Corps de l’article

This paper examines the emergence of an employee network at Siemens Hofmannstraße plant in Munich, Germany. It is one of a number of such networks to have emerged in Germany in recent years in an interesting development in worker self-organization that has been little discussed. When redundancies were announced among “Hi-Tech” white-collar workers at the plant, an employee network emerged, that operated in a creative but tension-ridden synergy with the works council (Betriebsrat) and trade union IG Metall (IGM). The network is not an alternative to these institutions, but both an adjunct to them and a critical faction within them. It proved a dynamic force in helping to resist redundancies, and has since linked up with other employee networks, posing significant issues for the union: was this an opportunity or a threat for them? The paper therefore addresses what is paradoxically a neglected area of union renewal, i.e. worker self-organization and how a union interacted with it. It does so in the German context, arguably the core of the “European model” of worker representation. If it were to renew itself, the wider international resonance would be great.

The paper is structured as follows. Initially, literature is reviewed to generate a set of theoretical propositions about Hi-Tech workers, their behaviours and interactions with unions. The case study forms the bulk of the paper, in which we outline the origins and activities of the employee network and its relations with the works council and union, using mobilization theory to analyze the different roles of each institution. Finally, we conclude by reflecting on the initial propositions, the experience and its significance. Inter alia, we offer some suggestions as to how mobilization theory might be further extended.


What Facilitates Mobilization among Hi-Tech Workers?

It has been widely argued that unions need new approaches to mobilizing and organizing workers (see for example Nissen, 1999; Turner et al., 2001; Voss and Sherman, 2000). Jarley (2005: 615) refers to a general need for unions to invest in “social capital” in order to make workers more likely to mobilize. Similarly, Johnson and Jarley (2005), drawing on experience in programmes aimed at young workers, suggest they are more likely to act in solidaristic ways if unions can draw on “network density” increased by associational activity. For several other North American authors, “associational” forms of organization (Heckscher, 1996; Wald, 1998; Benner, 2002; Van Jaarsveld, 2004) resembling professional associations, emphasizing mutual assistance, are particularly appropriate vehicles for Hi-Tech workers. In the WashTech/CWA case analyzed by Van Jaarsveld, WashTech emphasized mutual assistance and political action because of the difficulty in breaking into collective bargaining. Hence, all of these researchers stress the importance of networking, building associational ties at local level, political activity and mutual aid both for mobilizing and organizing Hi-Tech workers. Our first proposition is therefore: thatemphasizing non-collective bargaining approaches may facilitate mobilization. Our second proposition is that these workers are likely to show a preference for non-traditional forms of collective organization. A third, closely related proposition is derived from the same researchers: thatHi-Tech workers’ organizations are likely to prioritize political and mutual aid activities over other possible fields of activity such as collective bargaining.

Institutional Embeddedness and Institutional Competition

Unions internationally have reviewed how they organize and operate. There has been much discussion on unions in Liberal Market Economies (LMEs), but prospects for union renewal globally are strongly affected by unions in Co-Ordinated Market Economies (CMEs). German unions exert strong influence in both European trade union institutions such as the ETUC and more widely in Global Union Federations because of their strong finances and perceived strength (Müller et al., 2003). This reflects the national context in which they operate. CMEs are so-called because they have high levels of institutional co-ordination, building high-trust, stable and long-term relationships between state, employers and unions (Hall and Soskice, 2001). Unions in CMEs are therefore situated in a very different broad social, political and economic context from their counterparts in LMEs, with consequences for how they seek to maintain their institutional positions. As Frege and Kelly (2004: 40) suggest, “Union power and influence is secured in different ways in different national systems,” with “organizing” approaches most likely to occur in “Liberal Market Economies” such as the USA or Britain.

“Organizing” approaches assume that unions adopt external orientations. In Germany, programmes of change such as IGM’s “participation-oriented workplace politics” have tried to involve new groups over the last twenty years, and to this limited extent may be described as “externally oriented.” However, they have also laid more emphasis than unions in LMEs on internal “organizational development” (Frerichs and Martens, 1999). IGM’s leadership has been criticized for failing to generalize even these initiatives beyond specific localities (Frerichs and Martens, 1999; Behrens et al., 2003), possibly because the union’s core of male, manual worker members exerts great internal influence and shows only marginal interest in minority groups (Hassel, 1999). Yet both the internal orientation of “organizational development” and the union leadership’s apparent failure to generalize initiatives may also be connected to the greater institutional support provided in CMEs to worker representation in general and unions in particular. It has been argued that this arises from a “central dilemma” for all German unions, between developing autonomous capacity for organization building and collective action on one hand, and the constraints resulting from institutional embeddedness on the other (Behrens et al., 2003: 37).

Despite recent modifications, the German IR system remains essentially a dual model (Jacobi et al., 1992; Müller-Jentsch, 1997), functionally differentiated between industry-wide collective bargaining and workplace co-determination. In the first arena, industry-based employers’ associations and unions bargain mainly over wages and working time; in the second, elected works councils represent employees irrespective of union membership. Works councils operate in complex relationships with unions (Croucher and Singe, 2004). Works councillors are legally required to work with employers “in a spirit of co-operation,” and are forbidden to initiate collective action (Müller-Jentsch, 2004). The German system of co-determination permits the establishment of work councils in undertakings employing a minimum of five workers. However, it provides unions with the possibility of having unionists elected as works councillors to represent all workers, with strong protection against dismissal, and in fact unionists constitute the great majority of works councillors (Müller-Jentsch, 2004). For long periods after World War II, unions were also strongly linked into the political process (Hyman, 2001), buttressing their position, while simultaneously locking them into bureaucratic, hierarchical processes that under-emphasize mobilization (Martens, 2005).

As the percentage of employees covered by neither collective bargaining nor by workplace representation stands at 27% in the western and at 42% in the eastern part of the country (Ellguth and Kohaut, 2005) gaps have developed within the system. Union penetration is especially low in highly qualified white-collar work. Partly because of this, Germany provides cases of the converse of our third proposition: there, emergent alternative organizations claiming representative and bargaining functions exist explicitly in competition with large general unions considered to be unrepresentative of specific interest groups (Renneberg, 2005: 261).

The issue of new organizations emerging in competition with unions may also be relevant in North America. It has been suggested that attempts by US unions to organize Hi-Tech workers “on an industrial model of unionism” have “almost universally failed” and there is therefore “little union representation in the information technology sector” (Benner, 2002: 143). Benner argues that experience in Silicon Valley shows that professional associations, acting as labour market intermediaries and employee advocates, are in practice a more successful model. He also indicates that the boundaries between professional associations and unions catering for such workers have blurred in recent years, and we note that this gives rise to potential for institutional competition. In such situations, unions may be nervous that mobilization may lead to competitive institutions.

Thus, unions may, because of their institutional embeddedness and fear of competition, harbour reservations about mobilizations in their industrial sectors. Hence our fourth proposition, that unions will display ambivalence towards local mobilizations where possible rival organizations emerge.

Hi-Tech Workers, Self-Organization and Mobilization Theory

Some, writing in a long-standing analytic tradition, have advocated unions building (or in longer-run historical terms, re-building) wider social movements by developing community coalitions (for recent examples, see Voss and Sherman, 2003; Fantasia and Voss, 2004). Fairbrother (2000) argues on the other hand that no specific new or revived form of unionism such as “social movement unionism” is in fact emerging. He further suggests that unions’ scope as organizations for transformation “from above” is limited: change is most likely to come from below, by workers organizing themselves (Fairbrother, 1996).

Evidence of such mobilizations, i.e. workers organizing themselves in ways not directly initiated by unions, is thin on the ground. Yet, although unions are clearly central to interest representation, the question is ultimately less one of how unions renew themselves, than of how workers defend their interests. However, Hi-Tech workers would not seem an especially likely group to throw up organizations of collective self-defence. Indeed, it has been suggested that “Hi-Tech” workers, have identities, mentalities and relationships to their work that make them unlikely to show any interest in traditional forms of collective organization (see for example Milton, 2002). Recent German studies of highly qualified white-collar workers have found a distant relationship between them and unions (Baethge et al., 1995; Faust et al., 2000; Kotthoff, 1998). Kotthoff’s (1992) research on Hi-Tech workers found that they could be attracted to unions, but only if these developed a professional aura (professionalistische Ausstrahlung) and represented the specific mentalities of these employees, something very unlikely to be achieved by unions such as IG Metall. On the other hand, Van Jaarsveld (2004) provides evidence from the US context that collective responses are indeed possible, but that they take distinctive forms.

Precisely how mobilization is achieved is therefore an issue, and some theory is available here. Mobilization refers to the “process by which a group acquires collective control over the resources needed for action” (Tilly in Kelly, 1998: 25) and the theory emphasizes the importance of language, ideology and leadership. Mobilization theory takes as its point of departure the conditions required for workers’ mobilization. Mobilization stems from perceptions of and responses to injustice (Kelly, 1998: 126), which may be transformed into collective interests. This becomes the basis for collective organization, but for mobilization to occur, workers must “attribute” their problems to an agency. Attribution processes “both derive from and reinforce a sense of distinct group identity” (Kelly, 1998: 127). Workers overcome a sense that their grievance is not widely shared, and come to see their problems as collective. Leadership frames issues through language and sets issues into a wider context, emphasizing solidaristic values and discouraging “free riding” by individuals. The three key elements are therefore attribution, social identity and leadership. Recent extensions of mobilization theory (Johnson and Jarley, 2004; Lévesque and Murray, 2002) introduce factors from social exchange theory to emphasize social cohesion, since social identity may not be a sufficient condition for mobilization. Mobilization theory draws on empirical research in the classic industrial relations tradition of the 1960s and 1970s on blue-collar workers where social cohesion may be less problematic than in our case. Nevertheless, the theory operates at a high level of abstraction, and this encourages the view that it may be applicable to workers more broadly. Our fifth and final proposition is therefore: that mobilization theory will provide an adequate framework for analysis of Hi-Tech workers mobilization.

Research Issues and Methods

Our propositions are examined through a case study structured to show the developing relationships during the conflict between Hi-Tech workers, the company and the three labour institutions. A multi-tiered approach was adopted because multiple types of actors were involved. There are four research strands. The first and fourth strands were preparatory and validatory; the core of the research consisted of the second and third elements. The first strand is observation, through one author’s involvement with IGM for over twenty years on a range of projects. The second strand is a set of eighteen semi-structured interviews with union officials, works councillors and network activists held in 2003-4 involving twenty-one people in total. These were followed up by four repeat interviews for clarification and up-dating purposes, held in the summer of 2006. Semi-structured interviews were conducted, transcribed, checked back with respondents and followed up with further telephone and e-mail inquiries for clarification. A third strand monitored a large volume of discussion in Siemens electronic employee fora. This was designed to study directly the nature of the daily exchanges taking place there, and provided an alternative perspective to interview accounts. The electronic exchanges were printed to provide just over one hundred pages of script, which was open coded (Strauss and Corbin, 1996) to establish shared themes. Contributions to network e-mail discussion were anonymous. They involved many different people and contributors themselves defined and contextualized problems. These data gave a broader picture than that available from activists and may also be more reliable than that derived from alternative methods such as focus groups or interviews where social desirability effects and over-directional interviewing are potential problems (Kvale, 1996). A fourth strand was designed to provide an informed expert perspective. Many of these experts had themselves some involvement with IGM, and at least three of them had been involved with the Siemenskonflikt. This strand was participation in three one and a half day sessions on union OD-projects organized by the “Hattinger Kreis,” a group of union advisors and academics.

Case Study


Siemens is among Germany’s most important companies, employing 417,000 including 170,000 in Germany (<>). Siemens’ Munich Hofmannstraße site employs 12,000 and is their main location for developing telecommunication networks (ICN) and mobile communications (ICM). Hofmannstraße had for many years retained a paternalist “Siemensfamilie” culture, stressing consensual change and employee welfare. Nearly 70% of those employed at Hofmannstraße had university or Fachhochschule (University of Applied Sciences) degrees. Reflecting the high level of local qualifications, for nearly 50% of those working in ICN, wages and conditions were not determined by union-employer sectoral agreements as they were categorized as Außertarifliche Angestellte, i.e. their wages were above the scales negotiated in union-employer bargaining (Mayer, 2005: 2). Local union density in Siemens is low: estimates range from two to six per cent, but in a typical German pattern, IGM had a majority on the works council. Works councillors report that employees traditionally took little interest in voting in works council elections.

In order to address its weakness at one of Germany’s most important companies IG Metall had taken initiatives: By 1996 IGM had established a specific Siemens project in Bavaria which was extended to cover the whole of Germany in 2003 (Cramer, 2003). This indicated a shift from supplying a predominantly local union service for works councillors towards a national “task force,” i.e. a group of fully paid officers that focused solely on Siemens and tried to provide detailed analyses, research and advice to workplace representatives, “modernized” communication and fostered cooperation and exchange between different Siemens sites.

The last 15 years brought a new management approach. New managers demanded that employees take on wider responsibilities, allowed space for self-organization in working teams, performance targets were ratcheted upwards and pressure increased to work longer hours. A company previously recognized for its long-term orientation, high investment in research and development and good employee conditions was now managed according to short-term, “shareholder value” considerations (Schuhler, 2003: 20-21). Thus, the company itself, by breaking well-established norms and consensual social values (Kelly, 1998: 27) was creating the precondition for attribution.

Redundancy and Mobilization: Chronology

In August 2002 management announced that 2,300 jobs at Hofmannstraße’s ICN and 300 at ICM segments were to go within six weeks. The underlying company strategy was to greatly reduce the permanent core of highly qualified staff and to strengthen R&D capacities in “low wage locations” abroad. The remaining workers in Munich were to be supplemented by external contract labour on a non-permanent basis should the economic situation or certain projects demand so (Mayer, 2005: 2). Flexibility/redundancy thus “hit the core,” a strategy adopted by other German employers (Kratzer, 2003). The announcement triggered a conflict that lasted over 18 months.

As the company publicized its downsizing plans, it announced that employees were to be offered a choice between signing termination agreements including compensation payments and employment in a German “employment body” (Beschäftigungsgesellschaft) designed to place workers in new jobs and train them.

It was announced that employees not willing to accept either option were to be made redundant. The works council decided not to follow the path usually taken in similar situations, i.e. negotiating compensation payments and making sure that the legal obligations concerning social criteria for redundancy were observed. Instead it sought to save jobs through mobilizing employees and supporting processes of employee self-organization. The IGM members on the works council argued that in order to resist, employees had to understand that redundancies were not a necessity, and that management could act differently, especially since the company as a whole had just announced the second highest profits in its history. The works council developed an alternative plan for the company that recognized Siemens’ competitive environment. For that purpose it could use external resources provided by IG Metall. The union generated detailed analyses of IT markets and Siemens’ position. It also helped the works council use its legal right to involve external experts (under the Works Constitution Law, §80; §111 Betriebsverfassungsgesetz), establishing links with an economist who became a permanent advisor. Employees fed knowledge of technological developments, markets, customer demands and needs into the discussion, using the works councils’ intranet website but also new communication channels established through an employee network described below. The plan emphasized the importance of human capital. Staff continuity was required since technological development and the integration of existing technologies demanded staff continuity to avoid a “downward spiral.” Staff retention could be achieved through working time flexibility: all employees’ working weeks should be reduced by two hours with corresponding wage reductions, a move to be reversed if business conditions improved. Additionally, outsourced work was to be brought back in house (for more detail, see Schuhler, 2003). The plan was publicized by the works council on the net. The plan gained credibility amongst workers, since it recognized the company’s position, affirmed their value as qualified specialists, and shared adaptation costs equitably. In September and October 2002 the works council’s home page registered 150.000 “hits.” A typical e-mail to the works council read: “Well done, works council. The model you presented combines the demands of the business situation (quick, cost saving) and spreads the burden… Keep up the good work” (cited in Schuhler, 2003: 28).

It was only when the alternative plan appeared that mobilization theory’s “attribution” process was fully achieved since it convinced employees of management’s responsibility for the crisis. Employee willingness to resist was also furthered by de-legitimizing management’s offers. The works council and IGM pointed out that older workers (45 years +) were unlikely to find work through the Beschäftigungsgesellschaft. This institution was described as a “trick” to circumvent the provisions of dismissal protection, which provides cover especially for older workers, those with long tenure and families.

During this phase of the conflict, the main works councillors and their co-operators in the union thus strived to broaden and deepen employee involvement. Instead intensified discussion was required to enable employees to formulate and act upon their own interests (Mayer, 2005: 4). Thus between August and December 2002, nine workplace meetings were organized, each attended by 3,000–4,500. Additional meetings were held outside the workplace, using IGM’s facilities. Employees sought to publicize the issue through demonstrations and stalls in Munich’s city centre (some 200 participated in this). As the conservative CSU demanded that Siemens adopt working time reductions to save jobs, media attention grew further. However, Siemens’ first reaction to employee resistance was to harden its position and threaten 2,300 redundancies for operational reasons. The works council stated that it would contest each case individually, i.e. there would be 2,300 lawsuits. IGM took a pragmatic stance in relation to individual legal assistance, normally given only after three months membership. The local IGM offered immediate coverage to those willing to join the union, greatly increasing employees’ willingness to bring cases. Within three weeks, 900 joined IG Metall; many others took up private insurance in order to be able to bring cases. A demonstration of 3,000 exerted further pressure on management seriously to negotiate with the works council (Wanzek, 2005: 10). In a workplace agreement struck in late October 2002, management consented to a working time reduction of 2.5 hours/week and to insourcing some work previously given to external providers. The number of jobs to go at ICN was thus reduced from 2,300 to 1,100. These 1,100 were to either sign termination agreements or accept transfer to the Beschäftigungsgesellschaft. Should these measures not be sufficient to reduce employment by 1,100, management would resort to redundancies. The selection process (i.e. which individuals to dismiss) was completely under management discretion, the workplace agreement did not involve the works council in the selection process (i.e. selecting people according to social criteria). Whereas the workplace agreement (or Sozialplan, “social plan”) might be seen to indicate the end of the conflict, it actually escalated.

In early October 2002, 876 individual employees received the management “offer”: around 350 accepted transfer to the Beschäftigungsgesellschaft, 100 opted for a termination agreement or early retirement. About 400 remained and in January 2003 management named 366 individuals it had selected for dismissal (Mayer, 2005: 7; Wanzek, 2005: 8). As the works council had not participated in the selection process it was now in a position to contest management’s selection on the grounds that management had not followed legal dismissal requirements. The works council found that management had dismissed predominantly older workers (45 +), those with long tenure (20 years +), women, and even those with special protection (e.g., handicapped workers or those with a tenure of more than 25 years). The works council had to prove that a specific selection had violated the law, and that the person could be employed elsewhere in the company. The works council objected in 362 cases, but in 200 cases management stuck to its decision. In these cases employees could turn to labour courts, however as the then works council chair said: “just because you’ve got boxing gloves [legal support] it does not mean you put them on and fight.” While the works council could provide juridical advice and the union pragmatically offered “free” legal assistance, another actor grew in importance and greatly increased employee’s willingness to contest redundancies and file suits due to unlawful dismissal. We now describe the origins and operations of this employee network.

The Employee Network (NCI): Origins, Operation and Leadership

The Siemens conflict saw the development of a novel form of worker self-organization, the NCI (Network Cooperation Initiative). At the end of this research, it linked around 900 employees. It offers information and advice on a broad range of issues and remains an influential actor in Hofmannstraße and beyond (<>). In the three years of its existence, the website has received around 750,000 hits and has attracted much mainstream media attention for its novelty, depth of coverage and informative content.


NCI originated before management announced mass redundancies in the summer of 2002, since management previously sought gradually to reduce employee numbers. Management then targeted selected employees to participate in an outplacement programme. Nominally, participation was voluntary but management pressured those unwilling to participate. Some resisted because it signified acceptance of their redundancy. A prime mover in the network was a female employee, then not a union member.

This female activist wrote a letter to senior personnel, asking why she had been targeted. Management reacted by giving her a warning and, during a hearing of her case, threatened summary dismissal. She then turned to the works council for help. Until then she had little interest in workplace representation, as works councillors she had contact with “lacked the common touch.” She therefore established links with other targeted staff to share information, offer emotional support, overcome individuals’ isolation and develop an analysis of events. Small-scale meetings were held: “An employees’ network began to materialize: small, unnoticed [by works council, union and employer], without political impact” (interview notes). She had previously found Siemens management receptive to criticism and employee participation, and felt no need for union membership or other forms of representation. However, when threatened with redundancy, she was impressed by the support the works council gave her and her colleagues. Thus, close working relationships between the leading works council representatives and the female activist, the prime mover behind the nascent employee network, developed. This contact made her re-consider her position towards unionism and she joined IGM. In brief, relations between network and works council began and continued to be co-operative during the dispute.


While a small core of network activists had thus come together before the announcements of August 2002, its dynamic development began when individuals received letters offering termination agreements/Beschäftigungsgesellschaft or redundancy in November 2002. Immediate employee reaction was described by the female activist: “It was as if a meteor had hit … whole departments put their work aside.” (Wanzek, 2005: 8; authors’ translation). A first e-mail call for an immediate meeting was answered by 30–40 people. According to one participant, the introductory round lasted three hours as employees took the chance to talk about their personal situation and fears. Its main function was to establish space for open communication and break down individuals’ emotional isolation. Widely observable reactions such as self-doubt and panic attacks were tackled by developing a collective understanding of the processes and individuals’ situation. The network (at that stage it did not yet have an official name) grew as its members approached their colleagues individually. The woman prime mover now offered to share her knowledge about employment law. She toured the workplace, giving talks, and people attending were asked to join the networks’ electronic mailing list. She thus took on functions usually taken by works councillors.

By December 2002 the network had 120 members, in January 2003 it had grown to around 400 (Schuhler, 2003: 48). By mid-February 2003 around 400 staff participated in sixteen working groups on a range of issues, including political questions such as labour market reforms, publicity and labour law. When management named individuals to be dismissed (January 2003) the network e-mailed colleagues, suggesting a meeting. An activist described an early meeting (“the meeting of the 100”):

We just had the idea of bringing everyone together. We did not have a fixed programme, nor did we want one. We wanted to overcome the normal situation, where someone gives information while employees are little more than passive recipients, asking a question now and again. We intended to demonstrate what the network was all about, i.e. a completely democratic form of organization, without defined rules, with its own dynamic, coming alive through the ideas and thoughts of all of the participants, going through this period of change. We realized our concept was quite bold. We just gave a short introduction, put down some ideas for discussion on posters and then left people to themselves. It all could have gone terribly wrong; it was a “make or break” situation. But so was the employees’ situation.


The main idea was to create an open space for discussion in which people could participate as fully as they wished and to break with “representative/proxy politics” (Stellvertreterpolitik; importantly, the German word carries two meanings in English: “representative” and “proxy”). The network would spread information widely and give workers the resources to act as they saw fit. Employees published numerous accounts on the website, describing their financial, work and family situations and their fears and feelings. The network’s second key function was thus to act as a support group for individuals in stressful situations. Openness to emotionality contrasted with the “coldness” of management and with that of traditional “male” (as the network’s originator described it) forms of worker representation. Thus, it did not conceive of itself primarily as a representative, nor as a bargaining body, but as a mutual support mechanism.

This function took prime importance following October 2002. It was then that employees had to decide individually whether to accept termination contracts, accept transfer to the Beschäftigungsgsesellschaft or fight dismissal in the courts. Taking the company to court was something employees hard hardly envisaged before and which required some confidence. The works councillors, during the process, learnt that the NCI could fulfil a necessary function, a function that the established forms of workplace representation could not fulfil:[1] whilst the works council could enlighten employees about their rights in redundancy situations, it could hardly influence the conditions required for workers actively to pursue these rights. According to the councillors, NCI’s “emotional labour” was fundamental. NCI was a space in which individuals met colleagues they were familiar with, where they felt secure and could develop self-consciousness. And only on this emotional foundation were workers prepared to use their rights, try to defend their jobs and to risk a lawsuit. The network was central to turning the individual cases into collective issues; it built solidarity. When, in early 2003, the 200 individual objections against dismissal were to be filed, the NCI turned this into a public demonstration: 600 employees jointly marched to the Labour Court. As individual cases were held in court, the network made sure that the courtroom was filled with employees. Its website gave detailed reports of court proceedings and as employees kept winning cases, the confidence of the remaining employees grew.

Leadership and Structure

The network refers to its “members” (i.e. those who signed on to its e-mail list) although there are no subscriptions and members are free to act as they choose. Much of the network’s activity is conducted in small working groups; electing, or naming working group and network leaders has been avoided. This is part of a wider consensus that victimization by the company should be made as difficult as possible, which also made it difficult for the researchers to uncover further details on informal leadership. This “leaderless” feature was later criticized officially by IGM, as a refusal by NCI to name anyone as responsible for anything. Meetings could be called by anyone, “without anyone censoring the call,” but a co-ordinating group has been formed. Employees that would like to take the initiative on a certain issue just inform others via Email, whether or not something “happens” is then decided in practice by others taking up or not taking up the idea.

Elections have not been held for any positions, though the co-ordinating group recognizes that situations might arise requiring a ballot. The approach is “non-ideological,” rejecting organization and action according to “external” ideologies in favour of individuals’ interests, experience and objective requirements. It stresses the need to base activity on conviction, on comprehension generated through collective discussion and rejects leadership based on formal loyalties or “obedience.”

Network Self-Characterization

We summarize key elements of the networks’ self-characterization in table 1. The data on which this is based derive from personal interviews, the self-description of the slate that NCI ran in the works council election (<>) and NCI website documents. Though the network is a heterogeneous body, individual statements clustered consistently around the areas identified in these last two “official” statements. The qualities attributed to management were used frequently by both NCI and works council in political exchanges with management, pointing to the relevance of language and of framing issues.

Table 1

NCI’s Self-Characterization in Relation to its Characterization of Management







Reasonable, rational (vernünftig)

Unreasonable, irrational

Demands and defends normal, stable employment to generate a climate that supports innovation. Thus acts in company’s interest.

Abnormal employment situation, climate of fear hampers employees’ innovative potential. Harmful to company’s interest.





In favour of satisfying “fun” work

Creates situation in which fun and creativity can’t develop

Responsible for the enterprise and for Germany as a location of production

Irresponsible towards company and country, disregard for wider societal development

-> Voir la liste des tableaux

The regulated form of German stakeholder capitalism is thus counterposed to the shareholder driven, American version. This is crystallized on placards carried on a demonstration: “The economy must serve the people” and “We won’t have our future traded on the stock exchange.”

Trade unionism was very weak at Hofmannstrasse, according to NCI members. Initial perceptions of unionism in table 2 are derived from the same sources as table 1.

Table 2

NCI’s Self-Characterizations in Relation to its Characterizations of the Union



Non-hierarchical, decentralized

Strictly hierarchical, centralized

Responsive, flexible, quick

Bureaucratic, inflexible, slow

Encourages self-activity of Angestellte

Represents blue-collar workers

Non-ideological, functional


Encouraging full employee voice

Allowing controlled employee voice

Encouraging emotional reaction

Not encouraging emotional reaction

-> Voir la liste des tableaux

The three organizations therefore had the potential for co-operation. NCI’s specific contribution was to increase the “social motives” (Kelly, 1998: 34) for mobilization.

Employee Network, Works Council and Union

NCI – works council relations were mutually supportive, based on “open dialogue” and “constructive co-operation” (<>); the IGM works councillors regarded themselves as part of the network. The works council’s IGM majority actively encouraged NCI from the beginning. The strong tendency of representatives to “substitute” themselves for their constituencies (Schmidt and Trinczek, 1999) was therefore not present here. Councillors were aware of earlier OD processes in IG Metal and thus familiar with a “hidden” tradition of employee involvement. The works council majority argued that, faced with an employer who had decided to confront rather than co-operate, employees could not depend on the limited legal possibilities open to works councillors; they needed themselves to be active: “Proxy politics won’t yield any success in the future. […] Works council politics and trade union politics have to be founded anew and must be based on active workforces.” (Mayer, 2005: 2, author’s translation; also see Fieber, 2003). The main works councillors, whilst being easily identifiable as IGM members, in person contradicted the stereotypical picture many Hi Tech workers held of unionists. The works council showed high levels of expertise, was seen as competent, flexible and receptive. In public exchanges with management it was very articulate and used an appropriate language: its s orientation on employee involvement and participation fitted with the high regard for the self-reliance Hi Techs workers traditionally display. This self-reliance was to be turned into a collective effort.

During the first phase of the conflict IG, Metall established very good working relationships with NCI and for the first time established a serious membership base in Hofmannstraße. The union provided important juridical advice and support to both works council and individual members. IGM took a pragmatic and flexible stance in relation to individual legal assistance, normally given only after three-month membership. The local IGM offered immediate coverage to those willing to join the union, greatly increasing employees’ willingness to bring cases. It displayed high levels of economic expertise which it fed into an alternative plan publicized by all three institutions and which reduced redundancies from 2,300 to 1,100. It also offered logistical back-up: meetings could be held at IGM’s facilities and the union offered to host NCI’s website. As it was feared that the company would sue individual employees or the website owner for breaches of confidentiality, transferring the website and thus liability to IGM offered greater protection to the network’s activists. As employees have the right to access the trade union site from their workplace for information purposes, the move also secured that employees could access the networks site. These supportive functions generated trust in the union; IGM was seen to offer protection and support to a mainly non-union network, without attempting to control its content. Through paying legal costs and facilitating NCI’s communications, the union was in mobilization theory’s terms reducing employees’ costs and risks of participation (or “increasing their reward motives”) and increasing the likelihood of success (Kelly, 1998: 34). The IGM thus practically demonstrated qualities high tech workers did not usually attribute to unions. IGM’s successful “renewal” is indicated by the fact that it recruited 900 members. In works council elections held in 2004, it increased its share of the vote to 46% while the AUB, whose campaign was supported by the employer, lost 10% (IG Metall, 2004). A new slate of NCI activists scored around 6% and its representative on the council formed a coalition with IGM. In 2004, voter turnout for the works council elections had increased and exceeded 70%. The union, works council and NCI thus worked in a symbiotic way, and for all three “institutions” the outcomes were positive.

Dispute Outcomes

During the first phase of the conflict (August 2002–October 2002) the works council, supported by the union, was largely successful in de-legitimizing management’s plans. Mobilization was based on the condition that a credible alternative to management’s plans had been developed. As employees’ motivational problems became obvious and employees successfully influenced public opinion through demonstrations, etc., management finally agreed to negotiate with the works council. The resulting workplace agreement was a success for the council as it reflected many of its proposals and reduced job cuts from 2,300 to 1,100. The second phase of the conflict, beginning in October 2002, was successful from employees’ point of view. By January 2005 all 159 employees had won their cases against Siemens, with only a few cases still pending. The Munich Labour Court argued that the company had disregarded dismissal protection and had to continue employment for those dismissed. Siemens appealed in all cases but was unsuccessful in all but one case (Mayer, 2005: 7).

In February 2004 IGM declared an end to the conflict at Siemens. Works councillors felt that employees were exhausted and those not threatened by redundancy feared that the conflict could threaten the existence of Siemens locally. There was also a strong interest on management side as local productivity was low, management legitimacy had suffered, it was obvious that its plans to cut employment had failed and the Siemenskonflikt was becoming something like an exemplary case for representatives elsewhere.

According to the union, a final workplace agreement on the future of those with long tenure, handicapped and elderly employees marked the end. However, this announcement took many by surprise and NCI members continued to voice criticism of Siemens’ treatment of these groups of employees. The network was not prepared to adopt the union’s “declaration of peace” with the company.

Continued NCI Activity and Tensions between NCI, Works Council and Union

Looking back on IGM’s initial reaction to NCI’s development, an NCI activist remarked: “Within IG Metall everybody said: this will peter out after four weeks. They observed us as if we were children playing in a sandbox” (employee interview). When early collapse did not occur, IGM showed that it feared a competitor union. A leading activist remembered a union official stressing “the difficulties of setting up a union. They feared we were in the process of building an IT union, something we never envisaged.” These fears receded as network activists adopted “the basic idea of unionism,” by experiencing that “a union is something that you can shape yourself” (interview with NCI activist). However, NCI activists’ rapprochement to the “basic idea” idea of unionism co-exists with continuing reservations. As one network activist (an IGM member) argues:

Trade union and networks are separate structures which contradict each other, because unions are strictly hierarchical organizations. And the more powerful we are becoming the more frictions there are going to be. And now the union has got a problem. If they annoy us, we’ll go elsewhere. We have built contacts with other networks very quickly and our homepage can be moved within thirty minutes. We want to act freely, without prior consultation. We must be free to do our own publicity work. And we have considerable power. I think the union really doesn’t want to annoy us. So we lean on IGM but we are still ourselves.

The network continues to operate, providing support for Siemens employees and ex-employees. It has a monthly radio programme where, recently, a network member explained that the network is not a union but adopts a pragmatic but critical approach to unions; its main methods are openness and empowering people to act for themselves. It is linked with other similar networks or “projects” at other major companies such as Phillips, Electrolux/AEG, MAN and UPS. It also continues to allow members to express views on its website, some of which have criticized IGM agreements at other workplaces. Consequently, a union official urged the network to remove one such from its website. The dialogue between IGM and the network continues, but the union refuses to co-operate with the network. For reasons explained below, the union’s works councillors have expelled NCI councillors from the IGM group on the council.

How much the union learned from the experience remains an open question. One IGM official voiced scepticism: “systematic organizational learning just doesn’t happen.” The union presents a critique of NCI in an article that appeared on IGM’s Siemens Dialogue website in May 2005, declaring “The sad end of an exemplary project.” It praised NCI for having organized solidarity, arguing that it showed how self-help could work in a weakly unionized context, and proved that it could be effective. It was a model of self-organization; IGM had supported NCI because it achieved things that unions could not currently achieve on their own. Nevertheless, the union could not co-operate further with NCI. The network could not be represented as the future of workers representation; the network claimed to be a “grassroots” organization but in fact had opaque rather than democratic structures. The network had claimed the right to support other workers at MAN and Bayer, criticizing IGM’s agreements. Overall, NCI had adopted a “permanent opposition” stance that was easy but irresponsible; NCI could not and should not pose as “the better trade union.”

The Network’s female prime mover gives a different account. She argues that after the works council elections of 2004, where IGM achieved a good result, the works council became both less co-operative with NCI and less transparent in its dealings with employees. When the works council reached an agreement with Siemens about redundancies in another part of the company, the network issued a leaflet about it, leading to NCI councillors’ expulsion from the IGM group on the works council. On the website she and other anonymous participants argue that the conflict is one between a form of unionization that is based on hierarchy and control, where member participation and democracy is only selectively used and applied. To describe the conflict as one between NCI and IGM is simply a way of externalizing the problem and the opposition. She claims IGM’s local behaviour is damaging unionism and cites IGM own directions for trade union work in the workplace to show that these have been violated by IGMs local leadership. To quote her:

…this is a conflict between those hierarchy-oriented long standing members, fearing for their power positions and those new Metallers who don’t accept blind obedience to hierarchy and command. IGM has not come to terms with the fact that it has people on the ground who are not ready to just follow calls for action and returns to work if asked to. It has not come to terms with members asking: Why should we do that? What can be achieved? And that adds constructive ideas. That’s what the conflict really is about.

Discussion and Conclusion

Our first proposition was thatemphasizing non-collective bargaining approaches may facilitate mobilization and is strongly supported. The works council and union, even in the context of the latter’s break with NCI, both acknowledged the network’s significant contribution to mobilization. It provided employees with a “low-level” access to activity. It contributed to making the redundancy issue a collective one. It took individual emotions and interests as its starting point and created a discursive space no other institution could offer. Its “emotional function” facilitated employees’ willingness to actively defend their jobs. Secondly, the network stimulated demonstrative action because it increased the level of association between workers. All respondents agreed that this had been a sine qua non of the Siemenskonflikt. Our second proposition was that these workers are likely to show a preference for non-traditional forms of collective organization. This proposition must be rejected on our evidence. These Hi-Tech workers saw a considerable and continuing role for the network but as part of awider alliance between that organization, the works council and the union. Workers showed a strongly increased interest in all three forms of organization during and after the dispute. The network therefore contributed to union renewal. While the network allowed high levels of association and expression, the works council provided on-the-spot representation and the union had both sectoral collective bargaining rights and resources that the other bodies lacked. The union’s high vote in relation to the NCI’s in the works council elections indicates that workers see the union’s works councillors as effective; this is compatible with continued interest in NCI. Our third proposition was that Hi-Tech workers’ organizations are likely to prioritize political and mutual aid activities over other possible fields of activity such as collective bargaining. This proposition is supported, since the network certainly prioritized both types of activity. Its discussion showed interest in collective bargaining issues, but the network had never claimed a bargaining role. In any event, to displace the primary collective bargaining agent, the union, strongly supported by the law during a dispute would be extremely difficult. The network was therefore strongly encouraged by the situation to adopt its non-collective bargaining approach. Its nature had more in common with North American organizations and prescriptions than with some of its German counterparts; why this is the case requires further comparative research. Our fourth proposition was that unions will display ambivalence towards local mobilizations where possible rival organizations emerge.

This proposition was largely supported, since the union, and the works council’s union group, initially supported the network but showed that they feared a rival organization becoming established and eventually broke with the network after the end of the dispute. Although the reasons for the break are disputed, all parties agree that the network’s criticism of union deals was an important cause. It might be argued that unions as representative institutions, selecting which interests to further and which interests to suppress, are not capable of long-term co-operation with forms of rather “libertarian” self-organization.

Our final proposition was that mobilization theory would provide an adequate framework for analysis of Hi-Tech workers’ mobilization. The theory was indeed useful, but with reservations. Mobilization involved three bodies and we summarize schematically in table 3 the contribution of each body to the three central processes of attribution, social identity and leadership.

Our reservations with respect to the theory are as follows. First, “discouraging of free riding” (something that may well be more significant during strikes than during this type of mobilization) was the opposite of both NIC and the union’s successful policies which provided information and support to workers but left the decision as to whether they should take legal cases to them.

Second, the theory only mentions emotion as a factor in mobilization in terms of leaders appealing to emotion (Kelly, 1998: 35–36). In this case, demand for an emotional forum came from below and therefore appeared rather differently. Third, an important aspect of established practice, use of external experts, not mentioned in mobilization theory, played a vital role in achieving “attribution.” In this latter respect, union practice was in advance of theory. The theory bears the marks of its origins in an epoch in which classical IR concepts analyzed blue-collar mobilizations.

Table 3

Works Council, Union and Network Roles in Terms of Mobilization Theory



Social Identity


Works Council

Publicly attributed blame to irrational management. Traditional forms of communication not sufficient, sees NCI as important arena for attribution processes and identity development.

Initially had weak links to Hi-Tech workers. Leading works councillor develops into charismatic figure.

Rejected normal “social plan” solution. Exerts leadership function but breaks with traditional “proxy” forms, i.e. sees need to constantly inform and involve employees and decide on this basis.

Leadership not based on traditional class rhetoric but based on expertise, and sensitive appeal to Hi Tech workers’ mentalities/language, etc.


Publicly attributed blame to management but union membership exceptionally low at Siemens. Provided resources for alternative plan.

Initially strongly associated with blue-collar workers. Changed identity in course of action as receptive to Hi-Tech workers’ needs.

Provided resources for developing alternative plan and supported legal action. Too weak to run action from outside but some influence via its members on works council.


Attributed blame to management, de-individualized problems and had capacity to communicate this in effective ways to Hi-Tech workers.

Strongly based on Hi-Tech workers identity, which it reinforced and developed. “Raised social incentives” to participation.

Created open space and encourages employees to act as they see fit, rejects any form of formal leadership and insists that action must be based on comprehension and conviction.

-> Voir la liste des tableaux

In conclusion, the employee network played a vital role in raising the social incentives to Hi-Tech worker mobilization at Siemens. It did not represent an alternative to the existing institutions of worker representation, but rather an important complement to them. It differed in this respect from several other recently emerged German organizations for specific occupational groups (such as doctors or airline pilots), whose appeal is precisely that they present an alternative. Further research is required to examine why different types of organization emerge. In our case, it may be relevant that previous “organizational development” programmes had sensitized local trade unionists to a need for change, as Heery (2005) identified as desirable, encouraging union openness towards the network’s ideas. The network operated in ways anticipated by much existing theory, but this theory failed clearly to predict the simultaneous increased unionization that occurred, showing that workers recognized the institutional complementarity. The Network’s essential contribution was to provide both a vehicle for appropriate forms of participation in a situation where worker representation had no strong roots and traditions amongst Hi-Tech workers, and to create a bridge to unionization. The union initially reacted positively to the network and was a key beneficiary of the mobilization, but it is noteworthy that recruitment occurred not because of union organizing efforts but from an essentially external impetus and that marked frictions between the union and the network escalated after the dispute. The dialogue between the institutions continues and although it is too early to say whether it will lead to wider union renewal, the network has established links with other networks that may yet provide further impulses in that direction.