The role of productivity coalitions in building a ‘high road’ competitive strategy: The case of Denmark and Ireland

Non-market coordination mechanisms

Productivity coalitions necessarily engage in non-market collaboration, and those that can draw on existing institutions to aid this process will be at an advantage over those that have to develop new institutional mechanisms. Coordinated market economies (CMEs) have a comparative advantage in this regard, in that an extensive array of institutional mechanisms can facilitate non-market coordination between the actors. Moreover, complementarities arise from the interaction between various institutions (Hall and Soskice, 2001). In CMEs, existing non-market coordination mechanisms at various levels of the economy provide institutional supports for productivity coalitions and enable them to be more effective in achieving productivity innovations or addressing collective action problems.

The work of Streeck (1992, 1997) has been particularly influential in identifying the comparative advantage of CMEs in the area of training. In LMEs there is a disincentive to investment in training. Firms that do not train incur lower costs and therefore can pay higher wages to ‘poach’ trained employees; the rational response for employers is therefore not to invest in training but to ‘free ride’. The end result is insufficient training within the economy, which is then driven down the low-cost, low-skill and low-productivity route to competitiveness, creating what Finegold and Soskice (1988) termed a ‘low-skill equilibrium’. This prisoner’s dilemma is largely overcome in CMEs through the interaction of employers’ associations, wage bargaining and training systems. These provide a range of incentives and obligations to invest in training, and thus to compete on a high-wage, high-skill and high-productivity basis. For example, where unions achieve higher wages through collective bargaining, this gives employers an incentive to raise productivity through investing in their workforce. Institutionalized inter-firm linkages facilitate collaboration in the provision of training so that the risks of poaching are minimized. The end result is that the economy is driven along the ‘high road’ to competitiveness.

Because LMEs lack many of these institutions, productivity coalitions will require new institutional mechanisms to support collaborative behaviour. However, according to Hall and Soskice (2001), each institutional configuration has its own institutional logic, and new institutions which complement existing types of coordination will be more enduring. This suggests that new non-market coordination mechanisms in LMEs may be inherently unstable.

Union ‘embeddedness’

If productivity coalitions are not to become simply a mechanism for implementing an employer-driven agenda of efficiency gains and market-conforming reforms, they require well organized unions embedded at different levels of the economy: national, sectoral and workplace. They also require some semblance of power symmetry between the actors so that unions can assert their right to be involved in productivity coalitions and play a functioning role within the various coordination mechanisms, and hence effectively represent workers’ distinctive interests. How the productivity agenda is defined and who benefits will be largely determined by the power balance between the actors. Even potentially positive-sum issues like training can involve a range of tensions (Stuart, 2007), such as the extent of resource allocation or which employees receive training. Stronger unions will bargain successfully for access to training for their members (Dieckhoff et al., 2007), whereas weaker unions will be less effective at pressing employers to commit resources. If unions enter productivity coalitions because they have few other means of exercising influence, their ability to shape the agenda in ways which advance a balance between equity and efficiency will be limited. Productivity coalitions may imply unions ‘dancing’ with employers, but their effectiveness may also require the capacity for ‘boxing’ (Huzzard et al., 2005).

Ideological concordance

The state of capital-labour relations is the third determinant of the effectiveness of productivity coalitions. Katzenstein (1985: 32), in discussing successful corporatism, refers to an ‘ideology of social partnership’ and a ‘culture of compromise’. Unions and employers do not put aside their own aims, but there is a shared understanding of a wider, collective or public interest which moderates class conflict. Social dialogue facilitates mutual dependence, credibility of commitment and trust between the social partners (Visser, 2001), and non-market coordination mechanisms provide the space in which social dialogue can occur and consensus develop over time. Where there exists a high degree of trust and consensus, with an acknowledgement of the legitimacy of the roles that other actors play, productivity coalitions are likely to be stronger, more stable and more effective. Where it is necessary to develop new coordination mechanisms, building cooperation will be difficult if there is no history of trust (Culpepper, 2001). Where a strong anti-union ideology is pervasive, employers are likely to limit union influence within productivity coalitions or resist union involvement altogether.

Institutional constraints

Underpinning the success of many of the micro-foundational arrangements in CMEs are regulations and institutions that place constraints on the behaviour of firms. As Streeck (1997) argues, employers will not freely choose a path that appears to go against their self-interest. For example, a system that limits managerial prerogative will not be chosen voluntarily even though it may lead to more efficient outcomes. Similarly, he notes that if employers have the option to adopt a low-cost route to profitability, self-interest will guide them in this direction. However, by restraining them from taking what may appear to be rational action, society in fact transforms their preferences by opening up alternative strategies which they would not otherwise have considered. Streeck (2004) calls these ‘beneficial constraints’: society and the economy can benefit from placing constraints on the strategic options available, and in doing so prevent employers from wasting valuable resources on sub-optimal strategies. In this way, society can ‘educate capitalists’.

However, in a more liberal national context, employers can veto any attempts by the state to institute such constraints through the threat of relocation (Streeck, 2004). Given that the benefits of non-market coordination in addressing issues of market failure seem attractive to policy-makers, there has been some interest in voluntarist approaches to non-market coordination. Culpepper (2001), for example, in examining the issue of employer-sponsored training, argues that persuasion of actors by government, using incentive-based public policy supports such as subsidies, can be effective in bringing about collaborative employer behaviour. Similarly Teague and Donaghey (2004), writing in the Irish context, are equally positive about a voluntarist approach, arguing that innovative practices in areas such as partnership, networking and training can be encouraged and supported by public policy. In this way, consensus and cooperation will develop in a voluntary and experimental manner, and new ways of operating will be ‘owned’ rather than ‘imposed’.

As might be expected, there is significant debate about how effective such approaches can be at changing the actors’ behaviour. The logic underpinning a voluntarist approach is that rational argument and empirical evidence can convince employers of the economic benefits of a specific course of action. However, this ignores short-term market pressures. In a voluntary approach, firms that enter into productivity coalitions and adopt longer-term strategies based on investment in their workforce will incur higher costs than competitors who retain a low-cost approach. In an environment where there is an emphasis on short-term profitability, they would find this ‘high road’ difficult to sustain, and defect from agreements. Without constraints, productivity coalitions will be unstable as they cannot protect firms from market pressures (Bélanger and Edwards, 2007).

The voluntarist approach to coordination also ignores ideology. Some actors may simply not wish to participate in collaborative institutional arrangements for ideological reasons (Visser, 2001). For example, if employers believe that unions limit managerial prerogative and are thus bad for flexibility, it seems implausible that the majority could be cajoled into forming productivity coalitions with unions.

Danish productivity coalitions: A coordinated and consensual approach to national competitiveness

Denmark’s strong economic performance and high GDP per capita over much of the last 25 years is surprising, since its industrial structure is not dominated by large and successful high-tech firms. While it has some large companies and clusters of small successful companies in high-tech industries, its predominant economic base is smaller firms within medium and low-tech sectors, and its investment in research and development is comparatively modest. However, what makes these sectors internationally competitive is the adaptability and flexibility of Danish firms, and one of the key factors here is the capabilities of the workforce (Kristensen, 2006; Nielsen and Kesting, 2003).

This, too, is surprising in that there are two strong incentives for Danish employers not to invest in their employees. The first is the high mobility of workers as a result of the ‘flexicurity’ labour market system, with approximately one third of the workforce changing jobs each year (Aagaard et al., 2004). The second is the large number of small firms that dominate the Danish economy, since smaller organizations have less capacity for in-house training and incur greater disruption to their operations if they send staff to off-the-job training. These disincentives have been overcome through strong productivity coalitions.

The first characteristic of these productivity coalitions is the array of non-market coordination mechanisms. Denmark has a tripartite system with the social partners involved at national, industry and local levels in the development and review of training policy, colleges and courses. Their close involvement ensures that the training offered is reviewed and modernized to meet the changing needs of industry, and that there is feedback between delivery, design and policy formulation. This creates confidence in the certification of training, which in turn facilitates worker mobility and labour market flexibility (Egger and Sengenberger, 2003).

Social partner involvement also results in complementarities with the wage bargaining system, overcoming many of the potential prisoner’s dilemmas. Through the sectoral collective agreements, unions have negotiated annual leave entitlements of up to two weeks per year for employees to undertake training of their choice relevant to their career development. They have also bargained for several employer training levies, to cover the cost of reviewing and developing training courses, and more recently to address the lower uptake of training among low-skilled workers. This ensures that the costs and risks associated with training are shared among all employers covered by the collective agreements, and the potential issues of poaching are largely overcome. In some agreements the certification system has been embedded in the sectoral pay scales. This now happens less frequently with increasing wage decentralization, but where it does still occur, wage outcomes can be closely aligned with training outcomes. This provides an added incentive for employees to undertake training.

The Danish training system provides other complementarities in driving the international competitive success of Danish firms. One of these relates to the organization of work. Kristensen (2006) argues that as a result of high levels of training, Danish workers have come to expect work to be challenging, otherwise they will look for alternative employment. This forces employers to create jobs that utilize skills and provide worker autonomy, and it largely precludes cost-minimization strategies and repetitive and monotonous work processes. Several interviewees also pointed to the ‘shock effect’ of comparatively high wages, which force employers to compete on a high productivity basis.

Union ‘embeddedness’ plays a fundamental role in the effectiveness of Danish productivity coalitions. The high rates of membership density (67%) and collective bargaining coverage (71%) (Jørgensen, 2011) provide unions with a significant presence at national, sectoral and workplace levels and strong bargaining power. Their close involvement with the training system enables them to take a strategic view and to signal well in advance to employers’ associations their intention to bring training issues to the bargaining table. A good example is the levy negotiated in the 2007 bargaining round, which was aimed at unskilled and low-paid workers who are most likely to lose their jobs as a result of low-cost international competition and who therefore benefit particularly from up-skilling. Unions were concerned that while the state pays a proportion of the wages of employees off work for training, not all employers cover the difference, particularly in the case of unskilled and low-paid workers. This acted as a disincentive for such workers to utilize their training leave entitlement, and also allowed employers to manipulate the training leave system. As a union official representing sectors such as retail, hospitality, cleaning and agriculture reported, training in the past has often been limited to potential managerial staff or to employees undertaking training chosen by the employer rather than by the employee. Research in the retail sector revealed similar concerns about the manipulation of the training system by some employers (Esbjerg et al., 2008). This new levy covering the shortfall in wages is aimed at overcoming some of these issues.

Union officials report that these negotiations were not driven by any pressure for increased training entitlements from grassroots members, and indeed they have to convince many of their members that training is in their best interests. Thus shop stewards also have an important role to play at workplace level in encouraging workers to take up their entitlements to training leave. Additionally, with the increasing decentralization of wage determination, they also negotiate at workplace level for higher wages for employees who have undertaken training.

A key element of the ‘embeddedness’ of unions is their capacity to engage in ‘boxing’ when necessary and to counteract the short-term cost pressures that employers face. While employers recognize the benefits of Denmark’s highly skilled workforce, when it comes to negotiating collective agreements cost factors often take precedence, and unions report having to negotiate hard for any advances in the training agenda, despite the obvious benefits. As one union official noted, the new training levy ‘is driven by the unions, and this is actually good for the Danish economy and good for Danish firms. . . . But employers wouldn’t choose that path if we weren’t pushing them down it.’

Central to the success of these productivity coalitions are high levels of ideological concordance. There is shared discourse between the parties about the legitimacy of their respective roles as well as their differing interests. A consensus-based relationship exists between capital and labour, underpinned by a culture of dialogue and cooperation. Consequently, they refer to a ‘consensus principle’ that underpins the development of labour market policy (Due and Madsen, 2008). The positive attitude of employers towards trade unions at the workplace (Larsen and Navrbjerg, 2010) stands in sharp contrast to the anti-union stance generally found in LMEs. An official of one employers’ organization neatly sums up what many research participants noted:

We have a lot of aspects where we are not agreed of course, because we represent opposite opinions, but I think it is a professional attitude towards each other. We need the union and they need us. . . . You could say we have a common success criterion. . . a competitive Danish economy.

Another employers’ official described the relationship as one where ‘you have to live with them – it’s like a marriage’. He noted that this implied trust, commitment, honesty and fairness, but also disagreements and the potential for divorce if the relationship stopped delivering for one side or other. So, while the Danish productivity coalitions are supported by a high degree of consensus, achieving this consensus is often difficult and the actors will resort to ‘hard-nosed bargaining’. However, in such instances, it is the ideological concordance about the legitimate roles the actors play, and an appreciation of the benefits the system delivers for Danish competitiveness, which prevents the relationship from breaking down. Additionally, the various institutions provide the space to develop deliberation and consensus-building and for conflicts to be resolved.

While a high degree of consensus and cooperation is characteristic of the Danish productivity coalitions, their effectiveness also relies on various institutional constraints. These operate primarily through the collective agreements, with comparatively few laws regulating the employment relationship. For employers, membership of an employers’ association includes an obligation to abide by the terms of the relevant collective agreement, including any training provisions. One union official noted that associations representing small firms, in particular, would opt for a completely decentralized training system if given the opportunity, because of the short-term cost savings. However, because these associations are in turn members of larger associations, they are constrained by the collective agreements.

For employers who choose not to enter into collective agreements, the legality of secondary strike action acts as a constraint on them. Danish employers’ associations are not opposed to such industrial action, as they cannot otherwise deal with firms outside the collective agreements who gain a competitive advantage through lower labour costs. Some union officials reported tip-offs from employers about competitors operating outside the system, in the hope that the union would take industrial action to bring them on board. For this to be effective, it also requires strong unions prepared to take industrial action. One union that had done so reported that while relations with this particular employer were initially adversarial, within the space of a year they had developed very positive relations. Their interpretation was that by being forced into a collective agreement, the employer had come to appreciate the benefits of working with the union.

The state also plays an important role by creating the training system and mandating the involvement of the actors, thereby legitimizing the union role.

Irish productivity coalitions: Voluntarist coordination in a liberal market economy

The phenomenal economic growth of the ‘Celtic Tiger’ period was widely acclaimed by many commentators, and the sudden reversal in Ireland’s economic fortunes has been equally widely publicized. However, even after the economic crash, the country has one of highest levels of GDP per capita in the EU, alongside Denmark. Thus at first sight it may seem surprising that a key area of concern for the social partners from the mid-1990s onwards was whether enough was being done to build a sustainable high-wage, high-productivity economy. The explanation is that much of Ireland’s economic success was built on exceptionally high-levels of foreign direct investment (FDI). In manufacturing, foreign-owned firms account for 84 percent of the value of output while employing only 47 percent of the workforce (Forfás, 2011). The National Development Plan (NDP, 2001: 39) noted in 2001 that ‘a dual manufacturing structure [had] emerged with high-tech, high value-added industries dominated by foreign ownership, and low-tech, low productivity industry being largely of Irish ownership’. Tansey (2005) pointed to productivity rates in the leading manufacturing sub-sectors dominated by foreign-owned multinationals, six times higher than other parts of the Irish manufacturing industry. Yet the vision of the social partners was a long-term strategy that also included highly productive indigenous firms.

Given this segmentation in productivity levels, and given Ireland’s below-average performance on CVT, national strategy documents placed significant emphasis on developing a highly skilled workforce, enhancing employee engagement and implementing high-performance HRM practices (alongside other important drivers of productivity growth). However, despite agreement on the strategic aims there was little agreement about the institutional supports needed to bring these about. For example, the National Skills Strategy (EGFSN, 2007) included many training outcome targets, but few details about how these were to be achieved. The need for collaboration between the actors was acknowledged: as the Report of the High Level Group on Manufacturing pointed out (2008: ii) ‘a coordinated and collaborative approach [was] absolutely necessary’ between employers, employees, unions and government if the vision of a highly productive manufacturing industry was to be realized. What form this collaboration was to take was not made clear.

Developing the interlocking coordination mechanisms characteristic of CMEs was not on the agenda. Indeed Teague and Donaghey (2009) saw this as a positive feature of Irish social partnership, as to attempt this would run the risk of frightening off FDI. The extent to which such institutional mechanisms would have actually deterred multinational companies from investing is not clear – many of them operate in other European countries where such mechanisms exist – but the threat of relocation was enough to deter the Irish government. As a result, productivity coalitions were formed on a purely voluntarist basis. For example, significant state resources were invested in promoting workplace partnership. In the area of CVT, Skillnets was created, a programme where government grants are provided to support the creation of inter-firm networks to coordinate training. The failure of workplace partnership to develop in Ireland has been well documented (Roche and Teague, 2013), but there has been no analysis of attempts to build productivity coalitions at sectoral level. The analysis which follows fills this void by focusing on Skillnets.

Skillnets is a voluntary form of non-market coordination between employers to increase employer-led training in the Irish economy. It began in 1999 and since that time has funded over 350 networks involving more than 60,000 enterprises and 275,000 employees (out of a workforce of 2.1 million). Employers form inter-firm training networks and then apply to Skillnets for government funding to cover part of the cost of training delivery. Funding is available for a two-year period and networks may then re-apply. The large majority of participants have been smaller firms, which as noted above are less likely to train. Employers’ associations and government officials interviewed described this as making an important contribution to addressing Ireland’s skills shortage.

Skillnets is an innovative approach to building non-market coordination in a liberal market context. New coordination mechanisms take time to develop and the scheme was still in its infancy when the financial crisis hit Ireland. State funding was cut by two-thirds, reducing the number of functioning training networks from more than 120 in 2010 to fewer than 60 in early 2012. In any event, there are a number of limitations to Skillnets and even under favourable economic conditions it is doubtful whether it could fully address the prisoner’s dilemma facing organizations. The logic of state-subsidized coordination is to entice firms to experiment with collaboration, thus reducing their uncertainty about its potential benefits (Culpepper, 2001). Hence the hope with Skillnets was that initial government funding would lead to self-financing networks; but many collapsed when public funding ended. This was because employers outside the network incurred lower costs and could poach trained staff. Even at the peak of government funding, costs were a major disincentive for many firms to join a network. In the context of labour market shortages and high staff turnover in the mid-2000s, a frequent question put to Skillnets officials was ‘why should I invest in my people if they are going to walk across the road?’

In Denmark, the risk of poaching is overcome by the complementarities that exist between the systems of training and collective bargaining, which enable some of the costs and risks of training to be shared. In Ireland, though multi-employer bargaining is not widespread it does exist in some sectors. Retail is one such sector, with industry bargaining in the form of a Joint Labour Committee (JLC). Given that the largest and most successful training network is the Retail Skillnet run by the Irish Business and Employers Confederation (IBEC), there was an opportunity here to experiment with linking training outcomes and wages; this would have provided an incentive for employees to undertake training. Additionally, unions might have negotiated sectoral training levies. These would have provided a more stable funding basis for the training networks and better addressed the evident collective action failures. However, no consideration was given to linking any of the Skillnets training networks with the relevant industry bargaining structure.

Weak union embeddedness in Ireland is reflected in minimal union involvement in the networks. Skillnets originated from national social partnership and as a result trade unions have representatives on the board of Skillnets. However, there is no requirement for union involvement in the individual networks, even when there is a strong union presence in an industry. This makes the networks less effective, as Irish research finds participation rates in employer-sponsored training are higher among union members than non-union members (O’Connell et al., 2009). Additionally, data from Skillnets (2011) shows that participation in the networks is skewed towards those in professional, managerial or supervisory roles, with 69 percent of participants in 2010 being from these occupational groups. In the Danish case, as noted above, union involvement is central to increasing participation among unskilled workers.

Union exclusion from Irish training networks is evidence that training is largely driven by an employer agenda. Union officials were at pains to point out that, at their insistence, Skillnets is officially called ‘enterprise-led learning networks’ rather than ‘employer-led’, as they believe effective training systems ‘need worker and union involvement’. However, the reality is that the actual networks are very much employer-led with little union involvement. The explanation offered by union officials illustrates the asymmetry of power in Irish industrial relations: employers did not want unions involved and so ‘we were not invited’ to participate.

This lack of union involvement reflects the organizational weakness of Irish unions more generally: membership density and bargaining coverage both fell by some 20 percentage points during the period of social partnership. While unions were unhappy with the weak institutional supports for organizing, such as the absence of statutory recognition rights, they pursued ‘dancing’ through the social partnership process rather than ‘boxing’ to achieve their aims. This approach brought wage increases and tax cuts for their members, and gained them policy influence during the period of social partnership. However, their influence below the national level remained limited.

The lack of union involvement in Skillnets also reflects a wider ideological divide over the role of unions in building economic competitiveness. While social partnership persisted at national level for nearly 20 years, it operated within a wider sea of increasingly neoliberal employer attitudes. Indeed, anti-unionism increased during this period among Irish indigenous firms (D’Art and Turner, 2005). Sectoral employers’ associations interviewed by this researcher were opposed to engaging with unions in sectoral or workplace agreements. As the head of one association commented:

There’s a fear that once you get them involved you’re into difficulties. They slow things down. . . . They see the reasons why things shouldn’t be done like that. . . . Every change has to be paid for. … As a result there would be a reluctance to go down that road.

The example of the Retail Skillnet cited earlier is instructive. IBEC is the peak employers’ organization in Ireland and was the central employer representative in national social partnership. Although Skillnets was a social partnership initiative, it formed the Retail Skillnet without union involvement. This paradoxical behaviour reflects the ideological hostility towards unions among its members who, according to one union official, would find union involvement in the network ‘unpalatable’. Employers’ association officials did not perceive they had any role to play in convincing their members to engage with unions. Rather, they saw it as simply a matter of ‘employer choice’.

As noted earlier, because FDI is central to the Irish economy a voluntarist approach to non-market coordination was adopted. Rather than introducing legislation or other forms of institutional constraint, cajoling and encouragement were preferred. However, one of the limitations of voluntarist coordination is that it ignores short-term competitive pressures. This has been a significant issue for Skillnets, with cost-minimization strategies dominating in some sectors, and this predates the current recession: it was evident in interviews conducted in 2004 with industry associations in low-paid sectors. With low unemployment and higher wages available in other industries, they reported severe recruitment and retention problems. At the same time, they also pointed to poor profitability and intense competitive pressures as a result of an over-supply of businesses, particularly in the hospitality, restaurant and tourism industry. Their advice to members was to raise productivity by adopting a value-added product market strategy and investing in their workforce. However, in sectors where cost-minimization strategies prevailed, this message made little impact. One employers’ association official with responsibility for training in the restaurant sector stated:

I personally would feel that the employers don’t give a. . . . They don’t seem to see the advantage of a person who’s trained, and might cost them [more]. That person will sort people quicker, and will have customer care skills, and all the rest of it. But I don’t think that’s recognized in the industry.

In cases where employers in these sectors wished to move to a higher productive strategy, the wider competitive environment prevented them. A report by the Irish Tourism Industry Confederation (ITIC, 2006) pointed to the difficulties the industry faced in raising investment in training, including high staff turnover and staff exit rates from the industry, disruptions to business while employees were on training and the prohibitive financial costs for SMEs.

These sectors were caught in a ‘low-skill equilibrium’. In such an environment, it is in a firm’s rational self-interest not to invest in training, and no amount of cajoling to join a training network can overcome this. As one hospitality employers’ association noted, ‘there’s a flaw in the market. . . . It’s a shared problem needing an institutional solution, but we’re not sure what that looks like.’ However, employers were certain it did not involve constraints. For example, statutory training leave was discussed at national level among the social partners, but union officials reported that employers were strongly opposed because of cost pressures, and the idea never gained much traction. In the current recession, these issues will only have been exacerbated, with training perceived as a luxury few firms can afford. Additionally, the media discourse over the last three years has shifted from one of competing on a high-productivity basis, to one of restoring competitiveness through reducing costs and cutting wages.