of such hierarchisation, but only within their enclosed national spaces. The EU, through free movement of workers, opens up that space.
The opening of the EU labour market allows employers and workers alike to navigate a variegated regulatory landscape. The former can elect to engage with social protection systems or evade them (Berntsen and Lillie, 2015). The latter can potentially access social insurance resources from multiple locations (Levitt et al., 2017), but often only if their employers allow them. The transnational social policy literature puts the focus primarily on migrants’ agency in participating in the arrangement of their own social protection over a fragmented multiplicity of social protection ‘resources’. Migrants mitigate risk by locating and accessing state, informal, third-sector and market-based social insurance, available in different locations. This framework explains particular aspects of migrant behaviour and migrants’ ability to access resources from diverse spaces to compensate for their marginality. This reflects the lived reality of many migrants and compensates for the limitations of nation state-focused research. It also suggests an indeterminate outcome: a migrant might pull together more, or fewer, social protection resources than a non-migrant, depending on the situation and the migrant’s strategy. Our findings suggest that they are more likely to be disadvantaged than advantaged, though there is the theoretical possibility of windfalls from accessing multiple systems.
Building a pan-EU welfare system through mutual recognition
Social security coordination in the EU directly affects the economic well-being of millions of Europeans (De Wispelaere, 2023). The EU supports intra-EU labour mobility by ensuring that mobile workers remain protected when they move from one country to another (De Wispelaere et al., 2019). Previous research has shown that this coordination falls short for a variety of reasons, including poor information (Houwerzijl, 2018), and eligibility criteria that favour sedentarism (Baas, in this issue, 2025; Bruzelius et al., 2017; Fingarova, 2017; Gellérné Lukács and Gyulavári, in this issue, 2025). Precarious workers on shorter-term contracts tend to be particularly disadvantaged (Zabransky and Amelina, 2017). Migrant exclusion may be direct and intentional or indirect and incidental to other practices and circumstances (Bruzelius and Seeleib-Kaiser, 2021; Scheibelhofer, 2022).
The relevant legislative mechanism for addressing the portability of social protection is Regulation (EC) No. 883/2004 on coordination of social security systems and its implementing Regulation No. 987/2009. Article 12 of the Regulation applies to the case of posted workers and makes clear that such workers remain under the sending country’s social insurance system for a maximum of 24 months. Thus, for mobile construction workers who are posted from their home country only as part of a longer-term employment relationship with an (honest) employer, the situation is normally clear: they have a home-country employment contract, and social insurance deductions are paid by the worker and employer to the appropriate home-country funds or agencies, just as if they had never left home. However, mobile construction workers normally have careers consisting of dozens of employment contracts with a variety of employers, some as posted, some in the receiving country, and sometimes some in a third country. Many of these employers will not be honest in terms of paying social insurance contributions (Cremers, 2014; Lillie et al., 2024).
This multi-layered and complex system of regulations governing workers’ status presents employers with high potential for hierarchisation, not to mention employer non-compliance with the rules. Regulatory enforcement to ensure compliance would require an investment in expanded trans-governmental cooperation. At the bilateral level, such enforcement cooperation requires time, resources and consensus among the participants and it may prove unstable if migration patterns shift (Lillie et al., 2024). The system is scarcely seamless and provides unequal and differentiated security for different groups of workers (see Bruzelius et al., 2017).
The pan-EU labour market
The regulatory regime governing the pan-EU labour market virtually invites employer fraud (Cremers, 2014) and arbitrage (Berntsen and Lillie, 2015) between social insurance systems, reflecting, exacerbating and indeed encouraging east-west and north-south inequalities. For instance, workers on a given job site may be paid differently depending on where they are from. Despite free movement in the pan-EU labour market, wages between similarly skilled and productive workers are not equal: rather there is a strong home-country effect, meaning that a worker’s country of origin influences their pay in the host country. Leschke and Weiss (2023) show that, on average, migrants from Western EU countries obtain better wages when migrating within the EU than migrants from Eastern EU countries. The average skill level of Eastern EU migrant workers is lower, but the difference persists when holding skill levels constant.
This discrepancy could be also due to lower wage and social protection expectations (Ronchi, 2018). Eastern European workers, with few expectations concerning their coverage in the welfare states of the countries where they work, gauge their expectations based on their sending country, so that conditions in the latter continue to serve as reference point (Constant et al., 2017). This is particularly evident with regard to construction workers, many of whom remain embedded in their home-country employment and social context while working abroad (Caro et al., 2015). They are hired through social networks, on which they depend for continuing employment (Bagnardi et al., 2024). They strike an implicit moral contract with their employer, on home-country terms, which prevails even in the face of conflicting host country norms (Matyska, 2019). The flip side of this is that workers, too, have market agency. Many construction workers consider foreign jobs as a natural complement of domestic ones, moving readily between jobs both to find better opportunities and to exit from jobs with poor employers (Berntsen, 2016).
Firms are the more powerful actors, however. Different types of firms engage with the labour market in different ways. Established construction companies, for example, often post their permanent workers abroad to fulfil contracts, with the expectation that they will continue their employment in the sending country, or at other job sites, after the job is done. These firms sometimes violate host country employment regulations, but often maintain an implicit employment agreement with their own workers, which causes workers to cooperate in undermining enforcement (Matyska, 2019). Others strategise across labour market hierarchies in the transnational EU space, arbitraging wages and social payments to find the cheapest solutions possible (Berntsen and Lillie, 2015). Labour-only companies and temporary work agencies hire workers for specific jobs and send them abroad, or recruit migrant workers from abroad. These firms have little to lose if they are caught violating wage minimums or social insurance contribution requirements. Firms may transfer workers between different types of contract, such as between posted contracts and local ones, or direct employment contracts, and notional self-employment. In Finland and Norway, by shifting posted workers onto local contracts, firms can avoid paying
Contracts may also be concluded with firms incorporated in third countries. Often these are so-called ‘letterbox’ companies, created explicitly for the purpose of putting the employment relationship under the jurisdiction of a selected low-cost legal and welfare system (Cremers, 2014). Posted workers from non-EU countries, or third-country nationals (TCNs), are always ‘re-posted’, by definition, because the legal sending country is not the workers’ home country. The disembedded nature of their situation seems to push them into vague employment arrangements similar to letterbox firms, though this vagueness makes it hard to pin down from their narratives exactly what types of firms and contracts they are employed under. Third-country nationals who are resident and employed in an EU Member State can be posted to other EU Member States, which circumvents the normal migration restrictions, often regulated by strict quotas and occupation lists. This adds another layer of vulnerability and makes it even more difficult for them to access social insurance. Re-posting is frequently a red flag that an employer is evading regulations, also when the posted workers are EU Member State citizens.
Methodology
The project Secure Mobility: Uncovering Gaps in the Social Protection of Posted Workers (SMUG), funded by the European Commission, conducted 61 biographical interviews in six EU countries, namely Austria (n = 8), Denmark (n = 12), Finland (n = 13), Italy (n = 10), Poland (n = 8) and Slovenia (n = 10), as well as one focus group interview (FGI) per country. We directly reference only interviews gathered by our own (Finnish and Austrian) teams, or published excerpts of interviews from the other teams’ interviews, from reports, journal articles and policy briefs. We know from our joint analysis together with the other teams, however, that the interviews to which we do not refer show similar patterns.
The selection criteria for our interviewees were that the worker should have had at least one posted job experience in construction. In each country we interviewed both outgoing and incoming postings with respect to that country. This means, for example, that the Finnish interviews include Finnish, Estonian, Lithuanian, Polish, Ukrainian and Moldovan workers. We chose the construction sector because of its highly mobile and difficult to regulate workforce, as shown widely elsewhere in the literature (Bagnardi et al., 2024; Berntsen and Lillie, 2015; Cremers, 2014; Lillie et al., 2024). All interviewed workers were male because very few women work in the industry. There are mobile workers at all skill levels, from unskilled to university educated. Mobile construction workers are notoriously difficult to contact for a variety of reasons, and this was doubly so during the COVID-19 pandemic. We used a variety of approaches, including personal contacts, union contacts and directly approaching workers on sites. From these initial contacts, we used snowball sampling. Further details of the interview samples from Finland and Austria are presented in Appendix 1.
The interviews were conducted under a jointly developed protocol following past research (Mrozowicki, 2011), although some were done by phone or video call, because we interviewed during the COVID-19 restrictions. While ‘posting’ proved not to define a category of worker, it remained a good selection criterion as it led us to workers who had had highly mobile careers in construction. The interviews were transcribed, coded using pre-defined substantive themes linked to different aspects of social insurance (SI) access and use, and analysed thematically in a joint cooperative process with all project teams. Our biographic method enabled us to see how workers’ careers develop and intersect with social welfare planning. We then discussed our processed and anonymised interview data with sectoral social partners and government officials in focus groups. Focus group discussions were also coded and analysed thematically.
Findings
In this section, we present and compare how differently situated mobile construction workers talk about their working conditions, entitlements and social protection. Using these narratives, together with the focus group interviews with key stakeholders, we show that workers with higher capacities – better skills and education – use more transnational social protection strategies, and workers’ perceptions of access to home-country social insurance increase their confidence and security during their mobility experiences.
Workers in different positions in labour market hierarchies have different starting points, in terms of how they perceive their agency and their need for agency – in other words, the degree to which they believe they can affect their social insurance circumstances, and also the extent to which they feel they must try to affect their social insurance. Workers’ notions about what kind of leverage they have with employers and host country authorities, and their awareness of how much they can influence their employment situation also affect their confidence in dealing with employers and with authorities on social insurance issues.
Regardless of national origin, a marked characteristic of the labour market experience of these mobile construction workers was their reliance on personal and professional networks to find employment opportunities, and to a certain extent also to resolve problems at work. These networks sometimes blurred the difference between friendship and supervisor-worker relations. These networks are a source of labour market strength, but also a constraint, because workers who complain, and are known to present a negative attitude are less likely to be hired, meaning that complaints that are legitimate might go unvoiced.
Worker relations to home-country social insurance
Workers from western European or Nordic countries exhibited a high degree of confidence in their economic security and insurance systems. Many of these workers spent time under other social insurance systems as well, but rarely expressed problems or concerns. For example, one Finnish carpenter lived for a long time in Sweden, having moved there to work as an individual, and accessed the social insurance system there; on other occasions he was posted within the EU, and mentioned no problems in those postings. He was also posted to many locations outside the EU. In one posting to Russia, he and the entire crew quit because the conditions seemed ‘shady’ (FI 2). An Austrian construction engineer mentioned a similar experience (see below AT6).
Importantly, the Finnish workers were covered by either home-country or foreign collective agreements, depending on whether their mobility status was a posting. As one of the sectoral stakeholders in Finland put it:
Mika [a Finnish worker, pseudonym] can safely go out into the world, he has a safety network. When people come here [to Finland], then it is just a matter of luck, they’re dependent on the employer, in a bad sense. (FGI, FI)
The Finnish workers did not feel particularly dependent on the employer and had experienced neither social benefit exclusion because of employers or government policy nor being treated worse because of their nationality. Sectoral experts in Finland noted that Finnish workers do sometimes work without contracts, but this is so as to be able to (illegally) collect unemployment insurance while also working (this was also seen in the biographies, but mentioned only as something someone else was allegedly doing).
The workers in this group, coming from countries with generous welfare systems, shared generally positive experiences, including of social benefits and protection. A foreman from Finland (FI1) mentioned that the countries he had worked in, Switzerland, Norway, Sweden and Germany, had similar wages, workplace safety and insurance arrangements, and he regarded them as unproblematic. He brought up periods of unemployment between jobs, during which he was able to draw on his Finnish trade union unemployment insurance, even though most of his work was abroad. A lack of language skills made everyday life abroad challenging, although he saw this as more of a learning opportunity rather than a problem. In any case, it did not seem to influence social protection access as he could always fall back on the home-country system. On the whole, these workers tended to regard themselves as the authors of their own destiny and were confident in their ability to negotiate challenges.
Well of course, thankfully people have this way of pushing bad experiences aside, while the good always stay on the surface [. . .] of course small complications come along the way. [. . .] Of course people don’t always get along. But you can always negotiate. (FI1)
In this sense, the EU posting and reciprocity-based social insurance system seemed to be working for these workers. They often enjoyed their postings, and appreciated the opportunities, but also told of job opportunities in their own country:
So if you’re thinking, when you say security, you mean like laying off? [. . .] there is plenty of work. And on top of that there is also a large labour shortage, mostly for engineering and other skilled positions. So this danger does not exist. (AT5)
They were not willing to accept poor work conditions, and unlike workers from weaker welfare states or from outside the EU, these workers rarely talked about local workers earning more, or receiving better benefits, themes that often arose for other, less privileged groups. They rarely mentioned employers cheating them, although one Austrian construction engineer mentioned a case of his employer not paying insurance contributions on a site in Slovenia, causing the workers to rebel collectively. He left the site and went home (AT6).
Similar to workers from wealthier welfare states, as we found from the Finns and Austrians in our sample, workers in managerial or professional positions, and in more highly skilled positions, posted within the EU, regardless of country of origin, often retained ties to their home country’s social insurance system. Among the interviews in the SMUG project, some Polish and Italian posted workers related experiences of this kind. One marker for this was that these workers tended to hold secure employment contracts with companies in their home countries. These contracts, unlike those of more precarious posted workers, would extend before and after the posting. The Polish posted managers and professionals were under their own country’s social and welfare system during their postings and were aware of their home countries’ labour laws and rights. However, unlike the Finnish or Austrian workers, Polish highly skilled workers sometimes had to be more proactive in order to ensure their social insurance, as one Polish professional related (FI6). Like many other workers he was willing to accept a variety of employment relationships, including a self-employment contract, which had pitfalls of its own. The worker saw self-employment as a good opportunity because of the higher income but he did not know how social insurance would work, and whether it would affect his benefits in the future. This worker’s transnational contractual payments were quite complicated:
As regards [my work in] Netherlands [. . .] this worked strangely, because my salary was transferred directly from Dublin, and I don’t know how pensions work there. [. . .] In Sweden [when self-employed], it is also difficult to say. There are institutions and everybody has a tax number and so on. But the road map is missing, what will happen in the future with our retirement and the contributions we paid. (FI 6)
The difference between this worker, who is well aware of the problems and is planning for the future, and those in the next category, who express less trust in their own agency, may be more a matter of how he articulates his agency than an actual ability to ensure social insurance will be available.
Less skilled construction workers from weaker EU welfare states
The workers in this category are a diverse group, but were less confident in their ability to exercise agency, and also less confident in the security offered by their home country’s social insurance. They all held EU passports and reported that they went to work abroad because of the socio-economic situation in their home countries. Some were more vulnerable than others, and importantly, workers and sectoral stakeholders alike mentioned that some EU workers, particularly Bulgarians, Romanians and Poles, sometimes fell into situations similar to those of third-country nationals, as detailed in the next section.
Many of these workers regarded ‘grey’ employment simply as an economic reality: the competitive situation for firms in the lower part of the contracting chain is such that they must cheat to remain competitive. As one Estonian scaffolder put it:
if you want to run a company and pay all the taxes absolutely legally, paying every penny, you need to employ 50–100 workers. If you only have 10–20 workers and you pay all the taxes, you will not have anything for yourself. That’s why everyone uses all kinds of schemes. (FI7)
Although such firms operate illegally, they position themselves strategically to foster trust with target migrant workers. As in other segments, current and past workers form and register their own companies to post friends, family members and colleagues abroad, benefiting from their understanding of the ‘posted employment’ regime, and their rapport with fellow mobile workers.
Considering their personal employment potential and socio-economic conditions in home countries, the workers in this category were usually willing to accept a job with poor working conditions and rarely worried about the possibility of exclusion from social benefits. This is illustrated by the following statement from an Italian worker:
10 years ago it was about the money because posted workers could earn twice their Italian salary. [. . .] Now the main reason is that workers cannot find jobs in Italy and therefore they move in order to work. 10 or 15 years ago they could choose: A double salary or working close to the family. Now this second option is not there anymore, so you can find many that just want to work, they don’t even care what will they do abroad as long as you guarantee them a job. (Italian Team interview, cited in Ndomo and Lillie, 2022: 21)
Italian workers are in a somewhat ambiguous position, in that many mentioned being driven by unemployment to go abroad, as well as poor conditions and discrimination relative to locals and a lack of confidence in their social insurance system. However, others mentioned accessing unemployment insurance occasionally, similarly to the Finnish workers.
While a frequent reason for accepting poor contract conditions was a lack of job opportunities in Italy, workers from Eastern European countries referred especially to income differentials and the need to support their families at home (AT2). This meant working harder, but for less pay than local workers and often also having poorer social protection. It often also meant not getting the social entitlements of locals, and therefore not being covered adequately. Some workers noticed that employers were aware of what kinds of jobs and working conditions were acceptable to different groups of workers and offered jobs accordingly.
The sectoral social partners in Finland and Austria regarded self-employment in posting as simply another form of employer cheating. They explained that employers would present these contracts to migrants as desirable and better paying. The workers only found out later that they were getting less net income, because they also had to pay the employers’ share of the social insurance contributions (FGI, FI).
Unpaid wages and unpaid employer social contributions often go together, but unlike non-payment of wages, non-payment of social contributions often goes completely unnoticed by the worker. Some find out by chance, or because wages also went unpaid, as in the case of one Slovenian worker, who left for another job as a result:
I got a shock [. . .] Everything was fine except that the company was not paying any social contributions, which I did not know from the beginning. [Then] the wages fell behind. In the end, I did not get paid those wages and since May [. . .] no contributions have been paid. Well, now I have moved on, because I have no choice. (Slovenia Team Interview, cited from Lillie et al., 2022: 8)
Many told us they just assumed that social insurance contributions were not being paid, and took the attitude that there was nothing they could do about it, so they were not particularly concerned. This, however, often became a problem when the worker got sick. This was the case with regard to a Lithuanian working in Finland, who talked about interrupting his contract to go home when he fell sick as it was impossible for him to obtain support from KELA, the Finnish Social Insurance Institution. Often, they did not try to claim host country benefits. An Estonian worker told us he had not bothered to figure out how health insurance works, but instead paid out of his own pocket for health care.
I had a shoulder [injury]. The collarbone came out of the socket. [. . .] I visited the doctor in Estonia, never here [in Finland]. What else can you do [but pay out of pocket]? Many work through Estonian companies [as posted workers]; they at least have Haigekassa [Estonian social insurance]. I don’t have Haigekassa. Now KELA would perhaps pay something when I need surgery, but I haven’t given it much thought or found out the details. (FI7)
Some workers and sectoral stakeholders from both the Austrian and the Finnish sample pointed out that employers try to prevent workers from obtaining information on wage levels from locals or from investigating their eligibility for social insurance. This was exacerbated by workers not knowing how to find information, especially when they do not know the local language.
Posted third-country national workers
The third-country nationals we interviewed in this context were all posted workers, as they did not have free movement rights within the European Union. This meant that their mobility could be legally justified only under free movement of services; in other words, they could not migrate for work from one EU country to another, but their employer could send them. Many of those interviewed by our research teams were probably not posted legally, so this categorisation could be contested, but all experienced their status as a legal grey zone. Their right to remain could readily be challenged because their right to remain in their country of work derived from their employer’s free movement rights, not their own. Cases of working without contracts and being paid in cash were reported, as in the case of a Bosnian worker interviewed by Austrian team (AT7). Third-country national workers’ narratives were similar to the most precarious intra-EU posted workers, but generally worse.
Third-country national posted workers usually have little connection to their nominal ‘home’ country of EU residence, which should in theory also be the country where their employer is incorporated, where social payments are made, and from which social insurance can be claimed. From the stories of our interviewees, however, it is clear that many, if not most, third-country national posted workers had little awareness of and no practical access to social insurance or services, despite, in principle, having the same rights as EU posted workers to sending-country social insurance. They also told us that their income was much lower, and the workers perceived their tasks as harder, working conditions poorer and pay lower than those of EU workers and locals. One Moldovan said:
This guy I know, also from Moldova worked with me for some time, then he received a Romanian passport and found work elsewhere, hired directly but by a Finnish company, paid €15 per hour, also as a welder. But given that we are third-country nationals, we receive about €7, €8, €9 per hour, depending on experience. (FI4)
The third-country nationals accepted the situation, as their primary concern was to keep their jobs, and the pay was still better than in their own country. Many had complicated migration and posting trajectories. By definition, they had migrated to an EU country, were notionally resident in that country, and were legally posted from there. In other words, third-country national posting involves at least three countries, but many told even more convoluted stories of being moved from place to place, under varied and vaguely defined work arrangements.
Some of the workers were aware that benefits existed (such as extra provisions for overtime), to which they had no access due to their nationality, but the reason for this was not made clear to them. From their perspective, whether it was due to illegal actions by their employers, or to official legal exclusion, the result would be the same: they would not have access. Some interviewees did not know what would happen if they were to take sick leave. Some considered it normal to be asked to work when sick or not be paid during sick leave. In other cases, the workers were not sure how things worked, but trusted their employers to take care of things.
Furthermore, some workers reported working for extended periods for little or no wages, and these workers generally did not have proper paperwork and contracts, and therefore no access to social insurance. Often, workers would find out about social insurance payments that should have been made, but were not, when a problem arose, and suddenly insurance or sick benefits were needed. Re-posting, or sending a third-country national from one EU country to another, was the cause of many complications. One Ukrainian told us:
I work for an Estonian company, but this company hires me out in Finland, to a Finnish firm. But all my documents are connected to this Finnish firm. The social insurance, pension contribution, the health insurance I pay in Estonia, but in Finland I pay the income taxes. This is all legal, according to Estonian and Finnish legislation. (FI5)
Taking care of transnational pension matters is difficult and not something that a worker can do without professional help, as a Moldavan worker interviewed in Austria reflects:
You have someone to help you [. . .] is there an agency or someone?
No, just private with someone, paid, you know how it is, someone who knows how to speak, who knows the laws, someone from our country, from Moldova, you ask them, you pay and everything is settled. (AT2)
Finally, third-country national workers who we talked to did not feel they could challenge their employers, or even ask about social insurance access or rights. They seldom knew anything about unions and did not consider the authorities to be potential allies in accessing social insurance.
Conclusion
In this article, our analysis relies on worker narratives rather than on official data on ‘real’ social insurance access and use. We corroborated our findings with social partners and public stakeholders that we consulted in focus group interviews. We found that regardless of what the regulations and access rights are, the workers often only had a vague idea of their rights. However, clear patterns and categories emerge in the workers’ experiences. There is an EU system to ensure the social insurance of mobile workers, which does not function well in practice, especially for the construction workers who need it most. This is primarily because many employers seek to undermine and evade social insurance contributions, but there are other contributing factors, such as complexity, difficulty finding information, and the reluctance of host country bureaucracies to help mobile workers. Theoretically, all posted workers should be covered by their country of residence, and many are, but not all. Workers moving to sending-country employment contracts should enter that country’s social protection system. The situation becomes quite complicated from the viewpoint of highly mobile workers, who often have dozens of contracts of varying length in various countries over the course of their careers.
The construction workers we talked to mainly did not describe themselves as engaging strategically with their resource environment, as predicted by the transnational social protection literature, except for some purchases of private insurance, and some highly skilled workers from less encompassing welfare states. Many demonstrably exercised market agency in shifting between jobs after not receiving that to which they believed themselves to be entitled. From our sample of interviewees, the most successful exercise of agency tended to come from those in the most powerful positions. Some of these come close to using a multi-faceted transnational social protection approach, using both market insurance, host country social insurance, as well as insurance from their employers and sending country. It is worth singling out intra-EU mobile third-country nationals as a separate category of workers, not only in terms of work precarity, but also as regards their total lack of social insurance access. Third-country nationals were the most precarious of the groups we researched, with little prospect of obtaining protection or using transnational social protection strategies. Their contacts in the host country, and in Europe generally, were limited. As a result, they were the most dependent on the employer.
In line with the transnational social protection literature, our workers’ narratives about their own strategies and rationalities frame access to social insurance as uncertain, contingent and contained in time and space, rather than as a permanent, encompassing and risk-mitigating presence. In other words, this is exactly the opposite of the function ‘insurance’ is intended to serve: it might or might not be available when needed. Highly mobile and temporary migrants realise they cannot rely on social insurance and therefore focus on minimising contributions, and exploiting all possible benefits when they are available. There is, however, no point taking risks or exerting effort to participate in benefit systems that will probably never provide benefits or security. From the welfare state perspective, the incentive is to try to ensure the collection of contributions from employers but to minimise the cost of welfare provision by minimising access for workers. There is no incentive to spend resources determining and facilitating access; in other words, it is better to erect barriers, or to leave them in place where they already exist. For employers, the incentive is to avoid payment of contributions and avoid enforcement authorities, while ensuring that workers understand as little as possible of what is going on. The result is that employers often fail to comply with the system and there is no one who is able and willing to force them to comply.
Overall, the existing pan-European system of reciprocal social insurance between Member States does not guarantee insurance for mobile construction workers who make a career of mobility. This problem is more serious, as we noted above, the weaker the labour market position of the worker. The system relies on employer compliance, and employers cannot be relied on to cooperate with the system, but rather seek to circumvent and undermine it. Bruzelius and Seeleib-Kaiser (2021) liken the European Union’s social insurance reciprocity system to that of a federal state in which social insurance reciprocity has not been completely established. Particularly in the EU context, where the goal of free movement is supported by social benefit provisions, this represents a policy failure. Social welfare systems cannot depend on the agency of relatively powerless beneficiaries, who are supposed to strategise and fight to obtain the benefits to which they are entitled, as this assumes capacities and agency which workers in difficulty may not possess. Support for workers who are at risk from work mobility should be transparent, credible and easily accessible, or it does not fulfil the function for which it is intended.