Institutions or resources and capabilities? Explaining engagement in European sectoral social dialogue

H1a: Coordination of domestic institutions will predict participation in SSDCs

Institutional theories posit that social partners in more coordinated regimes have greater capacity. This is related to the more significant role of social partners in coordinated systems. Owing to the need to achieve coordination, social partners play key roles in systems of collective bargaining and typically have higher levels of membership density. In more liberal systems, in which firms engage in competition and social partners tend to be weak, the situation is inverse. Though binaries between liberal and coordinated systems have eroded, countries such as Germany undergoing liberalization (Streeck, 2009; Thelen, 2014), these processes are underpinned by market-friendly patterns of social partner coordination which are rare in LMEs. Different levels of coordination thus continue to characterize national systems (see Table 1). This hypothesis overlaps with hypothesis 1b, reflecting rational institutionalism’s emphasis upon actor capacity, yet coordination is distinct from resources and capabilities.

OPEN IN VIEWER

Table 1.

Operationalizing coordination.

	Social partner density	Collective bargaining practices and density	Overall coordination1 = low2 = low intermediate3 = high intermediate4 = high
German hospitals sector	There is little information available on social partner density in the sector. It can be estimated that density rates are not as high as in hospital sectors in certain European countries.	The sector is split into three sub-sectors: public, private and church/charity. Collective bargaining coverage is high in the public sub-sector, but low in the private sub-sector.	2
German metal sector	Trade union density is approximately 50%. a There is no precise information on employers’ association density, but data from the sectoral employers’ association Gesamtmetall suggest that the figure is at a similar level to trade union density.	Though a sectoral agreement exists, there has been extensive decentralization of bargaining to firms in recent years. Coverage rates for the sector and its sub-sectors range from 30 to 65%. Coordination between regional social partners nonetheless continues to take place.	3
Italian hospitals sector	Employers’ association and trade union density rates are higher than levels in other Italian sectors, though not as high as in hospital sectors in certain European countries.	A mixture of sector and firm-level bargaining. Collective bargaining coverage is 100% (Eurofound, 2009).	3
Italian metal sector	Trade union density is approximately 35%. Density by employed workers varies between 47% of big companies and 15% of small companies.	Though a sectoral agreement exists, there has been extensive decentralization of bargaining to firms in recent years. Collective bargaining coverage is over 90% (Eurofound, 2018).	2
Polish hospitals sector	Little reliable information is available; reports suggest that employers’ association and trade union density is low, even if certain unions have some members within the sector.	Owing to low levels of social partner density, little collective bargaining takes place.	1
Polish metal sector	Trade union density is 6.9%. Employers’ association density is very low; one measure is 0.11%.	Bargaining takes place at firm level, where there is a trade union presence. Collective bargaining coverage is below 25% (Eurofound, 2018).	1
Swedish hospitals sector	Employers’ association density is close to 100%. Trade union density is also high, particularly when compared to rates in sectors in other European countries.	Sector-level bargaining, with some decentralization.	4
Swedish metal sector	Social partner density rates are very high. By one measure, employers’ association density is 82%. There are no exact data on trade union density, though the figure is estimated at 75–80%.	Sector-level bargaining, with limited decentralization. Collective bargaining coverage is over 90% (Eurofound, 2018).	4
UK hospitals sector	Trade union density was 39.3% in 2016; this is significantly higher than the UK average. The employers’ association, NHS Employers, has a coverage rate of approximately 100%.	Pay is set unilaterally, though there is sector-level bargaining on topics such as working time and training. Other topics are negotiated at single-employer (NHS trust) level. Collective bargaining coverage is very high.	3
UK metal sector	Trade union density is 29.1%. Employers’ association density is 12.3%.	Bargaining takes place at firm level, where there is a trade union presence. Collective bargaining coverage is below 25% (Eurofound, 2018).	1

a

Includes retired members.

Owing to discrepant potential benefits of European social dialogue, it may be that certain social partners are reticent (see hypotheses 2 and 3). Given that both unions in coordinated regimes and employers in liberal regimes may be more reluctant, effects related to incentives might cancel themselves out; such effects may also be found to be insignificant. We thus hypothesize that the superior capacity of social partners in more coordinated systems will lead to more significant participation. Table 1 operationalizes degrees of coordination in our cases. There are external data on institutional coordination, for example the ICTWSS index, yet there are no relevant sectoral data. We rank sectoral coordination from 1 to 4, taking into account social partner density and collective bargaining practices and density. Though social partner and collective bargaining density rates are important to scores, we are attentive to the extent to which practices ensure coordination.

H1b: Resources and capabilities will predict participation in SSDCs

Several accounts hypothesize that social partner strength is contingent upon access to resources which are subsequently used effectively (Lévesque and Murray, 2010; Seeliger, 2017). Studies of the dialogue emphasize the wider importance of power (Keune and Marginson, 2013), yet research shows that two resources are particularly germane: language and money (Pochet et al., 2009; Prosser and Perin, 2015). Owing to the predominance of certain languages in SSDCs, mainly English but also French and German, those social partners who do not speak these languages have difficulty participating in meetings. Money exerts a similar constraint. There is funding for SSDC participation – the European Commission finances up to 28 delegates from each side in plenary sessions, 15 in working groups and five in steering group meetings – yet certain social partners consider this insufficient.

It is difficult to locate external information on the language competence and financial resources of social partners. Though data exist on language competence within countries, there is no specific information on social partners. These data rather emerge in the course of fieldwork (see later section). Owing to the difficulty of compiling data on language competence and financial resources prior to fieldwork, as we did with degrees of coordination, we do not prepare a table operationalizing the resources and capabilities of social partners.

H2a: Employers in more liberal regimes will be more resistant to regulation than employers in more coordinated regimes

A key tenet of institutional theory is that liberal regimes obtain competitive advantage from the limited systems of social protection which exist in these countries. This equilibrium is underpinned by the attitude of employers. In coordinated regimes, the situation is inverse. Not only are these systems characterized by more generous systems of social protection, but employers are more committed to the maintenance of such regulation. Despite the erosion of binaries between liberal and coordinated systems, recent accounts note continued differences in employer preferences (Bulfone and Afonso, 2020; Wright et al., 2016).

Owing to these differences in attitudes, we hypothesize that employers in more liberal regimes will be more resistant to SSDC regulation; such measures potentially undermine national competitive advantages. Employers in more coordinated regimes, who are accustomed to elevated conditions in domestic contexts and may wish to combat social dumping, are likely to be more relaxed. We understand ‘regulation’ as all SSDC output, though especially legally-binding measures; employers have long been opposed to this output (Keller, 2003).

H2b: Resources and capabilities will predict employer resistance to regulation

As asserted in H1b, literature indicates that the resources and capabilities of social partners significantly influence behaviour (Lévesque and Murray, 2010; Pochet et al., 2009; Prosser and Perin, 2015). There is reason to believe that there will be an effect on employer resistance to regulation. Owing to prerequisite needs for money and language competence, employers who lack these resources and capabilities will be unable to express preferences. As money and language competence increases, employers will likely become better at articulating resistance. Though some equate employer non-engagement in SSDCs with resistance to output, this is problematic. Traditionally, employer non-engagement is associated with the imposition of regulation by the European Commission.

H3a: Unions in more liberal regimes will be more desirous of regulation than unions in more coordinated regimes

As outlined in H2a, liberal regimes are characterized by lower social standards than coordinated regimes. Notwithstanding the basis of the competitive advantage of liberal systems, trade unions contest poor working conditions in domestic settings. At European level, union rationales are debated. Owing to the potential to achieve competitive advantage, it has been alleged that unions in certain liberal systems, particularly those from Central and East European (CEE) countries, adopt ambiguous attitudes to European regulation. Despite these concerns, literature suggests that unions in liberal regimes generally support European regulation; within CEE countries, this has been promoted by recent economic development (Prosser, 2018).

Research indicates that a more salient influence is the extent to which European regulation contributes to domestic policy contexts. In liberal systems, owing to lower domestic levels of regulation, unions are likely to attach greater importance to European regulation. Though unions in coordinated systems support such regulation, the better quality of national standards means that this is likely to be a lower priority. Despite the erosion of differences between liberal and coordinated systems, recent accounts note continued distinctions in union preferences (Wright et al., 2016).

H3b: Resources and capabilities will predict union desire for regulation

As argued in H2b, there is reason to think that the resources and capabilities of social partners will exercise greater influence on preferences. We consequently hypothesize that resources and capabilities will predict union desire for regulation.

Explaining disparate social partner participation

Participation among national social partners differed substantially. Among German, Swedish and UK social partners, respondents affirmed that they regularly attended and contributed to SSDC meetings. This is confirmed by other studies (Eurofound, 2018; Prosser and Perin, 2015) and our observations of SSDC meetings; the engagement of social partners from these countries is comparatively high. This was the case in both the hospitals and metal sectors. The extent to which participation in SSDCs was a priority for these organizations is a different question. Though a representative of a hospitals employers’ association stated that the European level was becoming more important for the sector, an official from a metal sector union asserted that participation in SSDCs was increasingly low, indicating limited interest in SSDCs among national social partners. Limited salience of SSDC output arguably meant that European social dialogue imperfectly reflected national balances of power (see below).

There was weak participation from Italy. In the metal sector, trade unions tended not to directly participate in the SSDC, a respondent asserting that participation was often contingent upon discussion of themes which were consistent with expertise of an available, English-speaking official. Despite this lack of direct participation, unions remained closely informed of the work of the SSDC, receiving updates and relevant documentation from Industriall. The metal employers’ association was more active in the SSDC, attending meetings and engaging in preparatory work. Similar patterns were evident in the hospital sector. Although Italian trade unions participated in SSDC meetings sporadically, engagement with EPSU (the European health trade union confederation) was more sustained. The employers’ association was more engaged in the SSDC; a respondent asserted that the organization was, amongst Southern European countries, the most active in HOSPEEM (the European health employers’ association).

Degrees of participation in Poland were very insubstantial; social partners from the country attended neither the metal nor hospital SSDC. This was partly attributable to the non-existence of employers’ associations in the metal and hospital sectors – in most Polish sectors no such organizations exist – though unions also failed to participate. Polish unions had once participated in the hospitals SSDC, yet this was no longer the case. PolHosTu1 had even ceased to be members of EPSU (see below).

These trends were affirmed by other interviews, who identified three distinct regional groups: Northern European social partners, who participated regularly; Southern European social partners, who participated more sporadically; and CEE social partners, who participated to a low degree. An official from a metal sector employers’ association affirmed that underdeveloped social dialogue structures and poor resources limited CEE social partner participation.

Issues related to resources underpinned many of the problems associated with participation. Though the extent of their own resources was not often referred to by Germany, Sweden and UK social partners, it is reasonable to conclude that comparatively abundant resources underpinned better participation; the superior resources enjoyed by these organizations is confirmed by studies (Prosser and Perin, 2015) and various interviewees spoke of the advantage of Northern European social partners. The case of UK trade unions is interesting. Despite domestic marginalization, comparatively strong finances and fluency in English meant that UK unions enjoyed a stronger position at European level.

Italian and Polish cases illustrate the problems faced by social partners with few resources. A key problem was finance. Owing to costs associated with travelling to SSDC meetings, social partners had difficulties meeting this expense. An Italian metalworking union raised this issue, though the problem was cited more often in the Italian hospitals sector: three social partner interviewees identified costs as a barrier to participation. Concerns about finance were more considerable in Poland. An interviewee from a Polish metal sector union asserted that although his union attended European meetings in which travel costs were funded, lack of funding precluded participation in SSDC meetings. The EPSU membership fee even meant that PolHosTu1 were no longer an affiliate,

This membership is simply too expensive for us . . . The union movement in the health sector is in decline . . . so we have less money. The funding we have is spent on more basic activities.

Aside from the issue of finance, a key impediment was language. The problem of competence in relevant languages was more prevalent in Italy and Poland. In the Italian metal sector, a trade union official asserted that participation in SSDC meetings was contingent upon the availability of English speakers. In the Italian hospitals sector, both sides of industry identified language competence as a problem; while respondents had some knowledge of English, this was not sufficient to make their participation as active as they would have liked (e.g. engagement in ongoing discussions). The problem was more considerable in Poland. Officials in Polish unions are older and tend not to speak the languages used in SSDCs; none of our interviewees was able to speak English. A representative from a metal sector union summarized this problem:

The main obstacle we face here is the language. Unfamiliarity with foreign languages, English mostly, stops us from participating in many of the meetings. If there is no translation to Polish, we simply cannot understand anything.

Language emerged as crucial. In Northern Europe, the capacity of most social partner officials to speak English facilitated participation. The ability to speak English as a mother tongue was a particular advantage for UK social partners; a representative from a UK metal sector employers’ association asserted that the organization’s command of English meant that it drafted documents for CEEMET (European Tech & Industry Employers organization). Social partners in Germany and Sweden also had an abundance of officials who spoke fluent English, particularly in the latter country. Though less important than language, the variable of location was significant. Given that SSDC meetings take place in Brussels, attendance is particularly easy for Belgian social partners. Certain national organizations, including IG Metall, Gesamtmetall and the NHS, also have offices in Brussels.

What are the preferences of social partners?

Nordic social partners have long advocated non-legally binding European social dialogue. This has been motivated by the desire to defend social partner autonomy within domestic systems of industrial relations. The preferences of Swedish social partners continue to reflect such rationales; representatives of hospitals and metal sector employers’ associations asserted that flexible European regulations allowed the Swedish model to be defended. Swedish employers also expressed wider preference for non-legally binding regulation. Interviewees from metal and hospitals sector employers’ associations asserted that the dialogue should take a flexible form, avoiding rigid standards. There was particular opposition to hard regulation and the coordination of wages, which threaten collective bargaining prerogatives, but less opposition to health and security issues, which are regulated by the Swedish state. Despite this preference for flexibility, aggressive opposition to dialogue was not articulated and conciliatory attitudes were sometimes expressed; a representative of a metal sector employers’ association stated that it was important to find topics which trade unions find interesting.

Swedish trade unions also expressed conciliatory attitudes about the dialogue. Though unions had a greater appetite for a harder dialogue than employers, particularly in the hospitals sector, calls for a more robust social dialogue were muted. Union interviewees in both sectors defined effective dialogue as involving joint learning and understanding, omitting more robust benchmarks; scepticism about the evaluation of concrete outcomes was even expressed. A respondent in the metal sector emphasized the greater importance of achieving understanding and respect:

[We established principles such as]: ‘okay, we unions understand that companies have a need for flexibility. In some way, we have to achieve that. But in return, employers must accept that consequences for employees must be as limited as possible, so that it becomes bearable.’ That was about as far as we came. And it may seem very banal, from a Swedish point of view, that: ‘okay, we can sit and talk about these things’. But I saw that it was a significant step for others to be able to acknowledge any kind of reasonableness in the position of opponents.

German employers espoused mixed views about the dialogue. Although an interviewee from a metal sector employers’ association emphasized the benefits of sharing good practice, there was opposition to a more robust dialogue. Not only were limits on the mandates of social partners called to attention, but the danger of Commission interference in the prerogatives of national social partners was emphasized. A representative from a hospitals sector employers’ association was more conciliatory, stressing benefits of the dialogue; it was asserted that the SSDC addressed practical topics and worked effectively.

The views of German trade unions were also mixed. In the hospitals sector, union interviewees were complimentary about the role of the Framework Agreement on Prevention from Sharp Injuries. This agreement, concluded in 2009, was considered to have improved conditions within the sector. A representative from a hospital sector union was also optimistic about the broader role of the dialogue:

When you take into account the amount of time which SSDC takes, then you see that roughly one topic per year is addressed and results are reached. And that is amazing given the setting; there are many countries, many employers and many different interests . . . Health and safety in Germany has been improved through the needlestick directive. From a trade union point of view, SSDCs are important.

Such optimism about the dialogue nonetheless fell short of advocating output on ‘harder’ topics – this is a traditional demand of more radical unions. In the metal sector, union respondents were less optimistic about the record of the dialogue, expressing scepticism about the value of the metal SSDC. Interviewees wished for the European Commission to assume a more proactive role and encourage the conclusion of sectoral agreements. It was asserted that digitalization would be an ideal topic for such an agreement.

Despite traditionally adversarial relations between UK employers and unions and contrasting attitudes to European regulation – employers have considered such regulation restrictive whilst unions regard it as a key source of substantive rights – fieldwork yielded modest evidence of these trends. Both sides of industry were somewhat relaxed about SSDC output. No union respondent mentioned wages as an issue which should be addressed by SSDCs, as occurred in Poland (see below), and a representative from a health sector union asserted that the outcome of ‘best practice’ should be targeted. A union interviewee in the hospitals sector also reflected that ‘good practice and joint guidance and joint statements is probably as good as it can get’, opining that SSDCs were more important for countries with lower social standards.

Trust in domestic employers was also expressed by unions. A hospitals sector interviewee ‘trust[ed] the employer side representing the UK, for representing the views of the sector accurately and not just taking a hard line to defend the employers’ position regardless’. Employers echoed such sentiments. A respondent from the hospitals sector employers’ association emphasized benefits of exchange of good practice with unions, whilst a representative of the metal sector employers’ association asserted that:

80% of the time we are in violent agreement with trade unions on most things at the EU as well national level. We agree on far more than we disagree.

This is inconsistent with the adversarial tradition in the UK metal sector, yet there were areas in which traditional disagreements between social partners emerged. A representative from a metal sector union asserted the union’s preference for agreements, adding that the capacity of topics to achieve a direct impact at workplace level was very important. The interviewee also compared UK employers unfavourably with European employers. UK metal sector employers expressed opposition to a more interventionist social dialogue:

There is no initiative issued by the Commission which employers would have initiated, they all limit competitiveness. Initiatives don’t take account of the needs of the metal sector . . . someone has to pay for [extra regulation] and it is either the company or member states through increased taxation. The EU doesn’t respect the subsidiarity principle. We need a single set of rules for everyone to play by, but within that set of rules there has to be freedom to compete. However the direction of the EU is to provide too little room to compete. If everyone is doing the same, how is a company in Germany competing with one in Italy?

Employers in the Italian metal sector articulated a tough line; an official stated that the organization wished to ‘contain any legal regulation from the European level . . . [and] avoid a third level of collective bargaining’. Associated with this was the goal of a dialogue in which autonomous positions were achieved by social partners; the Silica Directive was regarded as a success, given that it was produced by social dialogue rather than the direct intervention of the European Commission. The association expressed preference for topics which were ‘transversal and basic for all’, contending that this improved engagement in SSDCs; it was also asserted that member firms were more likely to participate if they had interest in a specific topic. Employers in the Italian hospitals sector were more conciliatory. Rather than outlining specific red lines, an interviewee expressed preference for a ‘negotiating culture’. Referring to the agreement on prevention of sharp injuries, it was asserted that such an attitude had allowed Italian employers to shape the agreement in line with their preferences.

Italian unions desired legally binding regulation and negotiations on topics which were central to the employment relationship. In the metal sector, a respondent asserted the commitment of the union to a European minimum wage and basic income, reflecting historic demands of Italian unions for European regulation and solidarity. Another union official repeated this demand for dialogue on ‘harder’ topics, suggesting discussion of wages in European companies. Metal sector unions also preferred ‘harder’ regulatory tools, yet this demand was particularly articulated in the hospitals sector. A representative of a hospitals sector union contended that directives, as opposed to non-legally binding texts, had potential to ‘reach workers [and] have legitimacy for policymakers’. It was also asserted that directives were particularly appropriate for topics such as prevention of sharp injuries and working time, owing to their legally enforceable nature.

Italian unions were divided on the record of SSDCs. Hospitals sector unions expressed preference for legally-binding output, yet it was contended that non-binding dialogue helped diffuse best practice and reinforce arguments in national-level negotiations. Metal sector unions were more critical. Though it was acknowledged that dialogue per se could be useful, respondents emphasized the need for output which went beyond non-binding declarations. Officials asserted that coordination in areas such as industrial development and environmental policy would make the dialogue more effective.

Notwithstanding problems faced by Polish unions in participating in SSDCs, compounded by the absence of employers’ associations in the Polish hospitals and metal sectors, Polish unions expressed preferences regarding European social dialogue. In the metal sector, a union official asserted that the most important issue was the development of the industry, reflecting longstanding domestic concerns. In the health sector, a Polish representative of a union in the process of applying for EPSU membership asserted that there was a need to achieve uniform sectoral standards throughout the EU. Similar views were advanced by another health sector union, the interviewee adding that this might stem pressure exerted by emigration on the Polish health sector.

A representative from a third health sector trade union expressed frustration with the output of SSDCs, affirming that SSDCs neglected core topics in favour of less important issues:

The problems which used to be raised at these meetings simply did not reflect the current situation in Poland . . . We face more serious problems than those . . . Work safety tends to be in the last place for us. Our priority is the employment type, the wages or the status of a hospital . . . I was a chairman when Poland was entering the EU. I believed that our accession would help us get better standards in terms of employment and wages . . . This has not yet happened.

This position may also be associated with weak Polish participation in SSDCs. Owing to dubious perceived benefits of participation, it can be argued that Polish social partners do not regard engagement with the dialogue as a priority; this issue is explored further in the conclusion.

H1a: Coordination of domestic institutions will predict participation in SSDCs

Evidence for this hypothesis was unconvincing. In some of our cases, there was an apparent association between coordination and participation. In Poland and Sweden, there was close correlation between these variables. The cases of Germany, Italy and the UK nonetheless suggest that the relationship between coordination and participation is contingent. Despite levels of coordination differing between domestic hospitals and metal sectors, rates of sectoral participation were similar within the three countries. This suggests that the relationship between coordination and participation may be collinear with a variable which is peculiar to countries.

H1b: Resources and capabilities will predict participation in SSDCs

There was strong evidence for this hypothesis. In all cases, money and language competence promoted participation in SSDCs. Not only did this explain high participation (German, Swedish and UK hospitals and metal sectors), but also intermediate (Italian hospitals and metal sectors) and low participation (Polish hospitals and metal sectors). The presence of these conditions also promoted participation in sectors in which levels of coordination were apparently unfavourable, for example the UK metal sector. This suggests that the variable of resources and capabilities is more significant than institutional coordination.

H2a: Employers in more liberal regimes will be more resistant to regulation than employers in more coordinated regimes

There was some evidence for this hypothesis. Swedish cases were broadly confirmatory; though employers were opposed to hard regulation and the coordination of wages, there was less opposition to health and security issues and conciliatory attitudes were sometimes expressed. In Italy and the UK, metal sector employers were more resistant to regulation than counterparts in the hospitals sector; the existence of differences within countries, in which control variables are similar, is potentially telling. German cases, in which metal sector employers were more resistant to regulation than less coordinated counterparts in the hospitals sector, nonetheless disprove such an inference.

H2b: Resources and capabilities will predict employer resistance to regulation

This hypothesis explained the basic capacity of employers to articulate resistance to regulation. In the Polish case, the non-existence of employers’ associations meant that employers were unable to express resistance. Beyond this rudimentary relationship, no association was found. Certain employers with abundant resources and capabilities, such as those in the Swedish hospitals and metal sectors, adopted a relaxed attitude towards SSDC output.

H3a: Unions in more liberal regimes will be more desirous of regulation than unions in more coordinated regimes

There was some evidence for this hypothesis. Among unions in the Polish hospitals and metal sectors, strong preference was expressed for social dialogue on more meaningful topics, including wages. This can be linked with less developed labour market conditions within Poland, which, although associated with the developing profile of the economy, are also related to liberal environments in these sectors. An inverse effect was observed in Sweden: in both the hospitals and metal sectors, little desire was expressed for a more robust dialogue and certain union interviewees argued for the use of softer tools.

This result can be linked to more coordinated systems in these sectors, though is also associated with a desire to defend the voluntarist form of Swedish industrial relations. In the UK, evidence for the hypothesis emerged: unions were more desirous of hard output in the less coordinated metal sector. German cases were nonetheless inconsistent with the hypothesis. Despite the fact that coordination is superior in the metal sector, unions were more desirous of hard output. This may reflect the weaker record of the metal SSDC, a potentially collinear variable. Italian cases were inconclusive. Though unions in the less coordinated metal sector desired harder output, this demand was replicated in the more coordinated hospitals sector.

H3b: Resources and capabilities will predict union desire for regulation

There was little evidence for this hypothesis. Despite poorer resources and capabilities, Polish and Italian unions expressed strong preference for social dialogue on more meaningful topics, including wages. As resources and capabilities increased, desire for regulation tended not to increase; in the Swedish hospitals and metal sectors, unions expressed little desire for more robust dialogue.

The results of the six hypotheses are summarized in Table 3.

OPEN IN VIEWER

Table 3.

Hypothesis results.

HP		Institutional theory (a)	Resources and capabilities theory (b)
(1)	Participation	Coordination of domestic institutions will predict participation in SSDCs	Resources and capabilities will predict participation in SSDCs
	Results:	(+) PL and SE (−) DE, IT, UK	(+) all (Explains high, intermediate and low participation)
(2)	Employer resistance to regulation	Employers in more liberal regimes will be more resistant to regulation than employers in more coordinated regimes	Resources and capabilities will predict employer resistance to regulation
	Results:	(−) SE (+) IT and UK (M) (+) DE mildly resistant	(+) PL
(3)	Union desire for European regulation	Unions in more liberal regimes will be more desirous of regulation than unions in more coordinated regimes	Resources and capabilities will predict union desire for regulation
	Results:	(+) PL (H and M) (+) IT (M and H) (+) UK (M) (−) DE (M) (−) SE	(−) all

(+) / (−) positive or negative predictability of the theory.