Abstract
In an era of fiscal austerity and dualization of social protection, has organized labor become increasingly split along skill and industry lines? Against recent political science accounts of trade union involvement in social policymaking, this paper argues that, in the specific area of pensions, unions representing high-skilled workers and the core industrial sectors of the economy have paradoxically been led to increase their cooperation with unions representing the less privileged segments of labor, in order to improve coverage of private pensions across the board. These unions’ motivations for doing so and the strategies they have employed have nonetheless differed according to the preexisting institutional design of domestic pension systems. The argument is supported with case studies of British, French, German, and Belgian unions’ involvement in contemporary pension reform.
In an era of permanent austerity and privatization of social protection, are trade unions characterized by greater occupational egoism? And has organized labor become increasingly divided? Students of pension reform have recently argued that, because of their members’ higher bargaining and earning power, trade unions representing primarily high-skilled employees have often favored private pension provision over more redistributive public plans and have consequently been in conflict with unions that include less skilled and low-wage workers. 1 Their analysis echoes that of Palier and Thelen on increased segmentation—or “dualization”—in the French and German political economies. 2 These two authors have suggested that, although they historically set relatively progressive social standards for the rest of the economy, workers and employers in the “core”—generally more skill-intensive—manufacturing industries have increasingly “turned inward” and shaped labor market and social policy institutions primarily “for themselves,” thereby creating a rift with other industries and segments of labor.
There is no doubt that the contemporary trend toward retrenchment and privatization of pensions may generate social protection dualism. 3 In this article, however, we challenge recent accounts of labor behavior in the field of pensions and argue that, in the face of growing political pressures for pension privatization, trade unions representing high-skilled and less-skilled workers have paradoxically been led to increase their cooperation with one another in this specific policy area, so as to extend coverage of supplementary pensions to as large a group of workers as possible.
The unions’ reasons and the strategies they have employed have differed according to the preexisting institutional design of domestic pension systems. In countries that had only basic public pensions and relatively large coverage of employer-provided occupational pensions after the Second World War, these traditional supplementary plans have come under mounting financial pressures, but have also been challenged by politicians and the financial industry because of their institutional fit with a fading era of stable industrial employment. To preserve their existence, trade unions representing high-skilled segments of labor have allied with the rest of the labor union movement to push for the expansion of supplementary pensions to the whole of the workforce, thereby reducing existing dualisms in coverage.
Political dynamics and unions’ strategies were different in countries that had more generous public pensions and much more marginal coverage of occupational pensions during the postwar period. In these countries, trade unions representing high-skilled workers, particularly those in the manufacturing industries, have continued defending generous public pensions. However, faced with governments’ increasing commitment to pension retrenchment and privatization, they have tried to counter the expansion of individual retirement savings plans managed by commercial financial services companies and sought to expand encompassing and not-for-profit occupational pension plans through industry-level——rather than firm-level——collective agreements with employers. Unions representing workers in the metalworking industries have been at the forefront of this strategy; they have seen it as a means of preserving their role as pacesetters for other sectors of the economy and other segments of labor.
We support our argument with case studies of British, French, German, and Belgian trade unions’ involvement in the contemporary politics of pension privatization. Unlike, for example, the Nordic countries, the four selected countries all represent likely cases of social protection dualization. Indeed, Britain embodies the liberal world of welfare where, because of the importance of private social protection, segmentation has been historically quite prevalent. By contrast, France, Germany, and Belgium all belong to the group of conservative welfare regimes in which dualistic pressures are considered to be on the rise. 4 The four cases also have the advantage of presenting useful variation on the organizational factors that might affect trade unions’ tendency to take either a “narrow” or a more “encompassing” view of workers’ interests. 5 Thus, whereas Britain and Germany have been moderately unionized, France’s union density has been among the lowest in Western Europe. By contrast, Belgium has had the highest rate of unionization after the Nordic countries. 6 Furthermore, unions’ organizational traditions have differed among the four countries, with the importance of crafts and occupational unionism in Britain, the dominance of industrial unionism in Germany, and significant fragmentation of organized labor along ideological and religious lines in France and Belgium. 7
Instead of an “inward turn” of organized labor, we find that, in the specific field of private pensions, unions representing different segments of labor have been led to build encompassing wage-earner alliances, which have in turn been able to enter into coalitions with parts of the business community. In Britain and France, where private-sector workers historically had access to limited statutory social insurance but were covered by a large, though in many ways still voluntary, occupational pension sector, skilled and less skilled workers’ unions have fought together in the 1990s and the 2000s to make these supplementary pensions compulsory for all workers. In Germany and Belgium, where public social insurance was generous and only a minority of the workforce benefited from private schemes, unions in the metalworking industry have pressed since the late 1990s for the creation of collectively negotiated pension plans and for their expansion across other industries.
The next section develops our argument in more detail. Section 2 presents our four historical case studies. The final section concludes and examines the implications of our findings for research on the politics of dualization.
Trade Union Egoism or Solidarism?
Over the past two decades, comparative political economists have called into question dominant class-based accounts of welfare state development. The power resources theory traditionally argued that social citizenship rights were institutionalized as a result of the working class mobilizing through trade unions and social-democratic parties, against strong resistance of employers. 8 Yet students of industrial relations cast doubt on these assumptions by highlighting the heterogeneity of workers’ interests over wage distribution 9 and by showing that centralized—and more redistributive—forms of wage bargaining had typically been established with the support of alliances between specific segments of labor and business. 10 The literature on the varieties of capitalism later contributed to extend the analysis of such cross-class alliances and intra-labor cleavages to the realm of social policy. 11
Recent studies have suggested that the divides between segments of the workforce have become increasingly consequential since the oil shocks of the 1970s. For example, students of party politics have argued that mass unemployment and the rise of atypical (e.g., fixed-term and part-time) work contracts have created a conflict between “insiders” and “outsiders” over social policy and have therefore presented political, particularly left-wing, parties with strong electoral dilemmas. 12 But structural changes in the economy, including the feminization of the workforce and the shift of employment toward services, have also added to the heterogeneity of trade unions’ constituencies and are thought to have made their strategic options more difficult. 13 Intra-union divisions are potentially most apparent in the area of social policymaking that generates the highest levels of expenditure, namely, old-age pensions. Indeed, in addition to dividing workers according to their age group 14 or their gender, 15 pensions may also split them by occupational or skill profile.
The skill divide is considered particularly salient when it comes to the choice between public and private retirement provision. In a historical study of the development of the British and Dutch postwar pension systems, Oude Nijhuis has argued that organized labor’s support for the redistribution of workers’ income in old age through public benefits has crucially depended on the interaction between unions’ skill profile and their organizational structure, particularly whether they represent large sections of the workforce or prefer to cater to the interests of narrower constituencies. 16 He found that in the Netherlands a strong tradition of industrial unionism, which consists in organizing both skilled and less skilled workers in encompassing industry-level unions, led Dutch labor to favor the creation of redistributive social security in the 1950s. By contrast, the prevalence of—more narrowly skill-based—crafts and occupational unionism among British skilled manual and white-collar workers led these groups to oppose, throughout the 1950s and 1960s, any redistribution of risks and income via the public pension system. By that time, these groups were already covered by generous supplementary employer-provided pension schemes. Given the narrowness of their membership and their constituencies’ vested interest in preserving these advantages, crafts and white-collar unions did not try to accommodate the pro-redistribution preferences of less skilled workers, who were generally not covered by occupational plans.
In her study of the contemporary politics of pension reform in Continental Europe, Häusermann has also emphasized the existence of a skill-based intra-labor cleavage over private pensions, but she argues that the transition to a postindustrial economy has led to growing conflict between workers. 17 She has found that, although trade union movements have generally remained unified in their defense of existing public pensions, they have become increasingly split over governments’ attempts to expand private plans. According to Häusermann, whereas low-skilled manual and service workers’ unions have opposed funded pensions on the grounds that they threaten public provision, unions representing high-skilled blue-collar or white-collar workers—such as the German chemical workers’ union (IG Chemie) and the French professional and managerial employees’ union (CGC)—have been much more accommodating, because their members could expect to be offered supplementary benefits by their employers.
In a similar vein, though their study does not focus on pensions, Palier and Thelen have argued that over the past few decades high-skilled workers in the core manufacturing industries have increasingly tended to reach deals with their own firms’ managers and have therefore left other industries and other, often less-skilled, workers behind. 18 They contend particularly that, whereas it used to bargain with employers over social standards that bridged the gap between skilled and less skilled workers and were diffused through pattern bargaining to all sectors of the economy, organized labor in the German metalworking industries (e.g., the IG Metall union) and in French “national champion” firms (e.g., automaker Renault) has lost its capacity to play this pace-setting role. By accepting to delegate more and more issues to firm-based bargaining, it has encouraged workers’ “local egoism,” which has in turn led to the institutionalization of new forms of labor market and social protection dualisms. Taken together, these studies suggest—or at least convey the impression—that instead of trying to prevent dualization, organized labor may on the contrary have exacerbated it.
We do not deny the existence of significant cleavages within the workforce and the trade union movement, but we take issue with the recent literature’s assumption that organized labor has systematically become more divided and more egoistic. There may be exceptions to this purported pattern. Against what is becoming a conventional understanding, we argue that, in a position of greater political weakness, trade unions representing distinct segments of labor may also be led to close ranks and foster solidarity among each other. This type of coalitional dynamics has arguably occurred specifically in the field of old-age pensions.
With the liberal turn at the beginning of the 1980s and growing emphasis on fiscal pressures facing the welfare state, public pensions have been a prime target in governments’ retrenchment initiatives. 19 At the same time, a political consensus has emerged over the need to shift pension provision to the private sector, driven by the promotion of new policy paradigms by international organizations 20 and, more important in the affluent democracies, by considerable lobbying on the part of financial services companies, which stand to profit from the development of a market in supplementary pensions. 21 In this context, organized labor’s initial, and unanimous, response has been to defend the existing public pension system. 22 However, as they have been unable to stem the tide of pension retrenchment and privatization, trade unions representing different segments of labor have been led to unify strategically in an attempt to mitigate the distributional impact of these developments and to extend coverage of supplementary pensions to larger sections of the workforce, through either collective bargaining or legislative changes.
Because of the variety of preexisting postwar pension arrangements and the feedback effects they have exerted, 23 the pathways that have led unions to increase their cooperation have differed cross-nationally. We distinguish between two groups of countries: those that, after the Second World War, had only modest public pensions, but complemented by a relatively large coverage of employer-provided occupational benefits, and those that had more generous public pensions, with only a residual role of occupational plans.
As has been the case elsewhere, public pensions in the first group of countries have been subject to cuts by governments. However, preexisting occupational schemes, which already played a very significant role in pension provision, have been challenged just as much. This has had to do, first, with the financial pressures they have faced. Postwar occupational plans were often of the more generous “defined-benefit” type 24 and, thanks to looser government regulations, they did not necessarily accumulate enough assets to meet mounting liabilities. With demographic aging, it has thus become increasingly costly for employers to provide them.
Second, they have been criticized for their lack of institutional fit with more service-oriented and flexible labor markets. Indeed, as they were used by largely industrial employers either to retain skilled workers or to pacify labor relations, occupational plans often favored long-standing employees over “early leavers.” To become eligible, workers frequently had to complete a minimum number of years of service. Moreover, their accrued rights were not necessarily portable across firms. Such rules could thus hamper job mobility. Last but not least, traditional occupational schemes have become a target for the financial services industry. This is because they were very often directly administered by employers and trade unions on a not-for-profit basis. Commercial financial firms have thus perceived them as a competitor in the market for private retirement savings and have targeted them in order to seize a greater share of the business. 25
As financial and political pressures on occupational pensions accumulate in countries where they already played a significant role in workers’ retirement income package, their preservation has been of strategic importance for organized labor. Since coverage of such schemes was not fully compulsory and was most prevalent among white-collar and high-skilled manufacturing workers, trade unions representing these segments of labor have tried to defend occupational pensions’ existence and legitimacy by pushing for their extension to the whole of the workforce. In doing so, they have found natural allies in unions representing less skilled—and traditionally less well covered—workers. But they have also forged cross-class coalitions with employers’ associations in the manufacturing industries, because these parts of the business community still have an interest in preserving schemes they had helped set up.
In countries with more generous social insurance, pressures for reform have concentrated on public schemes. And, despite their opposition to it, unions have been unable to stop the trend toward retrenchment. 26 As policymakers have sought to compensate for the declining generosity of public pensions with the development of private provision, unions have defended a specific type of supplementary schemes. Instead of accepting the expansion of individual retirement savings accounts managed by for-profit providers or of firm-level occupational plans, organized labor has advocated the institutionalization of more encompassing and not-for-profit industrywide pension funds. By favoring such “collectively negotiated schemes,” unions have effectively sought to make supplementary pensions accessible to as many workers as possible and to support the creation of a form of “social security through industrial agreements.” 27
We argue, in particular, that unions in the core manufacturing—more precisely metalworking—industries have played a crucial role in putting this strategy on the agenda. As stressed by Palier and Thelen, 28 these unions were traditionally leaders in Continental and Nordic collective bargaining and helped organized labor press for progressive social standards. 29 Their strong position stemmed not only from the centrality of manufacturing in the economy, but also from the fact that as industrial unions they represented both skilled and less skilled workers and therefore strove to unite different segments of labor. 30 At a time when industrial relations have been increasingly in crisis, 31 occupational pensions constitute a new item on the negotiation agenda, in addition to wages or working conditions. 32 Metalworkers’ unions—and their counterparts in organized business—have therefore seen occupational pensions as a way of revitalizing industry-level collective bargaining and of bolstering their pacesetting role.
In sum, whereas much recent literature emphasizes a growing dissension and egoism in the trade union movement over labor market and social policy issues, we argue that, in the conflict-ridden area of pension policymaking, different segments of organized labor have been led to reaffirm their commitment to solidaristic values. Where occupational pensions were already widespread, trade unions have fought to extend coverage to all workers. Where supplementary pensions have been on the rise, trade unions have tried to expand their reach to as many workers as possible through industry-level collective agreements. The next section illustrates our claims with a comparative historical analysis of recent British, French, German, and Belgian politics of pension privatization. We start with a general description of these pension systems’ postwar institutional design, which is necessary to understand their subsequent evolution.
Combating Dualisms in Private Pension Provision in Liberal and Continental Welfare Systems
Although Britain and France are generally considered to have belonged to two different worlds of welfare 33 —Beveridgean and Bismarckian, respectively—their postwar pension systems shared a remarkable similarity in that private-sector workers had access only to basic public pensions that were supplemented by private occupational schemes. This institutional design stood in sharp contrast to that of the German and Belgian systems, where statutory social insurance was much more comprehensive. In Britain, workers were covered by a contributory “basic state pension,” introduced through the National Insurance Act 1946. However, this flat-rate benefit was kept deliberately low and in fact barely reached subsistence level. In France, in contrast to public-sector employees, who were given access to generous final-salary statutory schemes, private-sector employees were entitled from 1945 to a mere 50 percent of past earnings, calculated only to a social security ceiling that roughly equaled the average gross wage.
In both countries, it was generally expected that workers would maintain their level of income in retirement through additional private provision. Occupational pensions expanded considerably in Britain in the decades following the Second World War; about half the workforce was covered by the mid-1960s. 34 Nevertheless, the voluntary and very decentralized—primarily company-based—character of such schemes meant that the British population was split into “two nations in old age.” 35 Whereas half of workers, overwhelmingly skilled ones, could benefit from a generous defined-benefit employer-provided plan, the other half, usually less skilled, were bound to face poverty in old age. Following almost two decades of political controversy, 36 including within organized labor, 37 a cross-party consensus was reached in 1975 over the creation of a “state earnings-related pension scheme” (SERPS) to complement the basic state pension for workers not covered by a more generous company scheme.
In France, occupational pension provision also played an essential role but developed through more centralized collective agreements than in Britain. As early as 1947, organized business and the CGC, a union representing white-collar managerial and professional staff (called “cadres”), signed a collective agreement that established a nationwide supplementary scheme—the AGIRC—for this category of employees. Companies in the metalworking and mining industries played a central role in the creation of this institution. 38 The AGIRC indeed continued and expanded schemes that these companies had set up for their engineers in the 1930s. But because those plans’ financial assets were wiped out during the Second World War, the AGIRC was financed on a “pay-as-you-go” basis, that is, through direct transfers between current workers and current pensioners. By contrast, blue-collar workers began to be covered by similar pay-as-you-go occupational schemes, after 1955 negotiations at carmaker Renault set the pace for further collective agreements, including the first industry-level one in 1957 in the metalworking industry. In 1961, organized business and labor founded a federation—called the ARRCO—to coordinate these various blue-collar schemes. In 1972, the state made coverage of AGIRC and ARRCO mandatory, but both schemes continued to be fully managed by the social partners and to incorporate a significant element of voluntarism; companies were allowed to choose between a minimum and a maximum contribution rate and, indirectly, benefit level, giving rise to significant social protection dualisms.
Compared with Britain and France, the German and Belgian postwar pension systems relied on much more generous statutory social insurance. Following a 1957 pension reform, German private-sector employees with an average wage and a full contributory record could count on a replacement pension of about 70 percent of their last salaries. 39 By contrast, after a 1967 reform, Belgian wage-earners were offered 60 percent of their average past earnings if they were single or lived in dual-earner households and 75 percent if they headed single-earner households. 40 Yet firm-level occupational schemes continued to exist, as in both countries contributions and pensionable earnings were calculated only up to a ceiling. Coverage reached as many as two-thirds of employees in Germany in the mid-1970s, but fell to half by 1990 and one-third by the early 2000s, after the public system had matured. 41 In Belgium, no more than about a quarter of private-sector personnel had access to employer-provided plans by the beginning of the 1980s. 42 In both countries, white-collar employees working in large (manufacturing, financial, or utility) companies were most likely to be covered by such plans.
Over the few past decades, all four systems have been considerably reconfigured and have seen the role of private pension provision increase. 43 The rest of this section presents a stylized account of trade union involvement in these changes. Each subsection starts by outlining how pensions in a country have been increasingly subjected to political campaigns for partial privatization, then describes organized labor’s initial opposition to those pressures. Finally, it shows how, in the face of creeping pension privatization, unions have decided to unite to extend coverage of supplementary pensions and reduce the magnitude of social protection dualisms.
Britain
In the early 1980s, Conservative politicians decided to shake the foundations of Britain’s postwar pension system. The first Thatcher government launched a first and relatively invisible wave of retrenchment by shifting the indexation of the “basic state pension” from wage inflation to generally lower price inflation. 44 However, in 1983, the Centre for Policy Studies, a free-market think tank, put forward major changes in the area of supplementary pensions. It considered that existing employer-provided defined-benefit schemes inflicted a “grave injustice . . . on those who change jobs” 45 and it suggested workers be allowed to opt out of such plans and to make their own provision through defined-contribution 46 “personal and portable” pensions. The second Thatcher government set up a committee to investigate this possibility. Among its five members were two private insurers, one of whom—the chairman of Hambro Life, a company that sold primarily defined-contribution plans—described himself as “commercial and not ashamed of it.” 47 In 1985, the Conservative government suggested it might even abolish the “state earnings-related pension scheme” (SERPS) and fully replace it with compulsory private pensions. 48 Nevertheless, after facing strong political opposition, it decided in 1986 first to cut the generosity of SERPS and, second, to offer workers strong incentives to opt out of SERPS or of their occupational scheme and to save in a personal plan provided by a financial institution.
Organized labor fiercely resisted the Conservatives’ proposals. It was skeptical from the outset about personal pensions and argued that financial institutions’ hard-sell tactics might lead individuals to make the wrong decisions. 49 The Trades Union Congress (TUC) campaigned actively against the abolition of SERPS and against changes to occupational pensions. It did so together with the Labour Party, but also with the UK’s two largest employers’ associations, the Confederation of British Industry (CBI) and the Engineering Employers’ Federation. 50 Yet it was unable to stop the trend toward pension privatization. In the 1990s, the Labour Party went back on its promise to restore the generosity of SERPS and decided to promote the expansion of private defined-contribution pensions. The main initiator of this change of strategy was Labour MP Frank Field, who campaigned for his party to “start thinking the unthinkable” 51 and make it compulsory for employees to save in private schemes. 52 Since support for public supplementary provision was still strong among the party’s rank-and-file and organized labor, Tony Blair’s first government decided to target SERPS (renamed “state second pension”) at low-income earners and to create voluntary individual “stakeholder pensions” for middle-income earners without existing private provision. 53
At the beginning of the 2000s, however, Britain’s looming “pensions crisis” 54 prompted organized labor to change its stance and to unanimously propose the expansion of private supplementary schemes to the whole workforce. A spate of high-profile closures of generous defined-benefit plans highlighted the steady decline in coverage of traditional occupational schemes. 55 After accounting standards were modified and regulatory changes forced employers to improve the portability of workers’ pension rights, occupational provision became increasingly expensive for employers and lost much of its effectiveness as a skill-retention tool. Unable to block the closure of company schemes through the industrial relations system, trades unions started to press for legislative changes.
Amicus, a union of high-skilled workers who were among the most strongly affected by the closures, played a key role. In early 2002, the union—formed by a merger of crafts/occupational unions AEEU (Amalgamated Engineering and Electrical Union) and MSF (Manufacturing, Science, and Finance)—started denouncing a “great pension robbery” 56 and encouraged the TUC to embark on a great crusade to introduce compulsory employer contributions for private schemes. 57 At the time, Amicus allied with general unions GMB and TGWU, which also represented less skilled segments of labor, as well as with public-sector union Unison to try to force compulsion onto Labour’s official policy agenda. 58 The TUC warned that, if the party did not commit to address this issue in its 2005 manifesto, it would “pay a heavy electoral price.” 59
Organized labor could count on the support of lobbies such as the National Association of Pension Funds and the Association of British Insurers; 60 the greatest stumbling block was employers. Since Britain’s largest employers’ association, the CBI, considered that compulsion was “all about punishment,” 61 the Blair government fretted about being attacked by the Conservatives for trying to introduce another “stealth tax.” The deadlock was broken after the 2005 general election, when the Engineering Employers’ Federation (EEF)—whose members were Amicus’s main partner in collective bargaining and had still relatively high coverage of occupational plans—broke ranks by announcing that voluntarism no longer worked and that a level-playing field was necessary. The TUC welcomed the announcement and said that “no longer can other employer organisations pretend that business is united.” 62
The EEF’s move allowed Labour and the Conservatives to work out a cross-party agreement 63 over a form of “soft compulsion” through the “automatic enrolment” of employees into workplace pensions. Moreover, to ensure adequate overall benefits and prevent the emergence of new dualisms in the quality of supplementary provision between large companies and small and medium-sized enterprises (SMEs), the Pensions Acts 2007 and 2008 significantly increased the generosity of the “basic state pension” and created a low-cost semipublic supplementary pension provider specifically aimed for SMEs, which was eventually named the “National Employment Savings Trust.” 64 It is also noteworthy that the collaboration between high-skilled workers’ union Amicus and general union TGWU resulted in their merger into Unite—Britain’s largest private-sector union—in 2007.
France
Political pressures for the introduction of private retirement savings became apparent in France in the late 1970s; and, as in Britain, they focused to a large extent on preexisting occupational pensions. The insurance industry was instrumental in bringing pension privatization on the public agenda. For instance, France’s largest state-owned insurance company—the UAP (Union des Assurances de Paris)—launched an aggressive advertising campaign that targeted the AGIRC and ARRCO occupational schemes and used the following slogan: “Babies born in 1949, don’t count too much on babies born in 1979 to pay for your pensions.” Simultaneously, the Geneva Association, a think tank set up by the global insurance industry, 65 financed a study that called for the expansion of fully funded—as opposed to only pay-as-you-go—pensions and received considerable public attention after it was published as a book. 66 In 1987, at the time “personal pensions” were being introduced in Britain, the right-wing Chirac government created voluntary individual retirement savings accounts (PER, plans d’épargne pour la retraite) that offered tax deductions for investments made in bank accounts, life insurance contracts, equities, and bonds.
Trade unions, which were involved in the management of the AGIRC and the ARRCO, viewed these developments with great suspicion. The two occupational schemes were indeed facing serious problems at that time, largely because of their unusual institutional characteristics. Whereas most occupational schemes around the world cover their liabilities with at least a partial accumulation of financial assets, 67 the French schemes worked on a fully pay-as-you-go basis. Moreover, despite the compulsory character of their coverage, the schemes allowed companies to set their contribution rates—and consequently the level of pension entitlements they would offer 68 —on a voluntary basis, between a minimum and a maximum. Since large manufacturing firms, such as Renault, traditionally paid the maximum rates, corporate downsizing and deindustrialization in the 1980s contributed to a rapid deterioration in the proportion between contributors and pensioners in many of the more generous companies. As they were well aware of these imbalances, insurance companies were keen on emphasizing them. 69 With time, any suggestion by insurers that—statutory and occupational—pay-as-you-go pensions should be cut increasingly infuriated organized labor. 70 In fact, trade unions played a key role in persuading a socialist government to repeal the PER retirement savings accounts. 71
As preexisting occupational pensions were subject to strong financial and political pressures, organized labor started pressing employers at the end of the 1980s to expand the AGIRC and ARRCO schemes’ contributory base by compelling all companies to pay the maximum contribution rates. The immediate aim was to shore up the schemes’ finances, but unions’ demands could also suppress existing dualisms in the system by improving the situation of workers covered by less generous schemes—often in SMEs or in sectors such as retail, hospitality, and textiles. Organized labor managed to implement its strategy in three stages. A first impetus was given in 1987 when a few occupational schemes for salaried senior executives, which were still voluntary, were made fully mandatory and were integrated into white-collar employees’ AGIRC. Second, in the winter of 1992–93, unions negotiated a collective agreement with employers that introduced a gradual harmonization of contribution rates—and, hence, benefits—within blue-collar workers’ ARRCO. Finally, in 1994, a similar agreement was reached within the AGIRC.
Although organized labor was united behind this strategy, the unions that had contributed to create the pay-as-you-go occupational schemes were crucial in making it successful. Force ouvrière (FO), the union that had negotiated the collective agreements signed at Renault and in the metalworking industry in the 1950s, and that traditionally presided over the ARRCO, argued that “in the name of equality and solidarity, one cannot refuse the right of textile workers to benefit from higher pensions.” 72 By contrast, even though—as suggested by Häusermann 73 —its white-collar members could have benefited from the private plans promoted by the insurance industry, the CGC—which historically chaired the AGIRC—insisted that organized business could “not put the cadres’ scheme in jeopardy by bowing to the mercenary interests of one of its industry associations [i.e., the insurance lobby].” 74
Both unions used their strong position in the governance of the schemes to persuade allied segments of business to support their plans. Although France’s peak employers’ association (the CNPF) prioritized a reduction in the schemes’ generosity to tackle their deficits, the Union of Metallurgical and Mining Industries (UIMM) was more willing to bow to trade unions’ demands. This organization not only helped found the ARRCO and the AGIRC, but also represented the companies that already paid the highest rates and hence could benefit from an across-the-board increase in pension contributions. Despite the vehement opposition of commercial insurers and many SMEs, 75 the UIMM, which happened to be the most powerful federation within the CNPF, induced organized business to accept a gradual harmonization of—and for many companies an increase in—contribution rates within the pay-as-you-go occupational schemes. In concession, unions accepted some cuts in the generosity of pensions. This policy of give-and-take between unions and employers was continued throughout the 1990s and eventually resulted in the integration of the ARRCO and the AGIRC into the European Union’s system of social security coordination, thus making them quasi-public schemes. 76
In sum, both in France and in Britain, where statutory pensions were traditionally relatively low, increased political pressures for the expansion of private retirement savings focused primarily on preexisting supplementary pension provision. As these schemes started to face financial difficulties, and new types of pension products from the financial services industry threatened, trade unions representing skilled and less skilled workers allied with each other and with the industrial segments of business to preserve preexisting occupational plans. They did so by extending them to the whole of the workforce, thereby also reducing inherited dualisms in coverage of private pensions. The difference between the two countries lay in the means organized labor use in pursuing its strategy: whereas in Britain’s voluntary and very decentralized system of industrial relations unions had to resort to legislative intervention, French unions could still achieve their aims using relatively centralized collective bargaining. We now turn to two country cases in which statutory social insurance was historically much more generous and occupational pensions played a lesser role.
Germany
Unfavorable demographic development, due to relatively low fertility rates since the 1960s and increased longevity, coupled with adverse labor market developments and the huge costs placed on the statutory pension insurance as a result of German unification, provided the socioeconomic underpinning for pension reforms in the late 1990s. High unemployment in the Eastern parts of Germany and the transfer of the West German pension system to the East led to a significant increase in social insurance contributions, which was said to undermine the competitiveness of German companies in an ever increasing global economy. A debate on the competitiveness of Germany as a business location (Standortdebatte) raged for most of the 1990s. 77 Political parties eventually “agreed” that the contributions to the old-age social insurance scheme should be capped. The CDU/CSU (Christian Democratic Union/Christian Social Union) proclaimed in their 1998 election platform: “The costs levied on work are too high in Germany. We will continue to comprehensively reform our social security system toward enhancing personal responsibility and private arrangements as well as strengthening efficiency.” 78
The rationale stated by the Social Democrats (SPD) for the 2001 pension reform, which included partial privatization, was to share the burden of demographic change equally among generations as well as to encourage more personal responsibility. 79 Although it argued for expansion of the old-age social insurance to certain groups, such as the self employed, the Social Democrats’ 1998 election campaign manifesto also highlighted the need to expand fully funded private and occupational schemes in the future. 80 Walter Riester, SPD minister for labor and social affairs (until 2002), and architect of the comprehensive pension reform of 2001, stated in parliament: “As necessary and as painful as it was in the past to indicate that the statutory pension system alone can no longer guarantee the achieved living standard . . . , we can declare today that those who participate [in private or occupational plans] will have a significantly higher overall old-age income.” 81
Thus the pension debate did not occur in a political vacuum, but was embedded in a broader debate about economic competitiveness. In addition to employers and employer associations, 82 the financial services industry was key in pushing for the privatization of pension provision since the mid- to late 1990s. As highlighted by Hockerts, 83 the Federal Association of German Investment Companies (Bundesverband Deutscher Investmentgesellschaften) as well as DB Research (subsidiary of Deutsche Bank) were crucial for the reform of 2001. 84 Among other instruments the financial services industry used media campaigns very effectively.
Labor unions, with the exception of the traditionally more accommodating Union for Mining and the Chemical and Utilities Sectors (IG BCE), were opposed to any comprehensive reform of the statutory pension system. 85 Although not in favor of private individual pension plans to “compensate” for the planned cutbacks in the public scheme, they supported the overall thrust of expanding voluntary occupational old-age provision and proposed that it should be integrated more comprehensively into collective bargaining agreements. 86 Even before the 2001 reform, the IG BCE negotiated a collective occupational pension agreement covering all workers in the sector.
Once the powerful Metal Sector Union (IG Metall) conceded that unions could not stop comprehensive pension reform, it pushed for collective and encompassing occupational pension agreements at the sector level. IG Metall signed a 2001 industrywide collective agreement, similar to the collective pension agreement covering workers in the chemical industry, creating the Metall Rente. The scheme’s aim was to offer the possibility of occupational pension coverage to all workers, especially for workers in SMEs that in the past were unable to set up their own schemes. Indicating his union’s continued ambition to set the pace for other industries, the president of the IG Metall, Klaus Zwickel, declared that the collective agreement constituted “a qualitative leap into the future for collective bargaining. We are shaping the future with it.” 87 The union added that: “The conditions are much more attractive than those offered by banks and insurance companies. With the Metall Rente, IG Metall has achieved more fairness and justice for workers in small and medium-sized companies, who so far largely had no access to complementary occupational pension arrangements.” Despite the increased possibility of opt-outs for struggling companies in collective bargaining agreements, the collective agreement governing occupational pensions allows opt-outs only if they are more generous for the workers than the Metall Rente. 88
Belgium
In Belgium, both financial lobbies and employers’ associations started pushing for pension reform in the early 1980s, but, as in Germany, financiers’ efforts to promote pension privatization targeted the relatively generous statutory pension system, rather than preexisting supplementary schemes, as was the case in Britain and France. Whereas employers—including the powerful export-oriented manufacturers in Belgium’s small open economy 89 —were concerned about reducing the impact of public pensions on rising nonwage labor costs, insurance companies emphasized the necessary “complementarity” between public and private provision and called for a “three pillar” system in which both occupational schemes and individual retirement savings accounts would play an expanded role. 90 A Catholic-Liberal government legislated the creation of voluntary personal pensions in 1986. In addition, over the 1980s and 1990s, Catholic-Liberal and Catholic-Socialist coalition governments gradually reduced the generosity of public pensions, in part through a relatively invisible process of institutional change: by freezing the ceiling on pensionable earnings, they slowly transformed Belgian statutory pensions into a basic benefit that failed to maintain the living standards of high-income and middle-income earners. 91
Trade union confederations of all stripes—such as the Christian-democratic ACV/CSC, the socialist ABVV/FGTB and the liberal ACLVB/CGSLB 92 —tried unsuccessfully to resist these developments. 93 They were also initially very skeptical about increasing the role of occupational schemes, which were traditionally fully controlled by employers who used them to retain their most valued workers, generally white-collar and high-skilled, and thereby dualized the workforce. Unions demanded that occupational pensions be subject to collective bargaining and that they be granted a right of involvement in their administration. Moreover, they wanted government regulations to improve the portability of accrued rights and to limit employers’ capacity to require that workers complete a minimum number of years of service in order to become eligible for a pension. Government actors held discussions on these issues with the social partners from the late 1980s; but because of employers’ resistance, legislative changes that met unions’ demands were introduced only in 1995. 94 Soon afterward, people close to organized labor—such as Greta D’Hondt, a Flemish Christian-Democratic MP and former general-secretary of the Christian ACV/CSC union—floated the idea of expanding occupational pensions, provided that they were “collectively generalized to all workers,” that is, to blue-collar and less skilled workers. 95
As in Germany, the metalworking industry provided a powerful stimulus to this conception. In 1999, the metalworkers’ unions and employers’ association Fabrimetal (renamed Agoria in 2000) signed a collective agreement creating a mandatory sectorwide defined-contribution plan for all the industry’s blue-collar workers, only about 15 percent of whom were then covered by a company pension scheme. 96 Not only was the pension fund to replace workers’ wages; the plan also contained “solidarity clauses,” as a result of which employees would receive pension contributions during periods of inactivity, such as sickness, temporary unemployment, or maternity. The plan was to be coadministered on a not-for-profit basis by employers and unions. In fact, rather than negotiating this scheme just for themselves, both the metal lobby and the metalworkers’ unions intended to set the pace for further industry-level agreements that could eventually result in a generalization of occupational pensions. 97 They also hoped that negotiations on supplementary schemes could help revitalize the Belgian system of collective bargaining at a time when the state increasingly intervened in the determination of wages. 98
Nevertheless, whereas the German metalworking industry created its sectorwide scheme only after relevant legislation was introduced, in Belgium the opposite was true. The Belgian metalworking industry’s collective agreement coincided with the 1999 negotiations to form a new federal government. Trade unions and Fabrimetal approached some of the negotiators to request changes in existing regulations for occupational plans so as to implement their specific conception, and a new Socialist minister of social affairs, Frank Vandenbroucke, pledged to respond to their demands. The Vandenbroucke law on supplementary pensions, eventually enacted in 2003, introduced important incentives for the expansion of encompassing and collectively negotiated occupational schemes. The government agreed to offer specific tax advantages to occupational plans that contained “solidarity clauses” such as those negotiated in the metalworking industry. 99 Individual companies were allowed to opt out of industrywide schemes only if they offered a more generous plan. To stimulate negotiations on the issue, the government agreed to exclude contributions to occupational plans from a state-imposed wage norm. In 2001, it also persuaded all peak-level trade union confederations and employers’ associations—which play an important role in Belgium’s centralized system of collective bargaining—to issue a declaration calling on the industry-level social partners to help “democratize” supplementary pensions by “introducing industrywide social pension plans in as many sectors as possible.” 100
To summarize, in both Belgium and Germany, campaigns for pension privatization focused their attention especially on the need to cut the generosity of public pensions, since these traditionally played a dominant role in workers’ retirement provision. However, as governments were increasingly committed to expanding private retirement savings through either individual or occupational plans, and organized labor was unable to block that development, both Belgian and German metalworkers’ unions decided to promote the expansion of collectively negotiated industrywide schemes that could cover workers whatever their skill level and whatever the size of their company. That strategy did not lead to growing political divisions with other segments of labor but, instead, set the pace for similar agreements in other sectors. Nevertheless, a major difference between the two countries is that German metalworkers started actively promoting their strategy only after the government had already set out its own vision, while in Belgium they were the ones who brought their conception to the government’s agenda and were therefore able to shape legislation in a more favorable way.
Conclusion
In this article, we have shown that the recent rise of private retirement savings in Western Europe has pushed different segments of organized labor to increase cooperation with one another to extend coverage of supplementary pensions to as many workers as possible. As unions have proved incapable of stopping the trend toward retrenchment and privatization of pensions, they have tried to mitigate its consequences and to avoid dualization, either by making coverage of preexisting supplementary pension plans compulsory, where such schemes already played a significant role in pension provision during the postwar period, or, where public provision was traditionally dominant, by negotiating the creation of encompassing sectorwide schemes, in the hope that they would eventually cover all industries. We have illustrated the first pattern with case studies of Britain and France and the second one with an analysis of reforms in Germany and Belgium, and the literature suggests that similar developments in the pensions policy area have occurred in other countries. For example, Green-Pedersen has shown that, in Denmark, where workers traditionally received a basic noncontributory “national pension,” the metalworkers’ union and its counterpart on the employers’ side set the pattern in 1991 for the introduction of industrywide schemes that now cover almost the whole Danish workforce. 101
Unions’ increased cooperation in response to pension privatization seems to be an exception to the growing conflicts and dualisms within labor to which recent studies of institutional change in affluent market economies have called attention. By highlighting this exception, we seek to contribute to the comparative political economy literature. We certainly do not mean to say that different segments of labor do not have distinct interests. Indeed, our study of the politics of pension privatization suggests that trade unions representing high-skilled workers or the core industrial sectors of the economy had their own, and not necessarily altruistic, motives in supporting the extension of supplementary pensions to the whole of the workforce. Thus, in countries where occupational pensions already played a large role, such unions were largely motivated by an interest in defending schemes from which their members historically benefited the most. By contrast, in countries with traditionally dominant but shrinking public pensions, the metalworkers unions’ decision to push for the expansion of industry-level collective negotiated schemes can also be seen as driven by their willingness to maintain their role as a pacesetter for the rest of the economy.
Nevertheless, we believe that our analysis has three broader implications for future research on trade union involvement in the politics of institutional change in capitalism. First and most directly, political scientists should be careful not to assume that intra-labor cleavages necessarily translate into trade union conflict and egoism, particularly in a context of economic liberalization and welfare state retrenchment. As highlighted by Thelen and Kume, “a strong opponent is often quite useful for unions needing to forge an external consensus across workers of different skill levels.” 102 In the specific case of pensions, plans by the financial services industry and some politicians to turn retirement provision into a commodity have certainly helped antagonize and unite different parts of the labor union movement. Only a better understanding of the specifics of each policy area as well as further empirical research can show whether similar dynamics have been observed beyond the field of old-age pensions.
Second, when analyzing the role that workers—and employers—in the core manufacturing industries play in current processes of institutional change, researchers should distinguish much more explicitly between the interests and attitudes of actors at the plant level and those of the industry-level associations that represent them. Palier and Thelen suggest that dualistic pressures in industrial relations have come primarily from the increased delegation of issues to company-based or plant-level bargaining. 103 However, their analysis is less clear about whether trade union and employer federations in the metalworking industries have willingly or unwillingly fostered this “inward turn.” Our evidence reveals that they have not lost their traditional willingness to set the pace for the rest of the economy. As industry-level associations, they may have their own organizational interest in the survival of more centralized forms of collective bargaining. 104 This is a question that should be investigated in the future.
Finally, and related to the previous point, researchers should also distinguish between the intentions of organized interests and their capacity to accomplish their aims. Although trade unions may see collectively negotiated schemes as a way of bringing about “social security through industrial agreements,” 105 do they still have the means to achieve this goal? Here, political alliances and the role of the state appear to be of utmost importance. In the four cases analyzed in this article, unions’ alliances with left-wing parties and employers from the manufacturing industry were crucial in helping them to push their strategy both in the political arena and in the industrial relations system. Moreover, as has already been highlighted by Martin and Thelen, 106 collective bargaining’s capacity to achieve egalitarian aims hinges on the arsenal that the state can use to support coordination. The history of the French system of pay-as-you-go occupational pensions shows how important the state’s role is in extending coverage of these schemes to the whole workforce. Whereas in Belgium the newly negotiated industry-level collective agreements on supplementary pensions have been extended erga omnes, it is unclear whether the lack of such state support in Germany, and the principle of collective bargaining autonomy, will allow German unions to achieve satisfying coverage.
Footnotes
Acknowledgements
For comments on previous versions of this manuscript, we would like to thank Christian de Visscher, the editorial board of Politics & Society, and the participants of the Twentieth International Conference of Europeanists (Amsterdam, June 2013), the ISPOLE general seminar (Louvain-la-Neuve, September 2013), and the Twelfth Annual ESPAnet Conference (Oslo, September 2014). Marek Naczyk is also grateful for Cathie Jo Martin’s and David Rueda’s comments on his doctoral thesis, which has provided much of the material for the present article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.