MPIfG Working Paper 01/9, December 2001

Institutional Change in the German Wage Bargaining System –
The Role of Big Companies

by Anke Hassel and Britta Rehder , Max Planck Institute for the Study of Societies

Abstract

Despite the emergence of new production systems, Europeanization and economic internationalization, the national arrangements of wage bargaining systems have not been eroded. The paper highlights the factors that counteract the pressures for a straightforward decentralization with reference to Germany. Firstly, the maintenance of peaceful labour relations is a major advantage for big companies in centralized wage bargaining systems. Secondly, big companies have been able to achieve labour cost control and the differentiation of working conditions by drawing up pacts for employment and competitiveness at the company level. Thirdly, they have succeeded in introducing a higher degree of flexibility into the collective agreement framework. The main argument will be that institutional change has taken other forms than expected. Whereas the formal institutional setting has remained relatively stable, the functioning of the German wage bargaining system has changed.

Zusammenfassung

Trotz der Verbreitung neuer Produktionssysteme, der Europäisierung und der wirtschaftlichen Internationalisierung sind nationale Tarifverhandlungssysteme in den meisten Ländern stabil geblieben. Dieses Working Paper hebt am Beispiel Deutschlands die Faktoren hervor, die dem Dezentralisierungsdruck entgegen wirken. Für Großunternehmen ist auch unter veränderten Bedingungen die Vermeidung von Konflikten einer der Hauptvorteile zentraler Tarifverhandlungssysteme. In dem gegebenen institutionellen Rahmen haben Unternehmen Möglichkeiten gefunden, über Standortsicherungsvereinbarungen auf Unternehmensebene die Arbeitsorganisation zu flexibilisieren und Arbeitskosten zu senken. Zudem konnten sie Tarifverträge flexibler gestalten und Öffnungsmöglichkeiten einfügen. Als Ergebnis wird festgehalten, dass institutionelle Veränderungen andere Formen angenommen haben als erwartet. Während die formalen institutionellen Rahmenbedingungen verhältnismäßig stabil geblieben sind, hat sich die Funktionsweise des deutschen Tarifverhandlungssystems verändert.

Contents

1Introduction

Since the early 1980s, a fast-changing political and economic environment has led observers to expect a stronger tendency towards the decentralization of wage bargaining institutions. Several reasons have been given to support this belief (Wallerstein and Golden 1997; Wallerstein, Golden et al. 1997). Firstly it was thought that more flexible production systems would require greater differentiation of pay and a stronger connection between individual performance and rewards (Katz 1993; Streeck 1993; Iversen 1996; Pontusson 1996). This would give companies a greater interest in locally designed pay systems as opposed to central wage agreements. Secondly, the fragmentation of trade union membership was expected to reduce the capacity of trade unions to act collectively. New groups on the labour market were undermining the dominant position of blue-collar workers. Less cohesion on the side of labour would reduce the ability of peak associations to obtain labour peace or wage moderation in centralized bargaining systems. Central wage bargaining institutions would become dysfunctional (Calmfors 1993; Moene, Wallerstein et al. 1993; Lange, Wallerstein et al. 1995). Thirdly, to the extent that centralized bargaining systems depend on government support, the decline of discretionary macroeconomic policy-making at the national level might reduce governments' interest in supporting centralized collective bargaining systems (Streeck 1992).

The evidence has to some extent been different to the expectations. Although a process of increased decentralization can be observed in several countries, this has not led to the breakdown of centralized wage bargaining systems. On an aggregate level, comparative studies have pointed to a relatively high degree of stability on the part of industrial relations institution (OECD 1997; Wallerstein, Golden et al. 1997). Where decentralization did occur, as in the case of Sweden and Denmark, it moved from the national to the sectoral level (Iversen 1996). In those countries where sectoral wage bargaining institutions were already in place, they generally remained in place (Traxler 1998). In some cases, such as Ireland, Italy and Spain, new central agreements even led to a reorganization of wage bargaining institutions at a more central level (O'Donnell and O'Reardon 2000; Perez 2000).

There is therefore some reason to believe that there are either strong factors supporting a system of central wage bargaining which qualify the pressures for decentralization or that, alternatively, the new challenges such as Europeanization, new production systems and economic internationalization impact on national bargaining systems in other ways than the compulsion to straightforward erosion. In other words, the institutional change might take other forms than expected.

Hence it is our aim to contribute to the emerging literature on the defining pathways and mechanisms of incremental institutional change, in particular to those approaches that look at institutional conversion, "as institutions designed with one set of goals in mind are turned to other ends" (Thelen forthcoming, 29). In the literature, institutions are almost by definition stable, since structuring devices need a certain time horizon to be effective (Genschel 1997, 47). Therefore, institutional change that falls short of fundamental shifts is difficult to conceptualize. Since adjustment pressure in a fast-changing economic environment is nevertheless strong, institutions can be expected to change their functioning without necessarily changing their form.

1.1The German Case

With regard to the transformation of wage bargaining systems, the German case seems to be especially interesting for mainly three reasons. First, the German model of industrial relations has impressed for decades by its stability and its performance. Second, during the 1990s the institutional arrangement had to meet extraordinary requirements by encompassing the East German regions. No other West European system of industrial relations had to cope with a task similar to this. Third, the German case is demonstrative of a process of institutional change as described above: transformation without breakdown. Wage bargaining institutions have remained relatively stable, conditioned by a slow, but steady, trend of shrinking coverage of collective agreements and works council representation (Hassel 1999). Moreover, unification has transferred these institutions in almost the exact form from the west to the east, although the pressure on the system has been high since the early 1990s. The Council of Economic Experts has criticized the German wage formation system continuously and demanded decentralization (Sachverständigenrat 1997). Employers' confederations have consistently complained about rigid and high wages.

Despite the criticism and complaints, very few major companies have resigned from central wage bargaining. Employers present themselves as weak and not in the position to withstand an industrial dispute. The unions have won the majority of industrial disputes, which leads some observers to conclude that the continuity of the German system is based on trade union strength (Turner 1998; Wallerstein 1998). The alternative view argues that, rather than union strength, it is the employers' weakness which maintains the system. Employers' associations are divided and unsure about the effects of decentralization and are therefore unable to reach a political compromise among themselves on how to proceed (Thelen 2000).

In this paper we would like to propose another interpretation of the German case, which is based on the study of the behaviour of big companies in Germany since the mid-1980s.[1] We will put forward three main arguments, which make up the remaining part of the paper. In the first part, we will argue that, under regulations of the German wage bargaining system, there are major advantages for big business to retain centralized institutional arrangements. In particular, large employers have a strong interest in maintaining the peace-keeping function of central industry-level agreements. For this reason our analysis focuses on the role of big companies in the ongoing development of the German wage bargaining institutions.

Instead of resigning from the system, and this will be our second argument, they have preferred - and, indeed, been able - to achieve a new degree of flexibility and cost-cutting within the system since the early 1990s. Large employers have strategically been able to increase their room for manoeuvre at the plant level. This has contributed to the stabilization of the formal institutions as the increased internal flexibility has prevented the employers from resigning from the centralized system.

Yet that does not mean that nothing has changed. In the last section, we will describe the type of regulatory change that evolved in the wage bargaining system instead of a straightforward collapse. Here we will argue that the emphasis on regulating working conditions has shifted from providing uniform regulations aimed at a high level of equality between companies to the introduction of the flexibility and hardship clauses which aim to ensure the survival of business operations under uncertain conditions. This process has implied a redefinition of the interaction between production model and labour market regulation. The model of diversified quality production was originally based on the German production regime of providing high wages and equal working conditions. However, the 1990s brought about the reversal of this interaction as the wage bargaining regime had to be adjusted to the needs of production.

2Employers' Interests in Centralized Wage Bargaining Systems[2]

We generally assume that stable institutional arrangements are based on coalitions of actors who obtain economic or political gains from them. Centralized wage bargaining systems are based on a coalition between large employers and trade unions. Thus we claim that not only the unions but big business as well will enjoy beneficial effects from bargaining centrally rather than locally.

For employers a centralized wage bargaining system has two major implications. The negative implication is that central wage agreements restrict the scope of local flexibility. Any agreement found at the central level has to cover a whole range of different companies. It therefore has to set certain standards in relation to wage patterns and working hours which does not necessarily suit an individual employer and his or her firm. The individual employer can only deviate marginally from pay schemes, usually by adding voluntary bonus schemes. He or she might have been able to negotiate a more suitable wage pattern at the plant level. The agreement might also entail regulations on the implementation of personnel policy that the individual employer would not have agreed to at the plant level. It also forestalls any swift changes in wage patterns and other regulations. It therefore slows down and prohibits certain personnel policies that the employers would have otherwise preferred.

It should be noted that the interest in local flexibility is not limited to the employers but impinges on the needs of the workforce as well. Interests which are to be realized at the centralized level have to be capable of mobilizing solidarity among all the main groups of union members (Streeck 1981: 156). While unions might call a strike for general wage increases, they would not be able to do so for the demands of the workforce of a specific company. Therefore employees need different channels of interest representation. The most important sectional interest of workers seems to be the economic survival and the competitiveness of the firm as the most important precondition for job security. With regard to this issue, we can assume that works councils tend to share the perspective of the management where the preservation of jobs is concerned. The emergence of plant-level pacts for employment and competitiveness in the 1990s supports this assumption (section 2).

Hence, the interest in local flexibility and the restrictions on the latter in the context of centralized wage bargaining systems does not reflect a class conflict but a multi-level conflict between the centralized and the plant-level actors.

The second implication of a centralized wage bargaining system is its peace-keeping function. By transferring the distributional conflict between management and the workforce to a level above that of the firm, employers in general gain a greater degree of peace within the company. Bargaining rounds take place between employers' associations and trade unions. Employers can maintain a good relationship with their workforce by staying removed from the process. Moreover, by being a members of a bigger association, employers are not as easily singled out for trade union activity and industrial action. Lastly, in many European countries, the employers' associations even provide an insurance against losses in an industrial dispute by maintaining strike funds. Nevertheless, centralized wage bargaining only remains a precious advantage for employers as long as trade unions are able to introduce conflicts in wage bargaining. If trade unions could not upset business interests in preserving industrial peace, the employers' interest in maintaining the system would be considerably lower.

The German case illustrates the interplay of peace-keeping regulations and the capacity of the unions to challenge them. On the one hand, the system imposes strict regulations on peace keeping and industrial action. Firstly, the right to call a strike is limited to the unions, while works councils are obliged to respect social peace. Secondly, industrial action has to be limited to collective bargaining issues (instead of wider political goals). And thirdly, legislation defines strikes as the last resort when any other mechanism of conflict mediation has failed. As a result, the peace-keeping potential of the German system is great.

On the other hand, German trade unions are, without a doubt, able to introduce conflicts. They have even managed to turn the majority of industrial disputes in their favour (Turner 1998; Wallerstein 1998). In this sense, Wallerstein and Turner are right in claiming that the continuity of the German system is based on trade union strength. Owing to the effective monopoly of interest representation, German trade unions are able to control industrial action without having to take into account competing organizations. In addition, the leadership of the metal sector union and its capacity to jeopardize social peace is of importance. One example is the wage conflict in 1995. Under the impact of the recession and the aftermath of unification, the metal sector employers' association adopted a particularly hard line by demanding that all wage increases should be paid for by new flexibility concessions. However, the union staged industrial action and won the conflict (Thelen 2000). After a two-week strike in February 1995, the employers conceded an exceptionally high wage increase.

Both implications of centralized wage bargaining systems - the peace-keeping function as well as the restriction of local flexibility - can be considered as a bargaining space defined by a class conflict and a multi-level conflict. They create the arena for negotiations between employers and trade unions as well as between different groups of employers.

The trade-off between the advantage of having industrial peace and the costs of being restricted in managerial discretion is determined to a considerable extent by the size of the company.[3] Company size affects both dimensions of the bargaining space. Big companies are likely to be targeted by trade union activities and give above-average pay awards. Therefore, they tend to benefit from participating in centralized wage bargaining systems that can offer them industrial peace. As big organizations, they are also more likely to have a developed and formalized pay system themselves. Wage agreements therefore do not necessarily restrict them further in their flexibility. Moreover, since the wages they pay are above average and thus above the rates which are laid down in the agreement, their scope for plant-specific bonuses is greater than in smaller companies.

Small companies, on the other hand, have less need for conflict-avoiding mechanisms since they are not the target of trade unions in the first place. Since they tend to pay less than big companies, they also run the danger of paying higher wages than would otherwise be the case. Moreover, their managerial discretion would probably be substantially higher without collective agreements. This is especially true for those small or medium-sized companies that do not have a works council. In these cases, local flexibility does not have to mean negotiated flexibility. Therefore small companies tend to benefit less from centralized wage bargaining, but run the same costs as big companies.[4] Hence, one can expect that big companies are generally more in favour of centralized bargaining than small companies. The empirical data on the membership density of employers' association bears this out. Small companies are much less likely to participate in collective bargaining than big companies (Hassel 1999). Also, small companies are more likely to voice stronger opposition to centralized collective bargaining systems, while big companies are likely to be in favour of them.

With regard to collective bargaining strategies, one can therefore assume that the bargaining goals will differ between small and big companies. Small companies are, firstly, more likely to go for bold bargaining strategies which would make a difference to all companies (such as a reduction in fringe benefits or holidays, lower wage increases and the like) and, secondly, they are more prone to conflict strategies which can be accommodated much more easily in small paternalist companies than in big companies. On the other hand, big companies are more likely to favour strategies that are conflict-averse and try to introduce a higher degree of flexibility into existing standardized wage norms.

Although employers' associations have to mediate between the diverging employers' interests we can assume that big companies control the process of interest representation. In countries with sectoral employers' associations such as Germany, the latter tend to have a highly uneven distribution of membership companies, with few very large companies and high numbers of very small companies. Membership dues are frequently based on the number of employees (or the aggregate sum of wages a company pays); voting procedures are sometimes attached to membership dues. In any case, they aim at reaching a high level of consensus. The dominant position of big companies in employers' associations is further strengthened by the fact that all employers know that decisive conflicts are fought over in big plants. If the management of big plants is not prepared to persevere with the trade union, the collective bargaining strategy is not viable. As a result, there is no bargaining strategy on the part of the employers against the interests of big companies. As big companies will always veto deterioration in the peace-keeping capacity of the bargaining system, the associations have to pursue a strategy to increase local flexibility without losing the peace-keeping function of the system.

To sum up: at the level of firms' preferences towards wage bargaining institutions there are positive and negative implications for firms when bargaining centrally. Firms might lose managerial discretion while gain in terms of labour peace and labour costs. We can also see that the costs and benefits vary with firm size: small firms benefit less while having similar costs. Big firms, on the other hand, have few if any costs from central agreements and benefit most. Ceteris paribus, there are good reasons why big companies are the main pillar of centralized wage bargaining institutions.

Empirical findings support our theoretical assumptions. During the 1990s, tensions between big and small companies within employers' associations rose. An increasing number of small and medium-sized companies in several sectors complained about their associations' high-wage policies (Völkl 1998: 172), and these firms rather than big firms left the associations during this period (Schröder and Ruppert 1996: 40f).

The next section of this paper will show how and why big companies in Germany have sought to increase local flexibility under the protection of the peace-keeping function of the centralized wage bargaining system since the mid-1980s.

3Plant-Level Bargaining in Big Companies

A number of factors have contributed to the heightening of competitive pressure on German companies since the early 1990s and forced adjustment processes upon them. The most important ones are the consequences of reunification, the recession in 1992/93 and the increasing internationalization of firms.

First of all, reunification was marked by the transfer of the industrial relations system from the west to the east. This political decision was supported by all major West German political actors. The high-wage strategy that had been successful for so many decades in the west was simply transferred to the east, although this meant that the wages had to be raised far beyond the productivity of eastern plants. Neither capital nor labour were interested in a low-wage area. The trade unions were afraid of the erosion of the high-wage regime in the west, whereas the employers wanted to prevent the emergence of a price-competitive production area (Lehmbruch 1994). Although the reunification process was supported by a massive financial transfer from the west to the east it could not prevent rising mass unemployment. As a consequence, public debt and labour costs exploded. Moreover, high wage hikes in the early 1990s were not only relevant to the east; the west also experienced exceptionally large wage gains, with trade unions claiming their share in what they saw as the unification boom.

In 1992/93 the economy was hit by the worst recession in post-war history, which was accompanied by major job losses, especially in the manufacturing sector. The failures of the past became visible. German business had been deprived of its leadership with regard to quality production and innovation of products. Japanese firms in particular had learned how to produce goods that were superior and cheaper than German products. As a consequence, German firms had to learn how to improve on price competitive innovations. Especially the implementation of "lean production" was accompanied by job losses that could not be compensated for by the reduction of working hours and a social policy which had always been previously a social net for the negative effects of the high-wage strategy (Streeck 1997). Consequently, the costs of social security and labour rose even further.

The declining competitiveness of the German economy became even more evident as worldwide production became more feasible. Companies exposed to the world market built up production sites abroad not only to be present at the most important sales markets; they also aimed to benefit from large and high-qualified workforces which they were able to employ for lower and more flexible wages than in Germany. Many companies institutionalized international benchmarking processes that compare production sites continuously in terms of their labour costs and the flexibility of their working conditions (Mueller and Purcell 1992). The plants that come out best are chosen for new investment, whereas others are threatened by closure. Workers who are employed by the same company, but work under different regimes of industrial relations in different countries, compete for investments and job security. The car industry in particular faces this new form of decision-making on investments, as do certain parts of the chemical industry, the tobacco industry and household appliance manufacturing.

The high costs of unification, the recession in 1993 and the internationalization process exposed the German economy and its institutions to new kinds of pressure. Rising unemployment and the declining competitiveness of German companies prompted employers to complain about high and rigid labour costs. As a result, they tried to expand their room to manoeuvre to modify the negative effects of these developments on their firm's performance.

However, big employers did not try to dismantle the centralized bargaining system in order to seek remedies for the competitive pressure they were facing. Rather, they turned to company-level bargaining for a solution. Bargaining at the company level has always played an important role in the German system of industrial relations. Employers have accepted the co-determination system as an element of the corporate governance structure in German companies, as it enables management to enforce social peace and cohesion by integrating the employees' interests into their decision-making process (Kommission Mitbestimmung 1998: 7).

Diversified quality production - based on the continuous improvement of skills, work organization, technology and products - benefits from co-determination as it supports the incorporation of employees into firm-specific coalitions: technological and organizational changes, which are discussed with the works councils and regulated by firm agreements (Betriebsvereinbarungen), are implemented more easily and more quickly. In addition, bargaining at the company level has introduced some kind of wage drift into a wage system that provides a high equality of incomes (Hassel 1999: 486). Central agreements neglect the different capacity of firms to afford a certain wage level. High-performing companies negotiate with their works councils paying "above the going rate" (über Tarif) as well as a system of special premiums to attract qualified employees.

However, since the late 1980s, a new type of plant-level bargaining has emerged which goes far beyond the traditional form of company-level bargaining. Rather than implementing or topping up the terms and conditions of central agreements, company-level pacts for employment and competitiveness (betriebliche Bündnisse zur Beschäftigungs- und Wettbewerbssicherung) have emerged which include a whole bundle of measures to improve competitiveness and job security. The rationale behind the emergence of these pacts is that both groups of actors - management as well as the workforce - suffer from the same fate of lacking flexibility in the regulation of firm-specific needs. In this constellation, management and works councils both share the same interest in increased local flexibility, in order to strengthen the competitiveness of the firm and to secure jobs. Neither side can realize this interest without co-operation. There is no possible way for the employees to force management to grant job security provisions as employment decisions are part of the management's prerogative. On the other hand, employers cannot control the production costs and other aspects of their firm's competitiveness on their own. At the collective bargaining level, they can realize their cost-cutting and flexibility interests only if the unions at the central level are willing to compromise. If the unions resist doing so and if the employers do not want to sacrifice social peace they have to look for solutions at the company level. Here, they have to compromise with their works councils: firstly, due to their co-determination rights and, secondly, with regard to the necessity of legitimacy for cost-cutting measures which can be provided best if the works councils are involved in the decision-making process.[5]

3.1Expanding Local Flexibility by Pacts for Employment and Competitiveness

In order to assess the scope and importance of plant-level pacts, we looked at the emergence of these pacts in the 120 biggest companies in Germany between 1986 and 2000.[6] During the 1990s, 55 (46%) of the companies negotiated a company-level pact. These companies employed 4 million employees in 1996. 1.7 million employees were affected by the pacts. Within these 55 companies, at least 156 agreements can be found.

Other companies, especially in the banking sector, have made use of different types of company-level agreements to pursue similar goals. For example, an increasing number of firms have negotiated the modification of payment schemes towards a higher share of variable wage components (Kurdelbusch 2001). However, our analysis focuses on pacts for employment and competitiveness as they, more than any other type of company agreement, have left clear traces in the institutional design, as we will show in the following sections.

The number of agreements per company varies: While Daimler Benz AG has concluded more than 30 agreements; the majority of companies have only negotiated one or two. However, there is a clear tendency that a successful plant-level agreement on employment and competitiveness has knock-on effects in not only the company but also the same sector. One can identify waves of plant-level pacts in the car industry (in 1993 and 1997) and in the chemical industry (1997 and 2000) (Rehder 2000).[7] 41 (77%) of the pacts are to be found in the manufacturing sector, where competitive pressure hit hardest.

When bargaining over plant-level pacts, three main dimensions of management's interests can be distinguished: to control labour costs, to increase productivity gains by improving the flexibility of the production process and to redistribute work among the employees, which can be another source of flexibility. Table 1 shows the three different types of pacts (Rehder 2001).

53% of the companies have agreed on a labour cost-cutting pact; they gain cost-cutting effects by directly reducing employees' income. Income cuts are realized in many different ways. The most prominent ones are the reduction of the wage drift between sectoral agreement and company payment, as well as the retrenchment of bonuses and premiums. In addition, several bargaining parties agree on paying below the collective agreement standards either for the whole workforce or for special groups of employees.

49% of the firms have negotiated productivity pacts. Most of them have been introduced under the impact of the international competition for investment and the intra-firm benchmarking procedure as described above. They aim at adjusting working conditions to a specific production process, to the life cycle of the product and to the changes in demand. The conditions of capital investment can thereby be optimized. Typical concessions made by the employees are, for example, the extension of working hours (in most cases without wage compensation), measures against absenteeism, changes in work organization, and the changes of pay schemes towards a higher share of variable wage components compared to fixed pay. In return, management offers investment or production orders. Because of the 1992/93 recession, the car manufacturers (and later other industrial companies) have experimented with the organization of production in order to increase productivity (Hancké 1998). These experiments have included the introduction of lean management, teamwork and a three-shift-system in one or several plants (Mueller 2000). Some firms have even partially returned to a modified model of fordist production (Springer 1999). Owing to the debate about "lean management" a range of opportunities for organizing production has opened up. As a result, a process of destandardization has taken place. Productivity pacts have been an instrument used to facilitate these developments in plants, as the reorganization of production always implies the reorganization of work.

42% of the firms have introduced work redistribution pacts. In this context, the flexible reduction of working hours has been the most prominent issue. It increases flexibility by adjusting the employees' working hours to the demands of the market. In addition, these pacts have been used by companies that have had to cut a large number of jobs. Through the reduction of working hours and the resulting redistribution of employment the firm increases its flexibility in cutting labour supply: the intensity as well the speed (and the costs) of the cuts can be shaped more carefully so that neither the production process nor social peace in the company is disturbed. Another prominent way of redistributing work is to increase labour mobility in terms of functional and geographical flexibility. This is usually organized in the context of workforce pools. The management can refrain from redundancies if core employees, whose jobs have been cut, go on to replace those on short-term contracts. External flexibility thus gives way to increased internal flexibility.

3.2The Role of the Unions

In the previous section we aimed to show why and how big companies have succeeded in increasing local flexibility. In this section we argue that they have done so without risking social peace by incorporating the unions into the process.

Hardly any agreement has been negotiated in the past without the unions' participation. In 55% of the firms the concerned union(s) usually joined the bargaining table and signed the agreement. In other cases they did not sign the pact but acted as a veto-player, meaning that the agreements had to be approved by the unions with regard to the legal aspects before the works councils signed them. In 35% of the cases the unions acted as a consultant for the works councils with regard to the legal or strategic aspects. In this way, unions always acted as a veto-player. Only in 10% of the firms were trade unions not at all involved in the bargaining process. This was, for example, the case at a printing plant owned by the Bertelsmann AG.[8]

The high level of involvement of trade unions can be explained by the features of the German corporate governance system. German industrial companies have always been marked by a high involvement of labour in the decision-making process. In most companies, more than 80% of works councillors are union members; in many cases the figure is 100%. In addition, unions are very often represented on the supervisory board of a company. In all companies of our sample that negotiated a plant-level pact, representatives of employees and/or trade unions occupied half of the seats on the supervisory board. Hence, employers usually accept trade union officials as negotiators in plant-level employment pacts. They prefer to involve them rather than risk social peace.[9]

The peace-keeping interest is also the reason for the fact that only a few of the pacts clearly violate collective agreements. In our sample, only 10% of the agreements actually broke the terms and conditions of the sectoral collective agreement. Consequently, the management can rely on the willingness of trade unions to improve flexibility and to be generous in exceptional cases. Deviations from collective agreements are usually controlled by unions.[10] Unions have agreed to sign several agreements that did not meet collective agreement standards in order to prevent job losses or the firm's bankruptcy. During the 1990s this voluntary practice entered the collective agreement framework, as we will describe in the following section.

4The Transformation of Collective Bargaining

Internal flexibility in the framework of high external rigidity has traditionally been the modus of the German industrial relations system. As we have pointed out, negotiated flexibility at the plant level and comprehensive rules in centralized collective agreements have been the trademark of the German model. Are the company pacts of the 1990s, therefore, a mere continuation of the successful path of company adjustment to changing economic conditions? In this section, we will argue that, despite the formal stability of German wage bargaining institutions (central collective agreements), the scope and power to standardize employment conditions by collective agreements has changed to a degree which implies a more fundamental transformation of the system itself.

Generally, it is indeed the case that plant-level bargaining beyond the central collective agreement has been an integral part of the German industrial relations system. During the 1960s, it was even trade union policy to negotiate additional agreements at the plant level to capture wage drift (betriebsnahe Tarifpolitik). The implementation of the metal sector wage agreement (Lohnrahmentarifvertrag II) in Nordbaden-Nordwürttemberg in 1973 required the management and works councils at the plant level to negotiate up to 30 supplementary plant agreements (Schauer et al. 1984; Billerbeck u.a. 1982: 176); similarly the framework agreement in 1978 (Sadowski 1985: 244).

Moreover, the agreements on working time reduction in the metal sector in 1984 opened the previously standardized regulation of working hours in central agreements for tailor-made plant-specific working time regimes (Thelen 1991; Bispinck 1997). More than 10,000 plant-level agreements were negotiated following the 1984 collective agreement. Within centrally defined parameters, plant level negotiations aimed at finding flexible solutions.[11]

The example of the dispute over working hours also highlights the double logic of interest representation on the part of the employers' associations: the acceptance in principle of reduced working hours (in spite of the protest of small business) helped to restore social peace at the collective bargaining level and the opportunity of working time flexibility increased the room to manoeuvre at the company level (Wiesenthal 1987: 173ff). Employers were obviously hoping to compensate for the costs of working time reduction by making productivity gains through working time flexibility. This was, however, a strategy mainly available for big companies, which developed highly sophisticated work schedules for their large workforces. In particular, the big manufacturing companies using high-technology equipment were able to decouple the production process from individual working time arrangements and thereby achieve high productivity gains (Silvia 1999). Research on the implementation of working time flexibility has shown that the majority of small firms continued to keep a standardized working week, while more than 80% of big companies did not (Hermann et al. 1999). The main long-term effect of this conflict was the increasing role of plant-level bargaining as the combination of working time reduction and working time flexibility had to be negotiated and implemented in each single company independently (Thelen 1991, 165).

In this sense, the developments of the 1990s might be interpreted in the same vain. The transfer of industrial relations institutions to the east in the course of reunification provided a further impetus towards the delegation of bargaining rights to the plant level. After bargaining institutions had been transferred to the east, the impact of high wage agreements was felt immediately. In 1993, the coalition of the various political actors which had pushed for the transfer of western institutions fell apart. The east German firms in the metal sector which could not afford to pay high wages any longer denounced the collective agreement (Bispinck 1993; Henneberger 1993;). The system was saved by the introduction of so-called 'hardship clauses' into the collective framework. Companies could apply for exemption from the collective agreement and would be granted this if they met certain conditions. For the first time in post-war history, a German firm which is legally bound by a collective agreement was allowed to fall short of collective agreement standards in order to survive. Research commissioned by trade unions reported 181 companies which applied for hardship in east Germany between 1993 and 1996 (Bahnmüller, Bispinck et al. 1999).

Hardship clauses in the east were introduced in the midst of the recession. Between 1992 and 1993 more than 0.5 million jobs were lost in German manufacturing. In this context, hardship and exemption clauses spread across all the industries and spilled over to the west in no time. The issues on which exemptions were to be made were similar to those in company-level pacts: flexible and longer working hours, working time reduction with pay cuts, and cuts in pay and basic bonuses. The most generous hardship clause was introduced in the chemical sector agreement in 1995, which allowed companies in hardship a cut in basic pay of up to 10% (Bispinck 1997).

While employers' associations in general aimed to find a peaceful way of introducing flexibility and cost-cutting measures into sectoral collective agreements, trade unions developed a strategy in which they opposed any general concession in principle, but accepted major concessions in individual cases. Hardship clauses therefore emphasize the singularity of the cases by introducing qualifying conditions such as "in particular justifiable cases" (as in the 1973 metal sector collective agreement of the powerful region of Nordwürttemberg-Nordbaden). In many cases, these clauses were an attempt to bring the regulations of collective agreements in line with reality since company-level social pacts were being agreed at a rapid rate without anyone bothering about the terms and conditions of the relevant collective agreement.[12] Especially in the car industry, company-level pacts had already become a matter of course and preceded the introduction of hardship clauses in the collective agreement.

4.1What Has Changed?

Compared to the traditional pattern of the negotiated internal flexibility and external rigidity of the German industrial relations system, two major changes occurred during the 1990s which cast doubt on the continuation of the current system.

Firstly, the hierarchical order between collective agreement and plant-level negotiations has been reversed. Previously, plant-level bargaining took place within a predefined corridor. The associations at the regional or national level decided upon the room for manoeuvre at the plant level and defined the scope in which plant-level actors were able to find their own solutions. Legally, the collective agreement supersedes plant-level agreements. Plant-level actors need explicit permission to reorganize issues of pay and working time at the plant level. If the collective agreement does not include this permission or a possible delegation of bargaining to the plant level, plant-level actors are legally requested to refrain from bargaining about these issues. The delegation of bargaining rights from the central to the plant level has been a process of enabling plant-level actors to deal with issues which were principally reserved for the associations. The associations took their right to restrict plant-level bargaining seriously and in particular trade unions tried to regulate as much as possible at the central level. Companies had to wait to be enabled by collective agreements to change practices at the plant level. The plant-level pacts for employment and competitiveness of the 1990s have in many cases superseded collective agreements. Even though the legal situation has not changed and even though many pacts do not violate agreements, associations have lost their capacity to limit or avoid plant-level bargaining by enforcing collective agreements. While previously plant-level bargaining implemented collective agreements, the agreements of the 1990s tended to implement the current practice at the plant level. The superiority of collective agreements vis-à-vis plant agreements on crucial issues of terms and conditions has been lost.[13]

Secondly, plant-level agreements are no longer to the advantage of employees. In the past, the delegation of regulation to the plant level was only used if the plant-level solutions were regarded as being to the advantage of the employees. The 1960s' idea of additional company agreements aimed at improving the terms and conditions by concluding supplementary agreements. The supplementary plant level agreements of the 1970s were agreed in the context of programmes of the so-called "humanization of work" (Humanisierung der Arbeit), which aimed at improving working methods. The working time flexibility that was agreed in the 1980s was at least ambiguous with regard to its effects on employees since they often gained financially from working time flexibility. The social pacts of the 1990s, however, were primarily concession-driven. Cost-cutting not only included further flexibility gains but also unpaid over time or outright wage cuts. The previous principle that the collective agreement sets the minimum terms and conditions, which are then topped up at the plant level, has been abandoned. Opening clauses in collective agreements enable plant-level actors to undercut the pay laid down in collective agreements.

In effect, the associations have lost their power to effectively set terms and conditions for individual employees through the practice of company pacts. Changes in collective agreements in the form of opening clauses are driven by plant-level practices and not vice versa. Furthermore, the collective bargaining associations have signed away these rights. While many opening clauses contain the right of the associations to veto plant-level agreements, in practice this rarely happens. Associations know that, once management and works council have negotiated an agreement, there is hardly any way of stopping them implementing it. The central collective agreement, which used to be the main piece of regulation on which the standardization of the terms and conditions of German employees rested, has turned into a document which offers guidance to remuneration, but leaves considerable scope as to its regulation at the plant level.

Does this mean that trade unions have lost their capacity to call social peace into question? If this was the case, big companies might lose their interest in centralized wage bargaining as its peace-keeping function becomes redundant. While trade unions are increasingly unable to prevent the emergence of plant-level pacts, they still have the power to resist special issues. At the political level they define the limits of cost-cutting and deregulatory policies. In these cases, they can rely on the willingness of their members to go on strike - even if employees accept a competitiveness pact at the plant level.

One prominent example is the conflict over sick pay in 1996. The employers' association of the metal sector wanted to cut sick pay, but the union resisted this issue, and the employees of several car manufacturers and their suppliers immediately went on strike. In order to re-establish social peace, business backed down. A wave of additional plant-level pacts was negotiated. Management and works councils regularly agreed on the retrenchment of bonuses and premiums in order to maintain the level of sick pay. Hence, the union was able to define certain taboos of cost-cutting policies at the collective level (i.e. sick pay) while it was unable to prevent plant-level cost-cutting programmes.

5Institutional Change in the German Wage Bargaining System

In the preceding sections we have offered three arguments as to why large German employers have not abandoned the centralized bargaining system. Firstly, industrial peace has become of major importance for companies in an uncertain business environment. All things being equal, the system guarantees a much higher level of industrial peace than would be the case in a decentralized bargaining system. Secondly, companies have been able to offset cost increases and the rigidities imposed by collective agreements through plant-level pacts for employment and competitiveness. Thirdly, collective agreements themselves were opened up to provide a range of options for companies to deviate from the standardized regulations of the agreement. This was possible as unions accepted the possibility of local flexibility being extended without calling social peace into question.

While the reasons for keeping a centralized wage bargaining system outweigh the pressures for outright decentralization, the adjustment of the German wage bargaining system to an internationalized economic environment has left clear traces of change on its institutions. To some extent these changes resemble the notion of organized decentralization - the dual system combining strong centralized co-ordination with substantial decentralized elements - as described by Traxler (1995). The term 'organized decentralization' describes a wage bargaining system in which "bargaining tasks have been deliberately delegated to lower-level associations in a way that does not eliminate co-ordinating control by the higher-order associations over the bargaining process at lower levels" (Traxler 1995: 7). Because of their peace-keeping interest, big companies seek flexibility within the system and accept the unions' role as a veto-player. Thus the institutional setting remains relatively stable. Nevertheless, under the cover of formal institutional stability we can witness a process of institutional transformation towards a competition-driven model of wage regulation that rests on the local deviation from centralized wage setting rather than on its implementation. While collective agreements at the sectoral level still provide a general framework for pay and working conditions, the actual terms and conditions for individual employees and plants can vary substantially with the performance of the firm. Highly flexible working time regimes enable production on demand to take place without an increase in pay for overtime. Opening clauses in collective agreements and plant-level pay agreements allow companies to adjust their labour cost increasingly to the business cycle.

The German wage bargaining system has traditionally provided high wages and standard working conditions for a large number of employees and thereby eliminated low cost competition within Germany. It has also created a high degree of social peace. Moreover, institutionally supported employment security, which was inherent in the German model for decades, enhances the commitment of the workers and encourages employer investment in improved skills - both features of high-quality production (Streeck 1997).

As a consequence, companies have had to seek quality leadership. Diversified quality production (DQP) has emerged as a strategic response by firms to the constraints of the rigid external regulation of wages, workplace co-determination and labour legislation. DQP does not refer to the individual business strategy of a firm but reflects a type of industrial order. It depends on a specific institutional and organizational setting to offer collective goods which firms are not able to produce on their own. For instance, social peace is one of the general and unspecific 'redundant' capacities which make the model work well (Sorge 1985). Social peace at the company level contributes to better quality and higher productivity, which makes high wage settlements pay off.

Today the interaction of wage regulation and production model has been reversed. Under the impact of internationalization, German companies have become price-takers not price-setters for their products. This does not mean that they have left the path of diversified quality production. It does mean, however, that, today, even quality production has to face the necessity of being cost-conscious. Previously, companies had to adjust their production regime to the labour market regulation, which aimed at providing equity and equality of working conditions. The newly emerging wage-setting regime follows the need of production strategies aimed at market performance and competitiveness. In this sense, the German wage bargaining system has changed its functioning in the political economy without changing the formal institutions. This transformation is driven by the changing product market strategies of big German firms that have to adjust to international cost competition.

The emerging cost competition-driven model of wage regulation is still based on the beneficial effects of the traditional centralized wage bargaining institutions. In particular, the presence of high levels of social peace and moderate and predictable wage negotiations still serve as a precondition for introducing further flexibility. At the same time, however, competition-driven forms of wage regulation introduce competitive elements that are set to undermine the co-ordination capacity of the system.

Pacts for employment and competitiveness and the emergence of opening clauses in the collective agreement framework that introduce price competition into the German model cause the free rider problem to reoccur. In competitive markets, companies are tempted to deviate from collective agreement standards in order to exploit a short-term advantage - and they know very well that their competitors do the same. The history of pacts for employment and competitiveness in Germany reveals that cost-cutting and flexibility agreements spread in no time between firms and sectors. This is particularly true in highly oligopolistic sectors with massive overproduction - such as the car industry.

The cost-cutting interests of big companies and the emergence of plant-level pacts for employment and competitiveness make the interaction between co-operation and competition more difficult. Whereas quality production depends on co-operation between firms, cost-conscious production does not. On the contrary, cost-cutting effects only provide rents as long as the competitors do not close the gap. Hence, the introduction of cost-consciousness into the model of diversified quality production leads to a new form of competition between firms in which cost-cutting is not necessarily based on productivity growth any more, but on reduction in pay.

As we have shown, this new competition has not led to the breakdown of the centralized wage bargaining system. Yet its capacity to solve collective action problems has clearly shrunk. The liability of the collective agreement standards is to some extent no longer an institutionally guaranteed collective good. Rather, it depends on the processes and results of private face-to-face negotiations between specific unions and specific firms in reference to an opening clause.

The developments of the 1990s have shown that the character of social peace has shifted from an institutional effect of the wage bargaining system to a private good. Firstly, employment security at the company level has been questioned by employers and has become contractual in the form of pacts for employment and competitiveness. For example, some workers in the metal sector are protected by a firm-specific formal employment pact while others are not. Secondly, at the sectoral level, social peace no longer is a "redundant capacity" in the sense of a beneficial institutional effect, but a private good that firms and employers' associations try to deal with. By negotiating cost control and flexibility without risking the peace-capacity of the system, firms and sector-specific associations try to find their private "social peace equilibrium" (Streeck 1989: 57), which may differ between companies and sectors. It is the privatization of previously generalized institutional effects that is at the centre of the transformation of German wage bargaining institutions.

Appendix: The MPIfG Database on Big German Firms

The sample of 120 companies (100 biggest of 1986 and 1996) in Germany was taken from the German Commission on the concentration of German industry (Monopolkommission), which since 1978 has bi-annually ranked the largest 100 German companies on the basis of net value added (in Germany). In contrast to sales, which are a more common variable for ranking companies, net value added has several advantages. First, it is a more stable factor, which allows for the inclusion of banks and insurance companies. Second, it ignores different price developments across industries, which would bias the company sample. Third, net value added can indicate the vertical integration of different industries. For example, in retailing companies, which have a low degree of vertical integration, the ratio of net value added to sales is frequently lower than in companies in other industries (Monopolkommission 1998: 153).

The selection by size (measured in value added) produces a bias towards the largest employers since labour costs are a major component of value added. The firms in the sample employ 3.8 million employees in Germany; about 16% of all employees in the private sector. Similarly, they contribute 18% to the gross national product produced in the private sector. Also, in terms of international activities, the sample covers a proportionally large share. The companies employ about a third of all employees of German companies abroad (1.4m compared to an estimated 3.5m employees). They are therefore on average much more internationalized than the average German company.

In our sample, we have 76 manufacturing firms, 39 firms in the service sector and five companies in the construction sector. The manufacturing firms include the chemical sector (14), industrial machines (13), automotives (11), electronics (3) and others. The service sector firms break up in banks (11), insurance firms (8), retail (10) and general services (10).

References (25/-25)

Bahnmüller, R., R. Bispinck, et al. (1999). Tarifpolitik und Lohnbildung in Deutschland. Düsseldorf, WSI-Diskussionspapier Nr. 79.

Billerbeck, U., C. Deutschmann and R. Erd (1982). Neuorientierung der Tarifpolitik. Veränderungen im Verhältnis zwischen Lohn- und Manteltarifpolitik in den 70er Jahren. Frankfurt, Nomos.

Bispinck, R. (1993). Kampf gegen die "Tarifwende" - Die Tarifbewegungen 1992 in den alten Bundesländern. WSI Mitteilungen 46, 129-153.

Bispinck, R. (1997). Überreguliert, undifferenziert, unbeweglich? Zum Flexibilitätspotential des Tarifvertragssystems und zu den Anforderungen an die künftige Tarifpolitik. In: T. Dieterich (ed.), Das Arbeitsrecht der Gegenwart. Jahrbuch für das gesamte Arbeitsrecht und die Arbeitsgerichtsbarkeit. Nachschlagewerk für Wissenschaft und Praxis. Erich Schmidt Verlag, 49-67.

Bronfenbrenner, K. (2000). Uneasy Terrain: The Impact of Capital Mobility on Workers, Wages and Union Organizing. Cornell University.

Calmfors, L. (1993). Centralisation of Wage Bargaining and Macroeconomic Performance - A Survey. OECD Economic Studies 21(Winter), 161-191.

Genschel, P.(1997). The dynamics of inertia institutional persistence and change in telecommunications and health care. Governance 10(1), 43-66.

Hancké, B. (1998). Industrial Restructuring and Industrial Relations in the European Car Industry. Discussion Paper FS I 98-305. Wissenschaftszentrum Berlin für Sozialforschung.

Hassel, A. (1999). The erosion of the German system of industrial relations. British Journal of Industrial Relations 37, 483-505.

Henneberger, F. (1993). Transferstart: Organisationsdynamik und Strukturkonservatismus westdeutscher Unternehmerverbände - Aktuelle Entwicklungen unter besonderer Berücksichtigung des Aufbauprozesses in Sachsen und Thüringen. Politische Vierteljahresschrift 34, 640-673.

Hermann, C., M. Promberger, H. Seifert and R. Trinczek (1999). Forcierte Arbeitszeitflexibilisierung. Die 35-Stunden-Woche in der betrieblichen und gewerkschaftlichen Praxis. Berlin, edition sigma.

Iversen, T. (1996). Power, Flexibility, and the Breakdown of Centralized Wage Bargaining. Denmark and Sweden in Comparative Perspective. Comparative Politics (July), 399-436.

Iversen, T. (1999). Contested Economic Institutions. The Politics of Macroeconomics and Wage Bargaining in Advanced Democracies. Cambridge and New York, Cambridge University Press.

Katz, H. C. (1993). The Decentralization of Collective Bargaining: a Literature Review and Comparative Analysis. Industrial and Labor Relations Review 47.

Kommission Mitbestimmung (1998). Mitbestimmung und neue Unternehmenskulturen. Bilanz und Perspektiven. Gütersloh, Verlag Bertelsmann Stiftung.

Kurdelbusch, A. (2001). The Rise of Variable Pay in Germany: Evidence and Explanations. Paper prepared for the 13th Annual Meeting of the Society for the Advancement of Socio-Economics, Amsterdam, 28th June - 1st July, 2001.

Lange, P., M. Wallerstein, et al. (1995). The End of Corporatism? Wage Setting in Nordic and Germanic countries. In: S. M. Jacoby (ed.), The Workers of Nations. Industrial Relations in a Global Economy. New York; Oxford, Oxford University Press, 76-100.

Lehmbruch,G. (1994). Dilemmata verbandlicher Einflußlogik im Prozeß der deutschen Vereinigung. In: W. Streeck (ed.), Staat und Verbände. PVS-Sonderheft 25, 370-392.

Mitchell, D.J.B. (1994). A Decade of Concession Bargaining in the United States. Labour and Society 11, 207-220.

Moene, K. O., M. Wallerstein, et al. (1993). Bargaining Structure and Economic Performance. In: R. J. Flanagan, K. O. Moene and M. Wallerstein (eds.), Trade Union Behaviour, Pay-Bargaining, and Economic Performance. Oxford, Clarendon Press, 63-131.

Monopolkommission (1998). Marktöffnung umfassend verwirklichen. Zwölftes Hauptgutachten der Monopolkommission gemäß § 24b Abs. 5 Satz 1 GWB. Baden-Baden, Nomos-Verlag.

Mueller, M. (2000). Taylorismus: Abschied oder Wiederkehr? Mitbestimmung 7/2000, 12-17.

Mueller, F. and J. Purcell (1992). The Europeanization of Manufacturing and the Decentralization of Bargaining: Multinational Management Strategies in the European Automobile Industry. The International Journal of Human Resource Management 3, 15-34.

O'Donnell, R. and C. O'Reardon (2000). Social Partnership in Ireland's Economic Transformation. In: Giuseppe Fajertag and Philippe Pochet (eds.), Social pacts in Europe. European Trade Union Institute (ETUI), Observatoire social européen (OSE), 237-256.

OECD (1997). Economic performance and the structure of collective bargaining. Employment Outlook 1997. Paris, OECD, 63-92.

Perez, S. A. (2000). From Decentralization to Reorganization. Explaining the Return to National Bargaining in Italy and Spain. Comparative Politics (July), 437-458.

Pontusson, J. (1996). Labor Markets, Production Strategies and Wage Bargaining Institutions. The Swedish Employer Offensive in Comparative Perspective. Comparative Political Studies 29, 223-250.

Rehder, B. (2000). Abweichung als Regel? Die Mitbestimmung 46, 12-16.

Rehder, B. (2001). The Impact of Plant-level Pacts for Employment and Competitiveness on the Institutional Change of the German System of Industrial Relations. Paper Prepared for the 13th Annual Meeting of the Society for the Advancement of Socio-Economics, University of Amsterdam, June 28th - July 1st 2001.

Sachverständigenrat zur Begutachtung der gesamtwirtschaftlichen Entwicklung (1997). Reformen voranbringen. Jahresgutachten 1996/97. Stuttgart, Metzler-Poeschel.

Sadowski, D. (1985). Betriebsverfassung und Betriebssyndikalismus - Zur gegenwärtigen Bedeutung klassischer Funktionsprobleme von Betriebsdemokratien in Deutschland. In: E. Böttcher et al. (eds.), Die Vertragstheorie als Grundlage der parlamentarischen Demokratie. Jahrbuch zur Neuen Politischen Ökonomie. Tübingen, Mohr, 234-249.

Schauer, H. et al. (1984). Tarifvertrag zur Verbesserung industrieller Arbeitsbedingungen. Arbeitspolitik am Beispiel des Lohnrahmentarifvertrages 2. Frankfurt a.M., Campus.

Schröder, W. and B. Ruppert (1996). Austritte aus Arbeitgeberverbänden: Eine Gefahr für das deutsche Modell? Marburg, Schüren.

Silvia, S. J. (1999). Every which way but loose. German Industrial Relations since 1980. A. Martin and G. Ross (eds), The Brave New World of European Labour. European Trade Unions at the Millennium. New York, Oxford, Berghahn Books, 75-125.

Sisson, K., J. Freyssinet, H. Krieger, K. O'Kelly, C. Schnabel and H. Seifert (1999). Pacts for Employment and Competitiveness. Concepts and Issues. European Foundation for the Improvement of Living and Working Conditions, Dublin.

Springer, R. (1999). Rückkehr zum Taylorismus? Arbeitspolitik in der Automobilindustrie am Scheideweg. Frankfurt a.M./New York.

Streeck, W. (1981). Qualitative Demands and the Neo-Corporatist Manageability of Industrial Relations. In: British Journal of Industrial Relations 14, 149-169.

Streeck, W. (1992). From National Corporatism to Transnational Pluralism: European Interest Politics and the Single Market. In: T. Treu (eds.), Participation in Public Policy-Making. The Role of Trade Unions and Employers' Association. Berlin, De Gruyter, 97-126.

Streeck, W. (1993). The Rise and Decline of Neocorporatism. In: L. Ulman, B. Eichengreen and W. T. Dickens (eds.), Labor and an integrated Europe. Washington D.C., The Brookings Institution, 80-101.

Streeck, W. (1997). German Capitalism: Does it exist? Can it survive? In: C. Crouch and W. Streeck. Political Economy of Modern Capitalism. Mapping Convergence and Diversity. London, Sage Publications: 33-54.

Thelen, K. (1991). Union of Parts. Labor Politics in Postwar Germany. Ithaca and London, Cornell University Press.

Thelen, K. (2000). Why German Employers Cannot Bring Themselves to Dismantle the German Model. In: T. Iversen, J. Pontusson and D. Soskice. (eds.), Unions, Employers, and Central Banks. Cambridge, Cambridge University Press: 138-172.

Thelen, K. (forthcoming). How Institutions Evolve: Insights from Comparative - Historical Analysis. In: J. Mahoney and D. Rueschemeyer (eds.), Comparative-Historical Analysis: Innovations in Theory and Method.

Traxler, F. (1995). From demand-side to supply-side corporatism? In: C. Crouch and F. Traxler (eds.), Organized Industrial Relations in Europe: What Future? Aldershot, Avebury, 271-286.

Traxler, F. (1998). Collective Bargaining in the OECD: Developments, Preconditions and Effects. European Journal of Industrial Relations 4(2), 207-226.

Turner, L. (1998). Defending the High Road: Labor and Politics in Unified Germany. Ithaca, NY, Cornell University Press.

Völkl, M. (1998). Krise des Flächentarifvertrages und Integrationsfähigkeit von Arbeitgeberverbänden. Eine empirische Untersuchung am Beispiel zweier mittelständischer baden-württembergischer Branchen. Industrielle Beziehungen 4(2), 165-192.

Wallerstein, M. (1998). The Impact of Economic Integration on European Wage-Setting Institutions. In: B. Eichengreen and J. Frieden (eds.), Forging an Integrated Europe. Ann Arbor, The University of Michigan Press, 185-208.

Wallerstein, M. and M. Golden (1997). The Fragmentation of the Bargaining Society. Wage Setting in the Nordic Countries, 1950 to 1992. Comparative Political Studies 30(6), 699-731.

Wallerstein, M., M. Golden, et al. (1997). Unions, Employers' Associations, and Wage-Setting Institutions in Northern and Central Europe, 1950-1992. Industrial and Labor relations Review 50(3), 379-401.

Wiesenthal, H. (1987). Strategie und Illusion. Rationalitätsgrenzen kollektiver Akteure am Beispiel der Arbeitszeitpolitik 1980-1985. Frankfurt/New York.

Zagelmeyer, S. (2000). Brothers in Arms in the European Car Wars: Management - Labour Pacts in the Context of Regime Competition. MPIfG Working Paper 00/2. Köln, Max-Planck-Institut für Gesellschaftsforschung.

Endnotes

1    The material is based on the research project "The German System of Industrial Relations under the Impact of Internationalisation", which is being carried out at the Max Planck Institute for the Study of Societies. We would like to thank the members of the MPI companies project Wolfgang Streeck, Jürgen Beyer, Martin Höpner, Antje Kurdelbusch and Rainer Zugehör for their comments and support throughout the project. Thanks also to the participants at the Workshop on Institutional Change in German Industrial Relations at the MPIfG on 8th and 9th December 2000, in particular to Reinhard Bahnmüller, Bob Hancké and Richard Hyman, and finally to André Kaiser and Armin Schäfer, who acted as internal reviewers at the Max Planck Institute for the Study of Societies.

2    Note the use of the term wage bargaining centralisation. Some of the political economy literature distinguishes between centralized bargaining systems at the national level and intermediate bargaining systems at the sectoral level (in particular Iversen 1999). In our understanding, a centralized wage bargaining system is defined here as any bargaining level above the level of the firm. 

3    The trade-off also depends on the company's production system and its product market. For the sake of clarity we confine our considerations to the issue of company size, since in our view it is the dominant factor for diverging company interests with regard to wage bargaining in the German case. 

4    In the past one could assume that small companies were more focused on domestic markets than international markets. If that is the case, central agreements take wages out of competition and thereby level the competitive base between companies. In an open economy, however, this only applies to a small and shrinking fraction of companies. 

5    While the emergence of pacts for employment and competitiveness at the company level is not only a German but also a European and American phenomenon (Zagelmeyer 2000, Hancké 1998, Sisson et al. 1999, Mitchell 1994), it should be noted that the necessity for management to come to co-operative solutions with the works councils distinguishes these agreements from mere "concession bargaining" as took place in the US during the 1980s.

6    See Appendix on the sample of companies. 

7    The simultaneous emergence of negotiations in companies of the same sector is not always an indicator of increased competition, but it may also indicate a high degree of inter-firm cooperation. For example, the management as well as the works councils of German car manufacturers usually exchange their ideas and experiences in the process of negotiating and implementing the pacts. Hence, processes of policy-learning seem to contribute a lot to the diffusion and appearance of the pacts. Owing to space restrictions we cannot deal with this issue here. 

8    Because of the fact that Bertelsmann is a media company, the company is exempted from the obligation to include employees' representatives on the supervisory board, although it would qualify for co-determination in other respects. 

9    One main difference between concession bargaining in the United States and in Germany is that German managers - since they have no choice - seek to involve the unions in the bargaining process, whereas American firms often combine company-level bargaining with union avoidance strategies (Bronfenbrenner 2000). 

10  This correlation is statistically significant (r=.86, p=.000, n=46 companies). 

11  These parameters changed over the years. Generally, the average weekly working hours had to be met over a certain period of time. Management could ask their workforce to work longer hours in some weeks and shorter hours in others. Since 1991 the corridor in which the average working hours have to be met is one year, so that workers in the metal sector can in principle work on annual working time accounts. A further reduction in working hours in 1993 has led to the exemption of 18% of the workforce from a shorter working day, mainly in order to enable companies to make precious highly skilled engineers and software specialists work longer hours without having to deal with the works council. In 1995, the collective agreement on employment security enabled companies to shorten the working week to 32 hours with the equivalent cut in pay (Herrmann et al. 1999). 

12  Although the violation of collective agreements is unusual in big companies, it is fair to assume that this is not the case in smaller companies, which have suffered more from the system and which are not covered by the unions. Moreover, it is often very difficult to establish whether a company arrangement is in line with the sectoral agreement because of the complexity of flexibility measures. 

13  Throughout the 1990s there have been debates about the legal abolition of the superiority of collective agreements which is laid down in paragraph 77 III of the Works Constitutions Act (Betriebsverfassungsgesetz). The changing practice of the 1990s has made legal reform largely superfluous.

Copyright © 2001 Anke Hassel and Britta Rehder

No part of this publication may be reproduced or transmitted without permission from the authors.
Jegliche Vervielfältigung und Verbreitung, auch auszugsweise, bedarf der Zustimmung der Autorinnen.

MPI für Gesellschaftsforschung, Paulstr. 3, 50676 Köln, Germany