Trade unions and the politics of occupational pensions in Denmark and Norway

Abstract

German

French

Around 1990, trade unions in both Norway and Denmark moved away from their previous hostile stance, and started actively to promote occupational pensions. Over time, full coverage of occupational pensions was achieved in both countries, yet with different scopes, degrees of risk-sharing and modes of governance. Why did both countries go through similar institutional changes, and why were the outcomes still so different? The roads to full coverage were different: In Denmark, this was achieved through wage bargaining; in Norway, through mandating. Denmark developed far more collectivistic schemes than Norway, but recent developments indicate convergence: efforts are being made in Norway to develop encompassing schemes more similar to the Danish ones, while the Danish system appears to be moving in a more individual direction. In both countries, strong trade union branches have at critical junctures forged alliances with their employer counterparts at the expense of trade union solidarity. Both mandatory and negotiated schemes thus have the potential to be sources of solidarity as well as of self-interest within the union movement.

Keywords

Occupational pensions industrial relations trade unions tripartite agreement risk-sharing

Introduction

As late as the early 1980s, the trade union movements in both Norway and Denmark were highly sceptical if not directly hostile towards occupational pensions. However, some time in the late 1980s or early 1990s their attitudes and strategies changed fundamentally, and they started actively to promote occupational pensions. Over time, full coverage of occupational pensions was achieved in both countries. The schemes’ scope, degree of risk-sharing and mode of governance however vary considerably. Why did both countries go through similar institutional changes, and why were the outcomes still so different? These are the questions we aim to answer in this article. The empirical analysis is guided by three research questions: a) how can the change of union strategies be explained, b) what form did it take, and c) what implications has it had for the scope and nature of occupational pension provision in the two countries?

The existing literature contains two broad strands of theorising about the dynamics of occupational pensions: a labour relations perspective and a welfare state/retirement provision perspective. The labour relations perspective emphasises that occupational pensions are a form of labour remuneration. From this perspective, it is argued that unions might prefer bargaining for occupational pensions in order to help their members gain access to the associated tax subsidies (Blinder, 1982) and/or because unions tend to give priority to the interests of older workers (Freeman, 1985). A third possible motive for unions to engage in the promotion of occupational pension schemes is membership recruitment and organisational interests of the unions themselves (Therborn, 1986; Pedersen, 1990). From this perspective therefore, unions might have reason to prefer occupational pensions to earnings-related public pensions.

The welfare state perspective instead emphasises that occupational pensions are a vehicle for retirement provision and hence a close substitute for public pensions. Consequently, the dynamics of occupational pensions should be sought in the social policy domain. Influential welfare state theorists have tended to entertain a dualistic view of public and private occupational pensions, associating the former with a strong capacity for redistribution and risk-sharing and seeing the latter as highly inegalitarian and a manifestation of unfettered market forces (see for instance Titmuss, 1955; Korpi, 1989; Esping-Andersen, 1990). Thus, a common assumption in this literature has been that (social democratic) trade unions at the end of the day will prefer public pensions to occupational pensions and only seek to develop the latter as a second-best solution if political power relations prohibit an expansion of public pensions.

This sceptical view of occupational pensions in the welfare policy literature is challenged in more recent contributions that instead emphasise the institutional diversity of this form of retirement provision (see Shalev, 1996; Ebbinghaus, 2011). The point, most explicitly argued by Trampusch (2007), is that occupational pension schemes can incorporate a significant degree of effective solidarity in the form of risk-sharing and power-sharing depending on their coverage, the mechanisms for determining contribution rates and the specific architecture of the schemes. This means that there is considerable potential for trade unions to further social policy objectives by actively engaging in struggles over the scope and nature of occupational pensions. In this context, much attention has been given to the distinction between defined benefit (DB) and defined contribution (DC) schemes, where the former are seen as the natural preference of wage-earners and unions. In this article, however, we will argue that this distinction is as much an oversimplification as the public-private divide, and that there is wide variation among DC schemes in their inclusiveness and degree of risk-sharing.

A strong trend in contemporary social policy studies is the study of institutional stability and change. In this literature, occupational pensions figure as paradigmatic examples of how changes in public welfare programmes can be driven by developments in the non-state sector (Hacker, 2004; Streeck and Thelen, 2005). As the argument goes, in many cases it is much easier to establish new institutions ‘on the edges of old ones’ (Streeck and Thelen, 2005: 31) than to change the existing institutions. This perspective draws attention to the interplay of various parts of the pension system, and to the potential occupational pensions have to influence other parts of the pension system.

As for our choice of cases, both Norway and Denmark are generally considered as belonging to the Nordic family of social democratic welfare states (Esping-Andersen, 1990), and they both feature strong trade union movements and a tradition for centralised wage bargaining (Hall and Soskice, 2001). If there is a potential for effective union involvement and the creation of fairly solidaristic systems of occupational provision, few cases would be more favourable. At the same time, the two countries diverge in both their public pension systems and in their industrial relations traditions, as we will outline below. This constellation of similarities and differences makes the two countries interesting cases for comparison.

We divide the chronological account of developments into two phases. In the first phase starting in the mid-1980s in Denmark and in the mid-1990s in Norway, unions departed from their previous anti-occupational-pensions stance and started pushing to expand coverage of occupational pension schemes. In both countries these efforts culminated in a radical expansion of the coverage of occupational pension schemes, but through very different mechanisms – in Denmark as a result of collective wage bargaining and in Norway as a result of mandating. In both countries the expansion of coverage came in the form of defined contribution (DC) schemes that now dominate the occupational pension markets.

In the second phase we see continuous developments in the nature of occupational pension schemes and conflicts involving issues like union bargaining rights, savings levels, risk coverage and risk pooling as well as union influence over pension capital and the management of pension plans. In Denmark, comparatively solidaristic and collectivistic DC schemes have over time become more individualised, while Norwegian trade unions have struggled with employers, the finance industry and among themselves to achieve a higher degree of collective control and risk-sharing in what is at the outset a system of employer-controlled and highly individualised schemes.

Before we embark on the analysis of occupational pension politics in the two successive phases, we give a brief outline of the distinctive historical and institutional legacies of occupational pensions in the two countries.

Historical context

Denmark has a long tradition of occupational pension provision that predates the national pension scheme. Civil servants were covered by special pension arrangements financed on a pay-as-you go basis and offering an old-age pension (a deferred wage) of about two-thirds of the previous salary. In 1937, legislation on private sector occupational pension schemes was reformed, requiring them to be fully funded. The schemes were taxed in accordance with a standard exempt-exempt-tax (EET) regime, i.e. contributions and returns on savings were exempt from taxation, while benefits were taxed during the retirement phase. In order to qualify for this favourable tax treatment, employers were not necessarily obliged to cover their entire workforce, and could opt for lump-sum payments rather than annuities. In 1982, a new tax on the real returns on occupational pension savings was introduced, moving the system from an EET to an exempt-tax-tax (ETT) regime (Kangas et al., 2010).

The first public pension system was introduced in 1891, when municipalities were required to offer means-tested benefits to elderly citizens. In the 1920s and 1930s, the national old-age pension scheme was expanded and means-testing streamlined. A universal flat-rate benefit to all elderly was introduced in 1957, but unlike its Scandinavian neighbours, Denmark never completely abolished means-testing in the public pension system, and also never introduced an earnings-related second tier to the statutory public pension system (Vesterø-Jensen, 1985).

Despite the lack of a public earnings-related second tier, the overall coverage with occupational pension schemes remained relatively low, reaching around 35 per cent in the mid-1980s. While collective occupational DC schemes had been set up in the 1950s and 1960s for particular professional groups in the public sector (like nurses, doctors and lawyers) as a replacement for the traditional civil servant DB scheme, lower-skilled white-collar workers in the public sector were left without supplementary pension cover. In the private sector only a small minority of employees – primarily white-collar workers – were covered.

Similarly, in Norway, occupational pensions predate national public pensions by several decades. The earliest regulations of private sector occupational pension schemes came in connection with a tax law of 1911. In 1917, a comprehensive occupational pension scheme for civil servants and other categories of state employees (Statens pensjonskasse) was established by law, offering old-age pensions at a replacement rate of two-thirds of the final salary. In the same year, the conditions for preferential treatment of private sector occupational pensions were further specified. To qualify for the preferential (EET) tax treatment, company schemes had to be DB, fully funded, and managed by an insurance company or a pension trust independent of the sponsoring company. From 1952, it was further required that company occupational pensions cover the entire workforce and not aim for higher replacement rates for higher-paid than for lower-paid staff (Hippe and Pedersen, 1992).

The first national public pension in Norway was a means-tested scheme introduced in 1936. Means-testing was completely abolished in 1959, so that all the elderly from the age of 70 would receive a flat-rate benefit. Occupational pensions were from then on required to be ‘integrated with’ (take account of) expected benefits from the public system, and adapted to the role of supplementing rather than replacing benefits from the national pension scheme. Their role was further limited when a relatively ambitious earnings-related second tier (of Swedish inspiration) was added to the public pension system in 1967.

A striking difference between the Norwegian and the Danish trade union movements in this early phase concerns their respective attitudes towards state earnings-related pensions: many Danish trade unions opposed such pensions as a matter of principle (Vesterø-Jensen, 1985), while a unified movement in Norway warmly welcomed the introduction of a second tier of earnings-related pensions in 1967, seeing it as the realisation of a longstanding demand for equal access to earnings-related pensions for all categories as wage-earners (Pedersen, 1990). In the wake of the 1967 reform, LO gave up its demand for negotiated occupational pensions, and even closed down a rudimentary scheme that had been introduced in 1961 (Hippe and Pedersen, 1996).

Nevertheless, public sector occupational pensions continued to exist, as did company schemes in the private sector. Overall coverage with occupational pension schemes continued to grow during the 1970s and 1980s, partly due to public sector employment growth and partly to expanding coverage with company-specific DB schemes in the private sector. Such schemes were comparatively cheap to establish at that time thanks to high nominal and real interest rates (Hippe and Pedersen, 1996).

Phase 1: struggles to expand coverage

In the late 1980s (Denmark) or early 1990s (Norway), trade unions changed their position on occupational pensions, moving from being basically hostile to actively engaging in efforts to expand coverage with occupational pension schemes. In both countries the process was characterised by internal conflicts between different union branches and between national union branches and the peak organisation, the LO.

Denmark

In the 1970s and early 1980s, the Danish trade unions were either disinterested in or directly hostile towards occupational pensions and the associated tax privileges offered to this form of savings. LO’s position on occupational pensions started to change in the mid-1980s. The low coverage among LO’s core constituencies in the private sector was increasingly seen as a serious problem. Growing awareness about the inadequacy of public pensions, particularly for dual-earner households, was an important part of the explanation. Another motive behind the reorientation were the aspirations of the LO leadership to advance the idea of industrial democracy through union control over pension funds. The latter could also help boost unions’ recruitment potential, image and administrative resources.

In 1985, a working group within the LO proposed a new policy approach favouring the expansion of occupational pensions to cover the entire Danish workforce (LO, 1985). At this point, the Danish LO insisted that initiatives at the bargaining table should be backed by legislation in order to secure the enrolment of the entire (private sector) workforce in one unified scheme (von Nordheim Nielsen, 1996).

The liberal-conservative government of the time responded by announcing its willingness to support LO’s demand for occupational pensions on the condition that the LO accepted wage restraint in return for pension contributions. In 1987, a historic tripartite agreement was reached (Den Danske Regering, 1987), and analytical work was initiated on ways to expand occupational pension coverage to the entire Danish workforce. However, the tripartite negotiations broke down in 1989. One of the reasons seems to be that the LO overextended its negotiating position by insisting on a general scheme with a centralised fund in which the trade union movement would be in charge of investment policies (von Nordheim Nielsen, 1996). This was seen as unacceptable by the employers’ associations who in addition were sceptical whether the LO had sufficient control over its national union branches to actually be able to deliver the necessary wage restraint in return for employer pension contributions.

In the wage negotiations in the private sector two years later, the Metal Workers Union and the Union for Unskilled Workers (the ‘SID’, now ‘3F’) broke with the LO position and sided with their employer counterparts to set up a series of industry-wide DC schemes in the construction, transport, trade and services sectors (Due and Madsen, 2005). The manufacturing sector followed suit in 1993.

In the following years, unions and their employer counterparts managed to develop these new embryonic schemes into a comprehensive system of negotiated industry-wide occupational pension schemes – resulting in an overall coverage with DC occupational pension schemes of about 80 to 90 per cent of the Danish employed workforce. The new schemes were in most cases controlled by the social partners in collaboration and managed by pension funds organised as special-purpose insurance companies. As a result, non-profit providers account for some 85 per cent of all private pensions in Denmark.

It should be emphasised that this outcome – the establishment of a network of industry-wide schemes in collaboration with the corresponding employer counterparts – was in direct conflict to the primary position of the LO leadership that had insisted on the creation of a unitary scheme for the entire private sector under union control and backed by legislation. The pension rebellion by national union branches against LO as a peak organisation not only changed the Danish pension system, but also marked a shift of power within the LO in which national union branches took full control of wage bargaining at the expense of the LO’s coordinating role (Due and Madsen, 2005).

Norway

In the decades following the introduction of an earnings-related tier to the public pension system in 1967, trade union officials and activists expected the public scheme to eventually make supplementary occupational pension coverage redundant for the vast majority of wage-earners. Attempts to strengthen occupational pension schemes were seen as potentially threatening the legitimacy of the public pension system.

However, the unions’ positions towards occupational pensions shifted gradually during the 1990s. An important driver was that replacement rates turned out to be lower than foreseen when the system was introduced. Pension rights were for a long period in the 1970s and 1980s not fully indexed to wages, and, as in Denmark, the inadequacy of the pension system to provide income security for dual-earner couples became increasingly obvious. Instead of addressing these issues, a minor retrenchment reform with cuts in earnings-related pensions was introduced under a Labour government in 1992. These cuts in the public old-age pension system made it clear to trade unions that the public scheme would never become generous enough to make supplementary occupational pensions redundant for ordinary workers. At the same time, coverage with occupational pensions in the private sector was limited to just half of the workforce.

Against this background, the LO leadership started voicing a more positive attitude towards occupational pensions for its private sector members in the 1990s, but without directly taking action. As in Denmark, the largest private sector union branch, Fellesforbundet (which includes the metalworkers), took the reins into its own hands in the mid-1990s by starting to lobby for the introduction of favourable tax rules for DC schemes established and financed by individual employers. Fellesforbundet saw this as the only realistic route to achieve higher coverage among its members.

Fellesforbundet’s demand for tax concessions for DC schemes was met with scepticism or outright resistance – both by the Labour Party and by other union branches affiliated to the LO. The public sector unions in particular feared that the introduction of DC schemes in the private sector could undermine the generous DB schemes found in the public sector. Due to these internal tensions, LO was forced to keep a low profile on the issue. The employers’ organisations and the finance industry, however, eagerly supported Fellesforbundet. During the 1990s, DB schemes became increasingly costly to employers due to decreasing interest rates and increasing longevity. Expansion of occupational pensions in the private sector had therefore come to a halt, and many employers were eager to find alternatives to the expensive DB schemes. The finance industry, for its part, was keen to find ways to expand the pension market that had halted at precisely the moment where more capital was needed to cater for increased longevity.

In April 1998, Fellesforbundet and its employer counterpart (TBL) sent a letter to the Prime Minister asking for the introduction of a DC option, and (interestingly enough) demanding concessions for female employees by requiring pension providers to use gender-neutral mortality tables (Fellesforbundet og Teknologibedriftenes landsforbund, 1998). This demand was met with broad political support, and a new law offering tax deductions also for DC arrangements was passed in 1999.

Taking effect from 2001, the new legislation followed the historical tradition of detailed regulations, but in other aspects was radical in its break with existing DB schemes. It assumed DC schemes to be purely individual savings accounts during the accrual phase, only requiring insurance coverage of future premiums in case the employee became disabled before reaching retirement. It was left up to the employer to choose insurance company or portfolio manager and to cover the administration costs. Upon retirement, the accumulated assets must be paid out in annual benefits for a minimum period of 10 years (lump-sum payments are not allowed). For the payment phase, the employer is further allowed a choice between savings accounts with fixed annual payments over fixed number of years (minimum 10) and a conversion into insurance contracts/proper annuities. In 2003, the law was amended, requiring companies to pay higher contributions for female employees if and when the annuity option (with gender-specific annuity calculations) is chosen. This was done to ensure that the annual benefits will be the same for women and men with a given earnings history.

The 2001 Act did not have the effect its proponents had hoped for. The rise in coverage rates was sluggish, with growth coming almost exclusively at the expense of traditional DB schemes. As a reaction, attempts were made by Fellesforbundet (this time with LO support) to negotiate a collective, industry-wide agreement requiring employers to establish company schemes in both the 2002 and 2003 negotiations. However, in contrast to their Danish peers, the peak employer organisation (NHO) and the affiliated employer association for manufacturing industries (TBL) firmly refused to negotiate occupational pensions as part of the national wage-bargaining process. In 2004, however, the parties – again – sent a joint letter to the government, asking for a law to make occupational pensions mandatory in the private sector. The timing was perfect: with a major retrenchment reform of the National Insurance old-age pension scheme underway, the government hoped that accommodating the LO on occupational pensions would help ensure support for the pending reform. Possibly the employers’ organisations shared this hope and, moreover, figured that a mandate requiring modest contribution rates would primarily represent a burden on smaller companies outside the collective bargaining system. Furthermore, mandating allowed the employers’ organisations to remain steadfast in their rejection of negotiating occupational pensions at the central level. Contrary to the situation in Denmark, the Norwegian occupational pension market is dominated by traditional for-profit insurance companies, and the life-insurance industry embraced the mandate, since it promised a market expansion that voluntary employer decisions had not been able to bring about.

The so-called OTP Act that came into force in 2006 mandated all private sector employers with at least two employees to establish a company occupational pension scheme with a minimum contribution of 2 per cent of annual salaries between a minimum threshold and an upper ceiling at about two times the average full-time wage. The mandate left it up to companies to choose between DB and DC schemes and, obviously, contribution rates could be set higher than the minimum level.¹

As a direct consequence of the Act, coverage of occupational pension schemes in Norway became virtually complete by 2006. However, practically all the new schemes were highly individualised DC schemes without insurance components and with only minimum contribution rates (Veland and Hippe, 2008).

Phase 2: debates and conflicts about scheme features

As we have seen, (almost) complete coverage with occupational pension schemes was achieved in Denmark in the early 1990s and in Norway in 2006. In both countries the vehicle for this expansion of coverage was DC schemes in which the financial responsibility of employers is strictly limited to the payment of fixed annual contributions. However, the nature of the DC schemes that emerged after the struggle for increased coverage differed widely between the two countries. The Danish schemes were at the outset relatively collectivistic with strong union involvement and extensive risk coverage and risk-sharing among members of each scheme, while the Norwegian DC schemes were typically very modest and highly individualised.

The second phase is dominated by struggles over the nature, scope and governance of the schemes, with tensions within the trade union movements becoming more pronounced.

Denmark

The schemes established and developed in Denmark during the 1990s were organised as insurance schemes offering risk coverage for disability and survivors’ benefits in addition to old-age (lifelong) annuities. In most schemes contributions were immediately converted to a ‘with profits’ insurance contract based on a fixed guaranteed return rate. The collective and semi-obligatory nature of the Danish schemes vis-à-vis individual companies and their employees made it possible to base the annuity conversion and risk premiums on unisex mortality tables. Since 1996 gender neutrality has been obligatory for Danish pension providers, implying quite significant redistribution from men to women among the members of each scheme – as long as they provide lifelong benefits. Typically, there was no individual investment choice, and little or no choice with respect to insurance coverage and the benefit profile. Workers were affiliated with a particular scheme without the right to move pension assets between schemes. Most schemes operated with a system to level out fluctuations in annual returns on the scheme’s total assets.

In sum, we can say that, despite being of the DC type with employers’ responsibility limited to paying fixed contribution rates, the Danish occupational pension schemes allowed for significant risk-sharing and redistribution among members of each scheme – both between genders and across cohorts. While this type of risk-sharing is strictly limited to members of each scheme, more comprehensive risk-sharing is arguably provided by the extensive use of income-testing in the public pension system (Sørensen, 2015).

In the high inflation environment of the 1970s, 1980s, and early 1990s, the pension providers had few problems honouring insurance contracts based on a high guaranteed nominal return rate. Also contributing to the sustainability of these types of scheme was the fact that the growth in longevity among the Danish population was comparatively slow up until the turn of the Millennium.

However, the external conditions started to change in the mid-1990s with the advent of decreasing and – as it turned out – persistently low interest rates and rapidly increasing longevity. These two trends both increased the capital needed to fund the promised benefit stream, and put the Danish pension providers under severe financial pressure. Signs to this effect became visible already in the 1990s and were further exposed by the ‘dotcom’ crisis in the late 1990s, the stock market crisis in 2003–2004 – and later the 2009 financial crisis.

The responses have developed gradually, and they have come from the state and pension providers as well as the social partners. The Danish financial supervisory authority reacted by tightening regulation. The most important steps were to lower the maximum allowed guarantee rate for new contracts (in steps to currently 1 per cent), to mandate market rate valuation on the liability side (2002), and to introduce stress testing in the solvency assessment, adopting a mortality benchmark framework (2011). The target of these efforts is to safeguard members’ rights and ensure the financial sustainability of pension schemes.

Partly as a result of demographic and economic pressures and the associated changes in financial regulations – but also in part responding to demands for greater flexibility from wage-earners – the Danish DC schemes have changed towards more individualisation and reduced risk coverage and risk-sharing between scheme members (Finanstilsynet, 2017).

The dominance of the classic guaranteed DC insurance approach has diminished, giving way to greater diversity. Pension plans with uniform, collective investment portfolios and mechanisms to smooth fluctuations in annual returns on capital have become the exception. Many providers now offer individual choice between different risk profiles and a few even offer members products with individual investment choice.

On the benefits side, the traditional insurance approach has increasingly been replaced by an approach based on annuitisation at retirement or combinations of fixed-term withdrawals and annuities commencing at a relatively high age. In 2014 only 57 per cent of total pension funds were still dedicated to plans offering lifelong annuities, while 28 per cent and 15 per cent belonged respectively to schemes offering fixed-term payments, or lump-sum benefits (Finanstilsynet, 2017: 5). Limited sharing of longevity risk has reduced the degree of redistribution in favour of female scheme members.

Even after these changes, the Danish occupational pension market remains collectivistic in the sense that participation is compulsory for individual employers and employees and contribution rates are determined by collective agreements. Trade unions have, in collaboration with their employer counterparts, played a key role in setting up the Danish occupational schemes, in raising their contribution rates and in adapting their designs to changing economic and demographic conditions. Hence, many schemes now have substantially higher contribution rates than originally anticipated, with a view to compensate for increasing longevity and low interest rates.

In recent years, unions and pension industry representatives have started to question the interaction between occupational pensions and the public pension system. The strong role of income testing in the public pension system is an important risk-sharing feature, but it also implies that the marginal gain from increased occupational pension savings is seriously curtailed for large segments of wage-earners (Sørensen, 2015). From this perspective, one can argue that that the wage restraint offered by LO unions in the 1990s in return for pension contributions was of little use insofar as increases in future private benefits are to a large extent neutralised through automatic reductions in income-tested public pension benefits (Pensionskommissionen, 2015). The severity of the issue was illustrated in the spring of 2017 by the fact that the union representing low-wage workers in the municipal sector (‘FOA’) threatened to withdraw its support for the further consolidation of occupational pension schemes, arguing that members’ contributions are largely paid in vain.

As a response to these concerns voiced by both trade unions and the pensions industry, the former social democratic government appointed an expert commission to look into the interaction between public and occupational pensions. The commission published a preliminary report in January 2015, highlighting the high implicit marginal tax rates caused by income testing of public old-age benefits (Pensionskommissionen, 2015). However, the commission was closed down by the new liberal-conservative government that came into power in June 2015. In the spring of 2017, a broad political compromise was settled to increase incentives to defer retirement and improve incentives to save in the last years prior to retirement. At the time of writing (autumn 2017) new legislation is being prepared offering a tax credit to older workers saving for retirement, but there are no plans to increase the utility of private savings for current pensioners.

Norway

In Norway, all public sector employees are covered by relatively generous DB schemes. Since 2006, virtually all private sector employees are similarly covered by occupational pensions, but the majority of these schemes are relatively meagre DC schemes that mainly take the form of individual savings schemes. The discrepancy between the sectors is obviously problematic, not the least for the LO that comprises unions organising workers from both sectors. In addition, the 2009 pension reform, which transformed the public pension system into a notional defined contribution (NDC) system of Swedish inspiration (see Christensen et al., 2012), made the public sector DB pensions difficult to sustain in their traditional form.

As a response both unions and policy-makers started to look for middle-of-the road alternatives ‘between’ DC and DB schemes – so-called hybrid schemes. The Norwegian notion of ‘hybrid’ schemes implies something similar to the relatively collectivistic Danish DC schemes, including a higher degree of risk-sharing, lifelong benefits, and including a possibility for employers to guarantee to link the indexation of pension rights to wage increases.

Legislation allowing companies to establish hybrid pension schemes was passed in 2014. Though such schemes are preferred by most unions over pure DC schemes, as yet they have achieved very limited coverage. They are mainly found among private companies that compete with the public sector, such as private providers of education, health care, social services and culture.

The 2014 regulations for hybrid schemes did not meet LO’s wish to open the door to broader, industry-wide pension schemes – hybrids are still supposed to be company-specific. LO argues that broader schemes covering large segments of the private sector workforce will be more cost-efficient (due to economies of scale) and will facilitate greater risk-sharing than the existing company-specific schemes. In this, however, LO faces joint opposition from the employers’ association and the finance industry, both of which are committed to promoting a DC-based decentralised system. In fact, the finance industry has been running a campaign arguing against including pensions in national-level collective agreements. Its core argument is that the introduction of larger pension providers will reduce market competition (PA Consulting Group, 2014).

Although LO is united in its desire to put occupational pensions on the agenda in central wage bargaining, there are strong tensions over the preferred nature of the schemes. The LO leadership, public sector unions and a range of smaller private sector unions prefer the establishment of broader sector-wide or industry-wide schemes that could maintain and strengthen the collective nature of occupational pension provision and allow for risk-sharing among men and women. Fellesforbundet, by contrast, questions the realism of this position and insists on pursuing a further individualisation of occupational pension savings (Gabler, 2015).

In the 2016 round of wage bargaining, these tensions came to the fore. Fellesforbundet was instructed by the LO to demand an increase in minimum savings, to search for solutions to the problem of limited portability of pension rights for mobile workers, and – controversially – to seek the establishment of broad, encompassing schemes. The latter demand was added by the representative body in LO against the will of Fellesforbundet and against the explicit recommendation of the LO chairman. The employers’ associations however refused, again, to negotiate pensions, and as a compromise the parties agreed to ask the government to engage in analytical work to find solutions to the problems of portability. This can be seen as a bid to gain time in a situation where the positions – both within LO, and between unions and employers’ organisations – are rather antagonistic with regard to both the nature and the governance of occupational schemes.

The strong resistance of employers to negotiating occupational pensions should be seen in light of the fact that the social partners are already dealing with the so-called AFP scheme (Avtalefestet pensjon / Negotiated pension) introduced in 1988. Originally an early retirement scheme, AFP was transformed in 2008 to a top-up, lifelong pension supplement. The existence of the AFP scheme may be a partial explanation for employers’ resistance to negotiating occupational pensions, as they have already taken on heavy and risky obligations in the field, but it undermines their attempts to flag this resistance as a matter of principle.

Conclusion

Despite substantial differences in public pension provision between Denmark and Norway, there are interesting similarities in the politics of occupational pension provision over the last decades. In both countries, trade unions changed their views on occupational pensions from being highly sceptical to becoming active proponents of such schemes. In both countries, more or less complete coverage with DC schemes was eventually achieved, although through different mechanisms: collective agreements in Denmark and mandating in Norway. While unions in both countries have traded risk-sharing for coverage, there are important differences in the collective nature of DC arrangements.

Denmark started out with far more collectivistic and solidaristic DC schemes than Norway, but recent developments indicate convergence: efforts are being made in Norway to develop encompassing schemes more similar to the Danish ones – although progress in this area is slow – while calls for individual choice and a series of financial pressures have combined to push the Danish system in a more individualistic direction with less extensive risk-sharing. Even so, at the time of writing (autumn 2017), it is still the case that Danish unions have come closer to realising the potential for creating a fairly solidaristic system of occupational pensions envisaged by Trampusch (2007).

A rather puzzling difference between the two countries concerns the position of employers. Danish employers have accepted including occupational pensions in collective wage bargaining and, during the 1990s, offered increased employer contributions in exchange for wage restraint from national unions. By contrast, the Norwegian employers’ associations have so far firmly refused to take up occupational pension demands in national-level collective wage bargaining (see also Trampusch, 2013). Moreover, the pension industry in Norway plays a role that the tradition for labour market self-regulation appears to have denied it in Denmark. In other words, both the player constellations and their priorities differ between the two countries.

Through analysing political processes and the players’ priorities, we have been able to confirm some of the dynamics of incremental institutional change suggested in the literature. The inability of governments in both countries to update and maintain public pension systems in tune with changes in the surroundings – most obviously the increasing dominance of dual-earner couples – can be seen as examples of policy drift (Streeck and Thelen, 2005). These developments led directly to a classic situation of layering (Streeck and Thelen, 2005), where critics of the status quo gave up trying to change the existing institutions and established their own ‘on the edges’ of the old ones. This in turn has changed the status of the old system: in Denmark, the wide coverage of occupational pensions has led to a multi-pillar system where private pensions complement public pensions. This process however, has altered the distributional effects of the means-testing of public pensions, and – as we have seen – its legitimacy is now called into question. In Norway, mandatory occupational pensions helped pave the way for a far more comprehensive – some will say paradigmatic – reform of the pension system (in 2009).

While the processes of drift and layering are similar in the two countries, there are also important differences, indicating a significant degree of contextual variation and path-dependence. Players’ priorities and scope for action can at least partially be understood in the light of the countries’ different industrial relations traditions and historical legacies. The higher union density and coverage with collective wage agreements in the private sector in Denmark made it possible to achieve comparatively high coverage rates without legislative backing. A similar approach in Norway would have left behind a significant share of private sector wage-earners. Obviously, the Danish self-regulation tradition also plays a role here, but it should not be forgotten that the Danish LO originally preferred a tripartite agreement on a broad occupational pension scheme for the entire private sector, supported by legislation. Moreover, it seems that in Denmark the collective DC schemes under the mutual control of the social partners opened up for a relatively rapid and flexible response to the challenges of increasing longevity and low interest rates. With the decentralised and strictly regulated DB schemes in Norway, change was harder to bring about. The financial pressures on both employers and providers continued to increase until a radical transition towards pure DC schemes became the only viable alternative for employers and life-insurers.

Our most important finding and original contribution to the existing industrial relations literature is that in both countries, occupational pension politics has created strong tensions within the union movement. At critical junctures, key union branches have forged alliances with their employer counterparts and outmanoeuvred other sections of the labour movement. In Denmark around the 1990s, an internal struggle ended in a bitter defeat for the LO leadership and a fundamental change of the balance of power within the LO family in favour of national union branches. In Norway, there are ongoing tensions between the male-dominated and export-oriented Fellesforbundet pushing for more decentralised and individualised models with a view to increasing coverage and de facto savings rates, and other private and public sector unions who give priority to qualitative aspects and strive to maintain or establish schemes under collective control and with greater risk-sharing.

Full coverage by DC schemes has been achieved in both Norway and Denmark and private pensions will play a significant role in the retirement income for most pensioners – especially in the longer term. However, while this development is generally embraced as a major achievement, there are ongoing tensions and conflicts regarding the nature and the governance of the schemes in both countries. How should the conflict between distributional concerns and incentive concerns be balanced? What level of centralisation is desirable as regards the scope of the schemes – should they be company-specific and cater to single company preferences and interests or should they be industry-wide and seek to facilitate mobility? How should the need for individual influence and choice be balanced against the need for risk-sharing – including between women and men? Even in rather mature multi-pillar settings such as the Norwegian and the Danish ones, these issues remain controversial and unresolved.

Footnotes

Funding

Support from the Research Council of Norway [grant numbers 238205 and 238202] is gratefully acknowledged.

Notes

References

Blinder

(1982) Private pensions and public pensions: Theory and fact. NBER Working Paper No. 902. Cambridge: The National Bureau of Economic Research.

Christensen

Fredriksen

Lien

Stølen

(2012) Pension reform in Norway: Combining an NDC approach and distributional goals. In: Holzmann

Palmer

Robalino

(eds) Nonfinancial Defined Contribution Pension Schemes in a Changing Pension World, Volume 1: Progress, Lessons, and Implementation. Washington: The World Bank, pp. 129–144.

Den Danske Regering (1987) Fælleserklæringen af 8. december 1987. Available at: http://danmarkshistorien.dk/leksikon-og-kilder/vis/materiale/faelleserklaeringen-af-8-december-1987/?no_cache=1 (accessed 30 November 2017).

Due

Steen Madsen

(2005) Et tilfældig sammenfald af særlige omstændigheder. Arbejdsmarkedspensjonernes indførelse i 1980-erne. In: Petersen

Petersen

(eds) 13 reformer af den danske velfærdsstat. Odense: Syddansk Universitetsforlag, pp. 189–201.

Ebbinghaus

(ed.) (2011) The Varieties of Pension Governance, Pension Privatization in Europe. Oxford: Oxford University Press.

Esping-Andersen

(1990) The Three Worlds of Welfare Capitalism. Cambridge: Polity Press.

Fellesforbundet og Teknologibedriftenes landsforbund (1998) Brev til statsministern av 20 april 1998.

Finanstilsynet (2017) Pension når garantierne forsvinder. Hvad er implikationerne for produktegenskaber og forbrugerbeskyttelse? Et diskussionsoplæg 27. februar 2017. Copenhagen: Finanstilsynet.

Freeman

(1985) Unions, pensions, and union pension funds. In: Wise

(ed.) Pensions, Labor, and Individual Choice. Chicago: University of Chicago Press, pp. 89–122.

10.

Gabler (2015) Innskuddspensjon. Forslag til lovendringer som vil skape en mer bærekraftig pensjonsordning som sikrer mobilitet i arbeidslivet og trygg pensjonstilværelse for arbeidstakere. Oslo: Gabler AS.

11.

Hacker

(2004) Privatizing risk without privatizing the welfare state: The hidden politics of social policy retrenchment in the United States. American Political Science Review 98(2): 243–260.

12.

Hall

Soskice

(2001) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage. Oxford: Oxford University Press.

13.

Hippe

Pedersen

(1992) Når jobben betaler. En analyse a velferdsordninger i arbeidsmarkedet. Fafo-report no. 136. Oslo: Fafo.

14.

Hippe

Pedersen

(1996) Labor movement, social policy, and occupational welfare in norway. In: Shalev

(ed.) The Privatization of Social Policy? Occupational Welfare and the Welfare State in America, Scandinavia and Japan. London: Macmillan, pp. 187–210.

15.

Kangas

Lundberg

Ploug

(2010) Three routes to pension reform: Politics and institutions in reforming pensions in Denmark, Finland and Sweden. Social Policy & Administration 44(3): 265–284.

16.

Korpi

(1989) Power, politics, and state autonomy in the development of social citizenship: Social rights during sickness in eighteen OECD countries since 1930. American Sociological Review 54(3): 309–328.

17.

LO (1985) LO’s Pensionsreformudvalg. Forslag til en samlet pensionsreform. Copenhagen: LO.

18.

PA Consulting Group (2014) Kostnadseffektivitet og markedsdynamikk i det skandinaviske tjenestepensjonsmarkedet. Oslo: PA Consulting Group.

19.

Pedersen

(1990) Fagbevegelsen og folketrygden. LO’s målsetninger, Strategi og innflytelse i pensjonspolitikken 1945–1966. Fafo-report no. 110. Oslo: Fafo.

20.

Pensionskommissionen (2015) Det danske pensionssystem – internationalt anerkendt, men ikke problemfrit. Copenhagen: Skatteministeriet.

21.

Shalev

(ed.) (1996) The Privatization of Social Policy? Occupational Welfare and the Welfare State in America, Scandinavia and Japan. London: Macmillan.

22.

Streeck

Thelen

(2005) Introduction: Institutional change in advanced industrial economies. In: Streeck

Thelen

(eds) Beyond Continuity: Institutional Change in Advanced Political Economies. Oxford: Oxford University Press, pp. 1–39.

23.

Sørensen

(2015) The Interaction of Public and Private Pensions – the Danish Case. ATP Analysis report no. 2:2015. Hillerød: ATP.

24.

Therborn

(1986) The working class and the welfare state. A historical-analytical overview and a little Swedish monograph. In: Kettunen

(ed.) Det nordiske i den nordiska arbetarrörelsen. Papers on Labour History. Helsinki: Finnish Society for Labour History and Cultural Traditions, pp. 1–75.

25.

Titmuss

(1955) Pension systems and population change. Political Quarterly 26: 152–166.

26.

Trampusch

(2007) Industrial relations as a source of solidarity in times of welfare state retrenchment. Journal of Social Policy 36(2): 197–215.

27.

Trampusch

(2013) Employers and collectively negotiated occupational pensions in Sweden, Denmark and Norway: promoters, vacillators and adversaries. European Journal of Industrial Relations 19(1): 37–53.

28.

Veland

Hippe

(2008) Når tjenestepensjon er obligatorisk – hva så? Fafo-report no. 2008/5. Oslo: Fafo.

29.

Vesterø-Jensen

(1985) Det tvedelte pensionssystem. Forskningsrapport nr. 1/84. Roskilde: Institut for samfundsøkonomi og planlægning, Roskilde Universitetscenter.

30.

von Nordheim Nielsen

(1996) Danish occupational pensions in the 1980s: From social security to political economy. In: Shalev

(ed.) The Privatization of Social Policy? Occupational Welfare and the Welfare State in America, Scandinavia and Japan. London: Macmillan, pp. 241–260.