UNIONS, THE QUALITY OF WORK, AND TECHNOLOGICAL CHANGE IN SWEDEN
Andrew Martin
In recent years the Swedish trade union movement has been trying to work out a strategy for influencing technological development.1 The ability to do so has come to be seen as essential to the achievement of work-quality objectives on which the unions have been putting increasing emphasis. In Sweden, as elsewhere, this emphasis has many sources. Among them is the growing dissatisfaction with Tayloristic modes of work organization, not only by new generations of workers but also by managers seeking ways to meet intensified competition by harnessing those workers’ commitment to changing production strategies, particularly those facilitated by new computer-based technologies. The degree to which unions can articulate worker dissatisfaction and mobilize it to influence the character and extent of the shift toward new forms of work organization is bound to affect the viability of the unions’ roles in the workplace and hence in the society beyond it.2 Some union response to the challenge posed by pressures for change in work organization is thus an imperative of organizational viability.
The flux in work organization presents unions with opportunities as well as threats, opening up the possibility of reasserting demands for autonomy and satisfaction in work that are deeply rooted in the labor movement’s ideological heritage but that had been subordinated in the instrumentalist bargain struck with “Fordist” mass production.3 Of course, the strength of this strand in the labor movement’s ideological heritage and the extent to which unions tap it in their response to the challenge of changing work organization vary considerably from country to country.4 In Sweden this strand, couched in such familiar terms as industrial democracy, has emerged as a salient feature of the unions’ current stance, after having been almost entirely obscured in the earlier postwar period.
As work-quality issues assumed growing importance for the Swedish labor movement, it found itself increasingly confronted by technological constraints on change in the content and organization of work. The scope for improving the quality of work was perceived as being significantly limited by technologies already in place. The unions were accordingly driven to seek a voice in the selection of technologies in order to bring work-quality considerations, specified from a worker perspective, to bear on those choices. But they found that even a strengthened voice in those choices could not affect the quality of work significantly insofar as the scope for doing so was limited by all the available alternatives. Thus, unions were driven to seek ways of assuring that work-quality considerations were already incorporated in the very design of the technologies that became available. The urgency attached to this imperative was intensified by the fundamental technological change being precipitated by the computer revolution. Computerization was recognized as having widely diverse potentials for enhancing or diminishing the quality of work, depending on the values that were incorporated in the design of computerized systems. To press their work-quality objectives effectively, then, the unions thought it had become particularly important to find ways to influence the design of computerization.
The tasks to which the unions were thereby drawn are difficult ones for them, for the tasks are very different from those which unions have historically carried out. Unions have been geared primarily to bargaining over wages and hours and providing some protection against arbitrary exercise of managerial authority—maintaining a modicum of due process or “constitutionalism” in the workplace. As a device for organizing joint action by workers, they have served as a means by which workers could in some measure offset and limit the power conferred on managers by the command over productive resources institutionalized in the organization of the firm. Their effectiveness has depended on their ability to turn numbers into power; the essential skills have been those of mobilization and negotiation, including expertise in the substance of negotiation over such matters as the details of wage scales and contractual rights. Their members could be the source of the requisite skills as well as the solidarity on which bargaining power rested, even as the skills came to be embodied in full-time union staffs.
Swedish unions have amassed an exceptional amount of power in terms of such resources. Their members comprised about 85 percent of the labor force as of 1983. Their capacity to mobilize this membership is facilitated by a high degree of internal centralization. It is also enhanced by the concentration of the membership in relatively few unions organized primarily along industrial and sectoral lines, although within these categories the unions are divided on broad occupational lines. Thus, there are three sets of unions, affiliated to three separate confederations. The largest and oldest is the Swedish Confederation of Labor (Landsorganisationen i Sverige, or LO), consisting of twenty-four unions organizing largely blue-collar workers and accounting for 52 percent of the labor force. The second largest is the Central Organization of Salaried Employees (Tjänstemännens Centralorganisation, or TCO), consisting of most of the unions organizing white-collar employees and accounting for 26 percent of the labor force. The smallest of the three is an organization of unions organizing upper-level white-collar employees with university degrees in their professions (SACO/Sr) and accounting for 6 percent of the labor force.5
The confederations are also divided along political lines. Throughout its history, the LO has been politically identified and organizationally linked with the Social Democratic party, while the other two confederations are formally nonpartisan, embracing memberships with political loyalties spread across the whole political spectrum. Although the very high proportion of the labor force in unions consequently does not translate into a corresponding degree of electoral support for the Social Democrats, it goes a long way toward explaining why they have been able to control the government from 1932 to 1976 and again since 1982.
Extended Social Democratic rule has provided Swedish unions and particularly the LO with a political alternative to bargaining as a way of advancing their objectives. Although the unions have avoided resort to that alternative with respect to wages in the interest of preserving their autonomy in this central function, they have turned to it in connection with other objectives, including pensions and, more importantly for our purpose, union power to press workplace issues. Those issues had been kept beyond the reach of collective bargaining by the correspondingly encompassing and centralized private sector employers’ organization, the Swedish Employers Confederation (Svenska Arbetsgivareforeningen, or SAF), which had long insisted that all issues concerning the organization and content of work fell within the exclusive domain of managerial prerogatives. Unable to overcome this barrier through negotiation, LO and TCO jointly succeeded in getting it removed by legislation in 1976. That was the culmination of a series of laws enacted over the preceding years that strengthened union and worker rights with respect to health and safety, job security, and position of local union officials.
Reflecting the unions’ power in the Swedish political arena, then, the legislation of the 1970s opened up the way for unions to utilize their bargaining power in the market arena to press their work-quality objectives. It proved difficult for the unions to convert their newly won right to bargain over workplace issues into agreements regulating them in new ways, however, for in the changed political context defined by the shift in control of government from the Social Democrats to the three so-called bourgeois parties from 1976 to 1982, SAF felt able to resist the demands for a central agreement by which the general principles embodied in the legislation were translated into more detailed guidelines for industry and company level practice and for which LO and the negotiating organization of private sector white-collar unions Privattjönstemannakartellen (PTK) jointly pressed. But even if this source of difficulty could be overcome, as it was to some extent when a central agreement was finally reached in 1982, there remained and continues to remain another, in some respects more fundamental, source of difficulty.
The resources on which the unions’ power in both the political and market arenas has been based, and the organizational forms for deploying those resources, do not equip them adequately to carry out the tasks involved in pursuing their work-quality objectives, particularly insofar as they are contingent on being able to influence technological development. To do so, it is essential to draw upon scientific and technical expertise, which is not the kind of expertise that has primarily been built into union staffs, whose size is in any case limited. Nor can the membership be a source of such expertise except in the case of unions organizing scientific and technical workers—even if workers’ skills and experience remain a relevant source of knowledge about how specific kinds of work can be done.
On the other hand, the scientific or technical component of the relevant expertise, whose importance is growing as technologies such as those involving computers become increasingly science based, is precisely the kind of productive resource over which management is given virtually exclusive command by the organization of the firm. This gives management an overwhelming advantage over unions in defining the requirements that technological development is compelled to meet. That advantage can be offset only insofar as unions gain independent access to such resources and, in addition, insofar as both unions and those possessing the scientific and technical expertise learn how to utilize it for the unions’ distinctive objectives. At a minimum, this requires the addition of such expertise to union staffs simply to enable them to tap external sources of expertise effectively. But the latter can never occur on a sufficient scale unless the way in which access to scientific and technical resources is institutionalized is changed—that is, unless the financing of research and development is changed in ways that permit unions to autonomously define the requirements to be met by technological development.
Having come up against this fundamental source of difficulty, the efforts of the Swedish labor movement to work out a strategy for influencing technological development have been focused on overcoming it by change both in their own organizational practice and in public policy concerning the social distribution of access to scientific and technical resources. This essay describes some aspects of these efforts.
The first of the two parts of the essay is an account of the evolution of the policies of the Swedish labor movement concerning technological change. The forces underlying the evolution are referred to only briefly. Instead, the discussion concentrates on the rationale provided for the policies in the documents in which they are set forth. Because the rationale is typically presented in some depth, it is possible to see how the relationship between technological change and union objectives has been understood as the objectives themselves have changed, culminating in the importance now attached to work-quality objectives.
To illustrate the challenge posed by the pursuit of those objectives, the second part of the essay is a case study of a technology development project that constitutes an especially promising approach to meeting the challenge. At the same time, the project, evocatively referred to as the UTOPIA project, also points clearly to difficulties that have their source not only in the social distribution of access to scientific and technical resources but also in conflict over changing work roles among different occupational groups built into the structure of the trade union movement.
The Evolution of Trade Union Technology Policy
The position of the Swedish labor movement on technological change has been shaped largely by the interaction of two concerns. One is the role of technological change in maintaining the economic basis of full employment and the welfare state. The other is the impact of technological change on the character of work.
The first has been an essentially constant factor throughout the history of the labor movement, particularly in the postwar half-century. The labor movement has relied heavily on the growth of the economy as a whole and on government action to stabilize that growth and allocate a large part of its fruits to advance its members’ economic interests. This national, macroeconomic political orientation reflects the large portion of the labor force that the unions organize and the close links of the largest union confederation to the normally ruling party, along with a strong awareness of the vulnerability of Sweden’s small, open economy to international competition. From this perspective, technological change has been viewed as an indispensable ingredient in the international viability of the economy, so that it is to be not only accepted but also encouraged.
The second factor is one that has been changing as the labor movement has given increasing attention to the quality of work. This is illustrated by the successive widening of the movement’s conception of its occupational health and safety objectives. Initially confined to protection against physical injury in the narrow sense, the conception has gradually expanded to embrace psychological health in very broad terms, including not only protection against hazards such as stress and isolation but also provision of conditions for personal growth through participation and skill development. Stemming from a variety of sources, as suggested above, the increasing concern with the quality of work has evoked an increasingly critical attitude toward technological change and particularly toward the basis of management decisions determining the specific directions it takes. This has not led the Swedish labor movement to abandon its support of technological change, which is more than ever viewed as an imperative of the economy’s viability. Instead, unions have sought to reconcile the requirements of that imperative with the requirements of the other objectives that have become increasingly salient. In the process, the terms on which technological change is acceptable have been redefined, becoming successively more stringent.
The resulting evolution of the position of the labor movement can be schematically broken down into four phases, each characterized by the assertion of a new condition for supporting technological change that is added to those set in preceding phases. The conditions marking each phase can be summarized as follows:
• manpower policy so designed and on such a scale as to enable all workers displaced by structural and technological change to get alternative jobs;
• industrial policy capable of channeling financial and technical resources into industrial expansion sufficient to assure new jobs to replace those eliminated by structural and technological change;
• extension of collective bargaining to workplace and enterprise decisions that affect workers’ interests concerning the pace and composition of structural and technological change and the quality of jobs that result;
• union influence on the development of new technology that determines the scope for choice concerning the characteristics of jobs resulting from structural and technological change.
The evolution of trade union policy has thus been marked by an expansion in the scope of the conditions on which acceptance of technological change has been predicated. The successive conditions have also involved an increasing role in the process of structural and technological change for the state and, more recently, for the unions themselves.
In describing this evolution in trade union technology policy, the following discussion will draw primarily on the positions taken by the LO. The focus on LO policy rests on its dominance in terms of ideas as well as power over the postwar period as a whole. Thus, TCO’s policy has generally developed along the same lines as LO’s, although the multipartisan composition of its membership has compelled it to avoid positions identified closely with Social Democratic party policy. It is also only in more recent years that TCO has made technology policy a major focus of attention. It has now become equally active in this area, cooperating extensively with LO at the peak organization level, although the erosion of old demarcations between blue- and white-collar jobs by technological change has generated some jurisdictional conflict between LO and TCO unions.6 With respect to union participation in technological development, TCO has to some extent taken the lead, particularly where its membership gives it access to the relevant technical resources that have become crucial for union participation. There is, of course, an even higher concentration of technical expertise in SACO, which has also developed a position on technology policy. However, the technical expertise of SACO’s membership comprises the resources for research and development (R&D) addressed to problems defined from business and academic rather than trade union perspectives. How such resources can be mobilized in the service of the latter is primarily a problem with which LO and TCO are concerned.7
Phase 1: Facilitating Displaced Workers’ Transition to Alternative Jobs
Manpower policy, or labor market policy as the Swedes prefer to call it, was specified as a condition for accepting structural and technological change within the framework of an overall strategy for managing the economy in which such change was viewed as a central feature. This strategy was worked out in the late 1940s and early 1950s by two LO economists, Gösta Rehn and Rudolf Meidner, and is referred to as the Rehn-Meidner model, or simply the Rehn model. LO adopted the strategy as it was set forth in a document presented to its 1951 Congress, The Trade Union Movement and Full Employment.8
The Rehn model was designed as a means for attaining noninflationary full employment without state intervention in the wage bargaining process, and thus as an alternative to state-run incomes policy. In addition to labor market policy, it had two other main ingredients. One was a moderately restrictive fiscal policy, aimed at what subsequently came to be called a full employment surplus, so that fiscal and international competitive pressures would be combined to make it hard for employers to pass on increased costs in increased prices. The other was a standard rate, or “solidaristic,” union wage policy aimed at enforcing equal pay for equal work, regardless of employers’ profitability. The combination of such fiscal and union wage policies was expected to subject firms to a differential profit squeeze, depending on how efficient and profitable they were. By favoring firms that could expand out of profits while forcing firms that could not do so to improve efficiency or else contract or go out of business, continuous change in the structure of industry would be stimulated. The proportion of low-cost, competitive firms able to pay standard rates would consequently increase relative to high-cost, uncompetitive firms that are not able to.
The third ingredient in the strategy, labor market policy, was intended to help workers who lost jobs in firms forced to contract or go out of business move to jobs in firms able to expand. Differential wage rates to induce workers to move from the former to the latter were ruled out by both the equity and efficiency objectives of solidaristic wage policy. Moreover, reliance on differentials would impose on workers subject to job loss in declining firms additional costs and put them at a disadvantage in competing for jobs in expanding firms insofar as the skill requirements and location of the jobs were different. It was accordingly held necessary to provide information both about new jobs and about various forms of support to assist workers in making the transition from old to new jobs—for example, providing retraining and relocation assistance. The associated costs of the structural change was supposed to be thereby shifted from the individual workers affected to the society as a whole, which benefited from the change. An “active” labor market policy of sufficient scope and scale to accomplish these objectives was accordingly viewed as an essential condition for acceptance of the structural change involved in the strategy.
What distinguishes this condition from those in the subsequent phases in the evolution of trade union technology policy is that it involved no direct intervention by either the state or unions in the process of structural and technological change itself. It was assumed that the process could be left to itself or, more precisely, to management at the enterprise and workplace levels and to the R&D performers who developed the new technology on which the growth of productivity depended. In other words, as far as the process itself was concerned, all that the combination of state and union policies was designed to do was keep up the pressure on management to be efficient. Under these conditions management was assumed to be willing and able to adopt rationalization measures, including new technology, on a scale that was sufficient from the overall economic perspective from which the strategy was devised. The impact of the measures on the quality or volume of jobs was not in question, nor was the capacity of the R&D system to come up with new technology at a sufficient rate. This confidence that the process could be left to itself, which prevailed in the 1950s, was eroded during the 1960s, leading to the crystallization of the second phase in trade union technology policy.
Phase 2: Assuring Employment for Displaced Workers
Industrial policy came to be viewed as essential to effective implementation of the structural change strategy by the later 1980s. An important early expression of this position was a report presented to LO’s 1961 Congress, Economic Expansion and Structural Change.9 It analyzed the policy problem as one of increasing the economy’s “adaptability,” emphasizing the need for mobility of capital as well as labor. To improve capital mobility, the report prescribed a variety of measures to reduce obstacles and establish new channels for it, along with new planning institutions and procedures to coordinate these measures with macroeconomic policy over a longer time horizon more commensurate with the structural change process than the short-term perspective typical of stabilization policy.
The analysis was predicated on the disappearance of the “very favorable external conditions for the Swedish economy prevailing in the preceding postwar period.” As a result, “Sweden with its limited domestic market and significant export dependence . . . will in all certainty require greater exertion, more intensive investment activity, and a more rational utilization of our resources” in order to continue to be a “welfare society.”10
The 1961 report was not a policy document, however, and LO did not adopt the position it set forth until after the middle of the decade. It was only then that the difficulties of implementing the structural change strategy effectively in the face of increasing international competition became palpable, precipitating political reactions that compelled some response. In particular, there was a growing reaction against the costs imposed by the high geographical labor mobility associated with the strategy. Among those costs were some that were not compensated despite massive expansion of labor market policy, and others that had not been anticipated or recognized as costs. Such costs received much greater attention in a report to LO’s 1966 Congress, Trade Unions and Technological Change.
The report observes that “the costs associated with the fraction of labor turnover that was not ‘voluntary’ on the part of the individual worker” includes
not only the actual costs of removal but also the losses sustained in the process of disposing of old and acquiring new housing, the interruptions in the children’s school attendance, foregone non-wage benefits, reductions in income or additional periods of training. To these must be added the social and psychological sacrifices often involved in a change of environment and adjustment to a new social setting. At the same time, there are costs incurred by society, both by the depopulated local authorities in the form of lost economic and population bases for their social services, and for recipient local authorities which . . . must create new services.11
Nevertheless, the report rejects the conclusion that “the structural transformation of industry should be halted or slowed down. We urge instead that our national labor market policy program, which is at present ineffective in dealing with this problem, be expanded to the point where it can have a significant impact.”12
Thus far, the argument was not that labor market policy was inadequate but only the scale on which it was implemented. But the report went on to argue that even a much expanded labor market policy would not suffice, and reiterated the contention of the 1961 report that an “active industrial policy” was necessary as well. LO made this its official policy at its 1966 Congress. The shift in LO’s policy was paralleled in the Social Democratic party, in which it was given a powerful impetus by a sharp setback in local government elections a few months after the LO Congress. This was widely attributed to a reaction against the rapid and unstable economic growth that the structural change strategy fostered but could not control. To establish such control, new instruments for implementing an industrial or, more precisely, economic development policy were believed necessary. Joint LO and party committees were set up to formulate the policy. With a national election due in 1968, however, the Social Democratic government felt compelled to set up a cluster of industrial policy institutions without waiting for the committees to report.
The first of the institutions was a State Investment Bank to provide long-term capital for investment projects, preferably in areas of advanced technology. To plan the policy and coordinate the institutions being created for it, a new division of the Ministry of Finance was set up and then turned into a separate Ministry of Industry. A holding company, State Enterprise, Inc., was established to take over a miscellany of state-owned firms and gear them to policy goals. In addition, various forms of government support for technological R&D were brought together in a new Board for Technical Development.
The rationale provided for these measures by the LO-party committees displayed a shift in emphasis when compared with LO’s 1961 document, although most of the measures were foreshadowed by the latter, which had in fact proposed to go further. Though the earlier document stressed the removal of hindrances to factor mobility, the later ones stressed the importance of establishing conditions necessary to make structural and technological change compatible with security and justice, as well as environmental protection and other values in which social costs and returns diverge from the private costs and returns of individual enterprise. More explicitly than before, “technical and economic development . . . formed in large part . . . in other countries that we are dependent on” were recognized as posing threats to these values as well as being a source of increased material living standards.
Economic development presents us with the possibilities but also the demands of a new era. . . . But the promise they give of better living conditions for people in the future can only be realized if we have the political will to exploit the possibilities and overcome the difficulties that development creates.
Full employment and a secure income are not given when industrial life changes. If economic forces are given free play, restructuring will be accompanied by new economic and social injustices. Therefore we must in our policies constantly work to unite economic progress with security and justice for individuals.
Although resistance to change is declared to be self-defeating, as in the earlier document, the necessity of controlling change is stressed.
If we should try to create security by resisting the development . . . by clinging tenaciously to what there is today, the world around us will nonetheless change. . . . The longer we allow us to be lulled by the illusion that we can achieve durable security that way, the harder the awakening will be. . . . Our security lies in development, in change. But the development must be steered so that it is of benefit to all. Therefore we a need a conscious and active industrial policy, inspired by our fundamental outlook and our demands for equality and security.13
As suggested by the measures adopted, the industrial policy program relied primarily on expanding the role of the state. There is no more than an allusion to expanding the role of the unions. This is in connection with a brief discussion of “industrial democracy” as a means for enabling workers to influence decisions affecting work environment and conditions at an earlier stage and at all levels of the enterprise. The need for this had also been pointed to in Trade Unions and Technological Change. As emphasized there, however, a major obstacle to movement in that direction was posed by the SAF’s insistence on defining all decisions about who does what with what technology as managerial prerogatives beyond the scope of collective bargaining. This position was formally enshrined in the well-known paragraph 32 of SAF’s constitution, which required a clause to that effect in all collective agreements. In the words of the 1966 LO report, “Any consultation on an even footing can hardly take place as long as the clause concerning management’s exclusive right to manage and allocate labor as well as freely hire and dismiss workers remains in the collective agreement.”14
By the time its next Congress met in 1971, LO had decided that it was imperative to eliminate the obstacle posed by paragraph 32.
Phase 3: Extending Collective Bargaining to Workplace Decisions
In a policy statement entitled Industrial Democracy, adopted by LO at its 1971 Congress, extension of collective bargaining to issues concerning the organization of work was declared to be essential to assure forms of organization “which can take advantage of modern technical effectiveness without being burdened by its [social] disadvantages.” This marked a major shift in LO’s position. It now demanded a direct union as well as state role in the process of structural and technological change, which had earlier been left up to managerial decision making.
Various pressures for this shift were building up in LO over a long time, but they intensified in the later 1960s. There was a growing conviction that the rationalization of production by technological and structural change was imposing greater costs on workers than had been anticipated or than could be compensated by labor market policy, no matter how “active.” By intensifying the pace of work, creating new and often unperceived physical dangers, increasing stress and social isolation, and in other ways, the rationalization process was now seen as hurting workers who retained their jobs in firms keeping up with technological change as well as workers who lost jobs in firms that failed to keep up. Increasing labor turnover as well as absenteeism and recruitment difficulties were interpreted as reactions to the process, while complaints filtering up through the unions and repeatedly expressed in motions at union congresses pinpointed many specific sources of dissatisfaction. In addition, surveys by LO and some national unions recorded its content and widespread prevalence.
One consequence of all this was a changed conception of occupational health and safety policy, as reflected in a joint LO-Social Democratic party “action program,” A Better Work Environment, in 1969 and a policy statement, The Trade Union Movement and the Work Environment, adopted at the 1971 LO Congress.
Policy in this area had traditionally been confined to regulations aimed at reducing industrial accidents and well-known occupational diseases, but the 1969 and 1971 policy statements describe responsibility for the work environment in much broader terms. Besides extending the scope of policy to a wider range of physical risks, including new ones created by technological change, it extended the scope to psychological risks as well, so that occupational health was viewed as including satisfaction from work. Moreover, the limitation of policy to measures designed to avert risks associated with production equipment and organization once they were in place was rejected. Instead, it was declared essential to introduce work environment considerations at a much earlier stage, when the equipment and organization were planned. Nor could work environment policy be treated in isolation from other kinds of policy. Its requirements had to enter into the “formation of industrial policy, labor market policy, and structural change within industry.” But it was “not enough” to assure “better attention and integration of work environment questions” in business and government policy formation. It was also necessary to “analyze the basis of technical development and the rationalization process in order to determine the conditions that have to be satisfied in future development from a trade union point of view.” Thus, it was “urgent to utilize known technology and develop new technology” to improve the work environment. Moreover, the division of responsibility for providing resources for developing that new technology had to be more clearly defined, including the “degree and manner in which STU [the government R&D support agency referred to above] should be assigned a more active role in the effort.”
Steps in this and a variety of other directions were cited as necessary. Particular emphasis was placed on legislation to strengthen the position of local union officials to press work environment demands. Union power in the workplace was viewed as an indispensable condition for achieving the expanded goals defined for work environment policy. Thus, LO’s 1971 work environment statement concludes by saying that
the decisive question is to what extent the trade union movement can create the possibilities for influence by employees in the planning and decision-making process. It is a vital task to work for increased union influence through a successive advance of our positions in bargaining with our counterpart. There is no doubt that precisely the work environment area is especially appropriate for deepened industrial democracy.15
The adoption of more ambitious work environment goals was thus an important factor contributing to LO’s determination to overcome the obstacle to collective bargaining over workplace issues posed by the official employer stand on managerial prerogatives. This factor was powerfully reinforced by the unions’ response to the unprecedented wave of wildcat strikes in the winter of 1969–70. The strikes demonstrated a loss of authority by the unions that they felt compelled to repair. To do so, it was believed essential to strengthen the unions’ capacity to protect the interests of their members at the workplace, where most of them had their only direct contact with the unions. At the same time, abandonment of centralized wage negotiations (which was one source of discontent) in order to give unions more to bargain about at the local level was ruled out by LO’s long-standing organizational and economic strategy. Accordingly, it was all the more necessary to make it possible for unions at the local level to negotiate over nonwage workplace issues.16
Since that objective was believed unattainable by negotiation, LO called for legislation, which was eventually enacted in the form of the Law on Joint Determination (Medbestämmandelagen—MBL) in 1976. This was the culmination of a whole series of labor laws enacted during the preceding half-decade. Together, they enlarged union power at the workplace, especially with regard to the work environment, imposing stringent restrictions on the right of management to dismiss workers and obligations on management to negotiate over changes in the production process, thereby nullifying SAF’s paragraph 32.
Although the legislation was not regarded as satisfactory in all respects, it was seen as removing a major obstacle to more effective pursuit of the more ambitious work-quality goals to which unions had become committed. As pointed out in the report The Work Environment, presented to LO’s 1976 Congress, however, reconstruction of the legal framework was “only a first step.” To realize the possibilities thereby opened up would take much “difficult and laborious” effort. The focus of attention was accordingly shifted to specifying what had to be done and building up the requisite capability. Out of this has crystallized the growing demand for greater union involvement in the process of technological R&D, which characterizes the fourth, current phase in the evolution of union policy on technological development.
Phase 4: Extending Union Influence to Technological Development
Union influence on technological development has come to be regarded as essential to the achievement of union goals concerning the quality of work—its environment, organization, and content. This position is predicated on the conviction that there are severe limits on what can be accomplished to achieve those goals at the stage when negotiations take place over the introduction of new technology that has already been developed and adopted. By then, the scope for choice has been sharply narrowed. Therefore, those goals have to be pursued at much earlier stages. Unions must be able to influence not only the whole process of planning changes in production within the enterprise from the very beginning, but also the preceding R&D process by which the technological options that become available are themselves shaped. At the same time the exercise of such influence is recognized as a largely new function for unions, and one they are as yet poorly equipped to perform. Building the capability for doing so is consequently given high priority.
This position has been elaborated on in increasing detail since the mid-1970s. As the challenge of realizing the possibilities opened up by the legislation in the preceding half-decade was confronted, LO’s understanding of the tasks involved, including their formidable scope and magnitude, was enlarged, and the tasks were spelled out with increasing concreteness. The main directions in which LO’s position has developed were set at its 1976 Congress, soon after the enactment of MBL. Several reports to that Congress took stock of what had been accomplished and what remained to be done.
The report on the work environment cited earlier emphasized the need for local union action to assure compliance with existing regulations and agreements, not least with respect to known risks and known safeguards against them. As far as unknown risks were concerned, local union action was said to be hampered by serious shortcomings in the responsible authorities’ effectiveness in identifying the risks and prescribing safeguards. This was all the more serious in the face of new risks generated by rapid technological change, as in the case of chemicals and computers. The regulatory process, with its traditional approaches and inadequate resources, was declared unable to keep pace with the effects of technological change.
Even keeping pace would not be enough, however, for averting risks before they are built into the production process was a much more effective way of creating a good work environment than trying to remedy them afterward. “It is in connection with the construction of machinery and equipment, planning and forming work processes and places, that the work environment’s conditions are created.” Therefore, it is “already in the investment and planning stage” that the requirements of a good work environment have to be incorporated. To assure that they are, workers must be in a position to see that they are incorporated from that stage on. For workers to do so, they must participate right from the beginning in project groups set up within or by the health and safety committees prescribed by law, assisted by experts of their own choice, paid for as part of the firms’ planning and investment.
Even this would not suffice, however. The solutions that can be arrived at within firms, including the availability of substitutes for dangerous products or processes, depend on the extent to which R&D outside as well as within the firms is addressed to the problems for which workers seek solutions. “As consciousness of work-environment questions has grown and the possibilities for influence strengthened, so too has the workers’ need to be able to steer research toward the work environment problems experienced as most central.” While “the trade union movement . . . has certain possibilities for directing and influencing research in the work environment area,” research is still in “too high a degree steered and initiated by employer interests.” Therefore, the trade union movement’s “influence must be significantly built up and strengthened.” It must “provide itself with better resources to articulate and specify its needs and desires” and be “able to attach researchers to itself to a greater extent” than it has been able to.17
The conviction that greater trade union influence was needed across the whole spectrum of R&D, expressed as well in a report entitled Education for Work and Democracy, led to a decision to work out a detailed research policy program. A committee, referred to as LOFO (the initials of the Swedish words for LO’s research policy), was set up to prepare such a program for presentation to LO’s 1981 Congress. A special effort in the area of computerization was also initiated. Stressing the importance of computerization for the possibilities of improving the work environment and democratizing work, another report, Solidaristic Joint Determination, recommended that an LO computer council be established to focus union efforts in this area. Such a council was set up and called upon to report to the 1981 Congress as well. The documents prepared by LOFO and the computer council are major statements of LO’s fourth phase position. Each of them will accordingly be discussed in turn.
LO Research Policy
In its 1981 Congress report, The Trade Union Movement and Research, LOFO proposed a research policy addressed to the tension it perceived between the promise and threat of technological change. It grounded the policy in the trade union movement’s basic goals of “full employment, a good work environment, and a democratic working life.” These goals, the report warned, are threatened by the economic crisis and the fundamental structural changes in the international economy that underlie it. The possibilities of defending the progress toward the goals already attained and of realizing them more fully depend more than ever on research (the term being used to cover the whole R&D spectrum from basic science to industrial product and process development) through the technological development it permits and the economic and social policy solutions it yields.
It is therefore more important than ever that the trade union movement engage itself in research . . . by increased union influence on the planning and direction of research and by widened dialogue with researchers. Union organizations must have the possibility of defining the questions and problems that have to be the object of research in order to be solved.18
The report recalled the labor movement’s traditional view of “research and technical development as essential and necessary means for raising society’s material and cultural level” and its stress on freedom for research as a condition for assuring its quality. At the same time, the labor movement has understood that the directions of research and the questions it addresses are “strongly influenced by power relations and values in society.” In particular, power is exercised over research by those who decide which research is to be financed. In industry, this power is concentrated in the management of a relatively few large firms. Management’s power determines not only what research is done in firms but influences what research is done in technical universities as well, through the research for which it contracts, by being “well represented in state organs that allocate funds and decide on research staffing” and through effective pressure on the government and parliament. Business influence extends to the social sciences, particularly economics, as well as technical and pharmaceutical faculties. “It is business questions that stand at the center and it is business that forms the future labor market for students and researchers.”
Cooperation between industry and the universities was nevertheless acknowledged as essential. Indeed, LOFO argued, it should be increased, for university resources must be utilized for the industrial renewal that is crucial. However, cooperation between industry and the universities “should not take place solely on industry’s terms.” “We need development, but not at any price, not at the cost of the environment and people’s health and well-being.” Development must be “meaningful” and be “distributed justly.” For the trade union movement, therefore, research on questions of how to make development meet those conditions is urgent. Yet research addresses those questions “far too little.” One reason is that the trade union movement has far too little influence in the country’s research organization; hence, it has to secure greater influence.19
To do so, the report stressed, the trade union movement had to acquire the capacity to act independently in defining research problems and bringing research resources to bear on them. What the unions seek is not simply or primarily greater participation with others on research addressing problems they have in common. Much of union experience of research has been with efforts carried out jointly with their employer counterparts, as provided for in laws and collective agreements such as those concerning the work environment. But such “two-party common interest” research limits the scope for questions the unions are concerned with, for it precludes research employers do not regard as in their interest. “It is research on our terms, from our perspective, and with union problems and goals that the trade union movement wants to get done.”
Within the framework of the research organization we have and the researchers who work at the universities, different research institutes and out in firms, the main aim is to bring about changes in the directions of research and to influence the questions one chooses to work with. It is, so to speak, to bring about [union] interest-oriented research within a research system that is in large part already steered by other interests, not least from the employer side. What is new and unusual is that it must be to a greater degree the worker side that impels research.20
This, the report insisted, was not a call for research that was itself biased. “Research does not become inferior, viewed as research, when it proceeds from clear values or is conducted to solve specific problems.” For the research to be useful, however, its integrity and quality must be upheld. “No one has any use for deficiently performed research. Bad research is worse than none at all, because it is misleading and lays claim to valuable resources.” The unions’ interest in the utility and quality of research is best served by clearly formulating the problems they want studied and taking responsibility for applying the findings, while placing their confidence in researchers to freely carry out their distinctive task of developing knowledge to the best of their ability.
Informed by these general considerations, the LOFO report discussed a broad range of research policy issues and presented a number of specific proposals. It called for a change in the whole way in which government conducts research policy, proposing the appointment of a new cabinet minister in whom responsibility for a “coordinated research policy” would be vested. This was proposed to remedy serious deficiencies seen in the way research policy was conducted. Many of the deficiencies were attributed to the growth of a “sectoral” approach, by which support for research was divided up among various ministries and agencies responsible for specific policy areas, such as defense, industry, education, medical care, housing, and so forth. Though this has resulted in resources for research on important policy problems, it leaves no place in the research policy structure for attention to the relationships among the different efforts or to problems that cut across them.
Thus, there has been little coordination of policy for university research with industrial policy despite the critical interdependence of the two. And of special concern to the unions, there has been little research on how to integrate work-quality considerations into technological development, since it requires the combination of engineering and social science resources divided between different governmental bodies with different missions—e.g., STU, which finances engineering R&D on new technology, and the Work Environment Fund (Arbetarskyddsfonden, or ASF), which finances medical and social science research on the effects of existing technology on the work environment.
Such multidisciplinary research was among a variety of areas identified as requiring the initiation of immediate efforts, adding up to an overall increase in funds allocated to research. These would flow largely through existing agencies, but some changes in the organization of the research structure as well as in the funding agencies’ operations were urged. For example, establishment of state development companies to run major projects in selected areas and greater reliance on research institutes independent of both the higher education bureaucracy and industry was recommended. At the same time, the report called for not only greater union participation in research planning and decision making but also the resources required to build up union capacity to exert real influence through such participation as well as to take responsibility for projects addressed to questions unions define.
The new emphasis on union capacity for autonomous research initiatives reflected the limitations ascribed to participation in contexts dominated by other actors and their purposes. In principle, union representation was sought in all the bodies that plan and administer research, including the governing bodies of national funding agencies, universities, industry institutes for cooperative research, and companies, as well as reference groups for particular projects. Such representation was seen as a means of access to information about research that is and is not taking place, an opportunity to exert some influence on decisions about the allocation of research resources, and some possibility for initiating research on problems of concern to unions. But there are drawbacks and risks as well. Representation takes time from the many other functions unions have to perform, further stretching resources already strained. And since there can therefore rarely be more than one union representative in any of the bodies, the lone union representative can easily be put into a hostage situation. Without much influence on the decisions made, his or her presence may nevertheless confer a claim of legitimacy on the decisions, placing on unions a share of responsibility for decisions that may not advance union members’ interests much if at all. While necessary, then, representation can at best be only a “complement to other forms of influence.”
Among the most important of these is “direct cooperation and dialogue with researchers,” even though it is “demanding, requires time, and can also bring with it conflicts.” A broad network of contacts with researchers is required to give the unions their own access to expertise needed to exert any independent influence on research through representation and negotiation. And unions must have direct contact with researchers to initiate research on questions that concern the unions, to be carried out under contract with the unions, in addition to whatever research on these questions can be carried out within the framework of ongoing programs in which others are involved.
The main limitation on such an autonomous union research policy was seen in the scarcity of resources, particularly of staff with competence and experience relevant to implementing the policy. This imposes on unions the need to be selective with respect to both the bodies on which they seek representation and the research they seek to get done on their own. Finally, it underscores the need for public funding, which means that research addressed to problems defined by the unions as well as the development of union capacity to define the problems and administer the research must become part of national research policy.21
The general position elaborated in the LOFO report was applied specifically to computer technology in the computer council’s report to the 1981 LO Congress, Union Computer Policy. In it, “computer technology’s special possibilities and threats” were viewed from the perspective of the broad aims of the labor movement, reiterated as “work for all, democracy in working life, good work environment, higher standards, and continued expansion of collective services.” The long-standing belief that technical development is “one of the prerequisites for achieving better and more human conditions” was reaffirmed yet again, along with its insistence on “a more just distribution of the increased material resources that could be created with the help of the technology.” The economic crisis facing Sweden now made “the need for technical development” greater than ever. “A stagnant industrial development—attributable in part to technology that is obsolete compared with that in firms in other countries and to declining investment—constitutes today the greatest threat to the possibilities of achieving these aims.”22
On the other hand, the report emphasized, the trade union movement is no longer willing to confine itself to the distribution of the economic surplus resulting from the “technical development that took place largely on private industry’s terms.” The role now claimed by the movement was set in the context of the evolution of trade union policy reviewed earlier. Referring to the reaction against the social costs of structural change during the 1960s and 1970s, the report cited the growing “demand for economic democracy . . . and for a direct influence over decisions on investment and the shaping of new factories,” and to the legislation enacted in response to those demands. Now,
the whole labor movement . . . confronts a new and still greater challenge. [It] must take the initiative for an industrial development such that Sweden’s possibilities as an industrial nation are not destroyed. But it is as important to create the possibilities and instruments [needed] to determine which technology shall be developed and how it shall be utilized in practice, which thereby decides also the possibilities for achieving our demands for content-rich and meaningful work, solidarity, community, justice and a good work environment.23
The challenge is posed most sharply by computer technology because of its fundamental, all-pervasive character. Moreover, since the essence of the technology is the command over information it provides, and since “those who have knowledge and manage information possess power,” the technology is an instrument of power with great potential for those who can control its development and use. It therefore has a “dimension and significance for the future development of society different” from earlier technological change, affecting “the conditions for political democracy as well as economic democracy and the democratization of working life.” Thus, the development of computer technology opens up both enormous possibilities for and grave threats to the realization of labor movement aims. Which of these directions its development will take will depend on who controls it. Because it is now to “such a high degree controlled by private owner interests,” as preceding technological development has been, computer technology is being “developed in the wrong directions in decisive areas.”
This pessimistic outlook was elaborated primarily with respect to worklife, contrasting the way the technology is being developed with the way it could be. Because computers have been harnessed primarily to management’s efforts to “centrally direct and control” the work process,
the dominating experience of computerization today is that it brings to most of those affected an impoverishment of work content and tasks while new risks and health problems arise in place of those that are eliminated. Separation between people in production grows. Work contacts are replaced by conversations over computers. Central planning becomes more detailed. Isolation in work life is thereby increased. Control becomes all the more effective. Instead of democratized management, increasingly effective systems for giving orders . . . are built up. This is a development that seriously threatens the trade union movement’s and democracy’s fundamental values.24
Computer technology offers entirely different possibilities, however. It can give workers “an improved overview of the work process and thereby continuously increase and develop their knowledge.” The information it can make available can facilitate the organization of production by autonomous work groups, in which the content of work is enriched by responsibility for planning, rotation among tasks, and the combination of new computer knowledge with “traditional occupational skills.” Computer technology can thereby serve as “the workers’ tool” rather than a “control system dominating the workers.” For business strategic decisions as well as detailed production planning, moreover, the speed with which computers permit information to flow enlarges the possibilities for democratic participation.
While development of computer technology in this alternative direction is a “more effective way of exploiting its possibilities” for production as well as the quality of work, it is likely to occur only if “wage earners’ influence increases.” But achieving such influence is extremely difficult in the face of the powerful forces dominating the development of computer technology and “contemporary power relations in industry.”
Very few companies, those in which computer R&D and production is highly concentrated, above all IBM, dominate the direction of development. Powerful influence is also exerted by major purchasers, whose demands the suppliers strive to meet. Among the most important purchasers are governments engaged in the arms race. Also important are “the few large multinational enterprises that dominate each branch [of industry] and shape the technical development and forms of computer applications . . . according to their needs and specifications,” providing models for computer application by other firms. Since these multinationals also operate in countries where labor movements are politically weak or entirely suppressed, the models are likely to reinforce the development of computer technology in the wrong direction.
While the basic directions of computer technology development are therefore largely beyond the reach of Swedish unions, even the adaptation of systems to Swedish firms is difficult for the unions to influence, for the instruments placed at their disposal by legislation and collective agreements are insufficient. Thus, MBL does not provide for influence on development that takes place in a large conglomerate’s subsidiaries separate from those in which the systems are to be introduced. Even when development takes place in the latter, workers and their representatives enter into the development process “too late to influence the development to a decisive extent.” Given the formidable obstacles to increasing worker influence on the development of computer technology, “the circumstances can seem to be extraordinarily difficult. But this makes it all the more urgent to immediately take on the problem, giving priority to concentrating political as well as union resources on it.”25
The report went on to suggest how to proceed. To begin with, legislation and collective agreements must be changed to enable unions to enter the design process at the stage at which a firm first considers the introduction or modification of computer systems. This also “requires a significant strengthening of unions’ resources, including the availability of their own experts to develop their own alternatives and provide an informed basis for evaluating and negotiating over management proposals.”
Such “trade union methods do not suffice,” however. To provide the economic and industrial policies on which the effectiveness of trade union methods depends, as well as to create the conditions under which they can be applied in the first place, requires action at the national policy level. This is so with respect to many additional dimensions of the challenge of computerization, such as its effects on employment.
Two sharply contrasting beliefs about those effects were cited. One, resting on the long-run growth of employment despite repeated technological transformations in the past, is that “computer technology will—more or less automatically—create full employment.” The other is that computer technology “will lead to mass unemployment.” While stressing the impossibility of determining with certainty which is correct, the report rejected both. The first fails to recognize that the far-reaching structural transformation that computerization precipitates cannot take place on socially acceptable terms without “strong action of an organizational kind, to assure that the social costs are borne equitably and that profits resulting from productivity gains are reinvested in new jobs.” The second fails to recognize the possibilities for coping with the problems posed by computerization or the potential it offers for enlarging the resources with which to fill social needs. Moreover, it overlooks the even greater threat to employment if the technology needed to survive in international competition is not adopted. Accordingly, there is no alternative to adopting the new technology, but there are alternative policies that can make a great deal of difference to its employment effects. “The total level of employment and unemployment in the country is determined not by technology but ultimately by how society is organized, the distribution of available resources through collective bargaining and political decisions.”26
The report accordingly concluded by reiterating LO’s fourth phase position.
If the development [of computer technology] is allowed to continue along present lines, the divisions in society would increase. Sweden’s traditionally greatest asset, a broad and deep knowledge among the workers and internationally competent firms, threatens to be weakened. Wage earners would justifiably resist changes that entail worsened conditions. The new technology would not be utilized and Sweden’s leading position as an industrial nation would be seriously threatened.27
The new technology can be used to secure employment and welfare and provide the conditions for meaningful work and a good environment. Cohesion and solidarity can be strengthened. To achieve these requires that we use the possibilities for influence we have better than we do today. But it in addition requires means of control that neither society nor the union organizations have at their disposal. Time is short. Programs and measures have to be put in place quickly.
[Thus] the growth of investment in new technology is in everyone’s interest but wage earners cannot accept it if the conditions are increased gaps in wealth and the exclusion of employee influence over how the new technology is introduced and used. Forms must [therefore] be created to enable Swedish industry and the labor market to be renewed and developed in a democratic way in which changes take place with the participation of the wage earners.28
It is on developing forms for more effectively influencing the development and utilization of new technology generally as well as computerization in particular that the efforts of LO and its affiliated unions have concentrated since its 1981 Congress. These efforts have proceeded in the context of an essentially unchanged public policy framework, with respect to both collective bargaining rights accorded to unions by MBL and funding for union technology development initiatives by state R&D support agencies. What happened within that framework will be just briefly sketched.
Recent Developments
The main event in connection with collective bargaining over technological change was the conclusion of an agreement on implementing the MBL provisions between the private-sector employers belonging to SAF and their two principal union counterparts, LO and the negotiating body for private-sector white-collar workers, PTK, in 1982. Like much Swedish legislation, MBL laid down a set of general principles and guidelines, leaving the details to be spelled out subsequently. In this case, the details were to be spelled out in collective agreements between the employer and union organizations, adapting the principles to the distinctive situations in each of the sectors covered by the respective peak organizations and, with successively greater specificity, from the peak organization to the industry and finally workplace levels. Such joint determination agreements (MBA) were reached in the separate central government, local government, state enterprise, and cooperative sectors in the years immediately following the enactment of MBL. In the private sector, however, negotiations for such an agreement between SAF and LO and PTK were repeatedly deadlocked, broken off, and resumed.
While LO and PTK adopted a common bargaining position calling for provisions designed to reinforce the rights accorded by MBL, SAF rejected the union demands and took positions implying a dilution of those rights. It seemed as if SAF was bent on resisting implementation, confident that the nonsocialist or “bourgeois” parties that were governing since 1976 would not provide the unions with a legislative alternative to negotiations, as the Social Democrats had done. If this was the expectation, it was confirmed, although disarray among the bourgeois parties made them ineffective state arena allies for any employer offensive against the unions in the market arena, as demonstrated in 1980 when SAF failed to win its objectives in the largest lockout since 1909.
Whether it was the possibility of the Social Democrats’ return to office in September 1982 (which is what happened) or satisfaction with the extent to which the union demands had been worn down, SAF finally settled with the unions in April 1982, nearly six years after MBL’s enactment. Without summarizing the terms of the so-called development agreement (rather freely translated into English by SAF as “agreement on efficiency and participation”), it can be said that it did not appreciably strengthen or weaken the unions’ ability to influence workplace decision making. SAF won a lot of language about the importance of “developing” firms’ “effectiveness” for the maintenance of employment. The design of change in technology and work organization so as to utilize and develop workers’ skills was in turn treated as instrumental to effectiveness. The unions won language that reserved their rights under MBL and nailed down their right to bring in outside consultants and researchers to assist in evaluating the information about enterprise performance and plans and to participate as observers and analysts (but not actors) in technological and work organization change projects.
The evidence on how MBL and MBA have worked out in practice is mixed. A recent survey of employers showed them to be generally satisfied with the operation of the law, with very little indication that it had introduced the inflexibility SAF had warned against when MBL was proposed. The survey evidence of the employers’ contentment might be interpreted as showing that joint determination as prescribed by MBL and the negotiated agreements has not made much difference after all. Complaints to that effect have indeed been voiced by some unionists. Such complaints were the basis for motions at the most recent Metalworkers Congress calling for the negotiation of technology agreements that would guarantee more timely information and influence over technological change. But the motions failed in favor of the union leadership’s position that the law and agreements were working reasonably satisfactorily and should be given more time to prove themselves. This presumably reflected the atmosphere of compromise that had been renewed with the conclusion of the private-sector joint determination agreement, which the leadership was probably not inclined to disturb.
A parallel expression of this atmosphere of compromise was a large program of joint union-management computer application projects financed by the government and administered by ASF, the work environment research funding body. The projects, like ASF itself, are quintessentially corporatist—that is, they are two-party common-interest exercises of the kind whose limitations the LOFO report had pointed out. To some extent, the new program appears to be a restoration of the kind of union participation in management-initiated applications of new technology that was fostered by SAF’s technical department in the early 1970s. Dissatisfaction with this had contributed to the unions’ demand for an extension of collective bargaining to workplace issues and subsequently for autonomous union development of technological alternatives. In any case, the joint projects in the ASF program concern the introduction of technology that is available or on the verge of being available rather than the development of new technology.
No substantial change in national research policy in the direction of supporting such development under autonomous union auspices has occurred either. In response to continuing pressure from LO and TCO, the government’s technological R&D support agency, STU, did provide some funding for the formulation of suggestions for R&D projects aimed at questions of concern to the unions, as well as funding to the graphics workers union to engage in an actual technology development project. In addition, the importance of a union role in bringing R&D to bear on technological issues was for the first time cited explicitly by STU in a major program document in 1983. But this did not really entail a change of policy in principle and did so even less in practice.
Meanwhile, a great deal of experience about worker participation in the application of new technology is being accumulated, testifying to the wide scope for variation in the applications, in union roles in designing them, and in perceptions of what is at stake in alternative applications. Some illustrations are provided by flexible manufacturing system (FMS) installations being tracked by the Metalworkers’ union. These range from a case in which an autonomous group has achieved an extremely high utilization rate to one subjected to very tight managerial control, set up when an attempt to run a system with inexperienced workers resulted in too much downtime. Although these cases suggest that autonomous groups of workers with high, multiple skills may be most effective in achieving managerial goals of high utilization, management is evidently threatened by dependence on the workers’ knowledge, including the risk that they will leave and apply that knowledge elsewhere and the consequent weakening of managements’ wage-bargaining position. Managerial counterstrategies of cultivating workers’ identification with the companies they work for, including incentive schemes and applications that inhibit the buildup of skills that are not company specific are, in turn, seen by unions as a threat to their ability to retain membership identification while securing applications that further work-quality goals.29
These and other cases, along with the lack of significant change in the policy framework, suggest that the problems and possibilities of union influence on technological development remain pretty much as they were described in the reports to LO’s 1981 Congress. There is neither sufficient space nor data to provide a comprehensive review of the evidence for this judgment. We shall instead take one case, the graphics union project partially funded by STU cited earlier, and describe it in some detail. The project is referred to by the evocative name “UTOPIA,” an acronym formed by the Swedish words for its full title, undoubtedly chosen for that reason, “training, technology, and products from a quality of work perspective” (“Utbildning, Teknik Och Produkt I Arbetskvalitetsperspektiv”). The UTOPIA project is probably the most promising demonstration of how unions might bring technical expertise to bear on the development of alternative technologies; at the same time it illustrates especially clearly some of the most serious difficulties in doing so.
The UTOPIA Project
Broadly stated, the aim of the UTOPIA project was to reverse the tendency of computerization to impoverish or even eliminate the work of graphics workers—skilled craftsmen in the printing trades.30 It sought to achieve this aim principally by developing alternative computer technology that graphics workers could use as a tool for producing high-quality products, and by developing the work organization and training required to utilize the alternative technology. Initiated jointly by the Nordic Graphics Union (Nordisk Grafisk Union, or NGU) and the Swedish Center for Working Life (Arbetslivcentrum, or ALC) in 1980, the project focused on computerized text and image processing for page makeup in newspapers. It was carried out by a fifteen-member team consisting of social scientists at ALC, one of which was the project leader, graphics workers employed in several Nordic newspapers, and computer scientists at a Danish and a Swedish university. Representatives of graphics unions in Denmark, Finland, Norway, and Sweden also constituted a reference group to which the project team reported. After its inception UTOPIA entered into cooperation with a large Nordic commercial project aimed at developing technology of the kind with which UTOPIA was concerned. Referred to as TIPS, for Text and Image Processing System, the development project was led by a subsidiary of a large, Swedish state-owned publishing company, Liber AB. Cooperation with TIPS proved crucial for carrying out the UTOPIA project.
Forging a Development Coalition
The UTOPIA project had two sources. One was the concern of the graphics unions over the revolutionary changes being brought about in the printing industry by computerization during the 1970s. The other was a set of studies done at ALC on the implications of technological change in general and computerization in particular for the quality of work. The connection between the Swedish graphics workers and the researchers went back to 1973, when the union sought the support of researchers at Lund University. Some of those researchers went on to ALC, where the union contacts continued. The local branch of the graphics union at one of Sweden’s major dailies, Svenska Dagbladet, participated in one of four studies that were part of an ALC project, DEMOS, begun in 1975 and aimed at developing “knowledge about planning, control and computer use” needed by trade unions to engage in negotiations and training in connection with the introduction of computer technology.
The graphics union needed the researchers’ support in order to take advantage of the opportunities opened up to it by a “technology agreement” it had concluded with employers in the industry in 1974. This was the first such agreement in Sweden, preceding MBL’s enactment by some two years. The agreement provided for some employment guarantees in connection with the introduction of new technology, rights to retraining for changed work tasks, and rules concerning the manning of new equipment. Further provisions concerning the introduction of integrated text processing were added to the agreement in 1980.
The Swedish printing industry technology agreement illustrated a widespread trend in the Nordic countries. A General Agreement on Computer-Based Systems was concluded between the Norwegian Federation of Trade Unions (LO) and the Norwegian Employers Confederation (NAF) in 1975. A general technology agreement was concluded between the peak organizations of unions and employers in Denmark in 1981, which the Danish graphics union sought to supplement with employment guarantees like those won by its Swedish counterpart, though so far without success. In Sweden, as we saw, no general agreement in the private sector was struck until 1982. Throughout the Nordic countries, there are many agreements between individual newspapers and local unions concerning worker rights in connection with the transition to new technology. Of course, there have been and continue to be conflicts between graphics workers and other occupational groups, particularly the journalists, over how the radically changed tasks are to be distributed among them. Particularly in the Swedish case, conflict between the graphics workers and journalists has proven to be a serious obstacle to the realization of UTOPIA’S potential—a point to which we shall return.
The technology agreements reflected not only the problems with which the new technology was confronting the graphics unions but also the possibility of influencing the development of the new technology in ways that permitted the problems to be solved on terms acceptable to both the graphics unions and the employers. Creating a framework for cooperation in developing the technology, the agreements also offered a potentially significant market advantage for suppliers that could develop technology satisfying union as well as employer requirements. Formulating those requirements in such a way that they could be incorporated into technical development was a new and difficult task for the unions, however. Initial efforts underscored as well as concretized the difficulties involved. The lessons drawn from those efforts, including the DEMOS project, shaped the rationale and design of UTOPIA.
As stated in its prospectus, UTOPIA was specifically addressed to the difficulties encountered in previous trade union efforts to advance their members’ interests in the quality of work—skill content, democratic organization, and a good work environment. The difficulties were ascribed to two main factors: (1) disparities of knowledge and power between management and unions, and (2) limitations imposed by available technology. With respect to the first, management was described as having “relatively unambiguous and well-defined targets for technological development (e.g., productivity and efficiency measures) which can conflict with trade union demands, and . . . extensive experience of technological development from the traditional company-oriented viewpoint which dominates over trade union demands.” With respect to the second, it was pointed out:
It is often difficult to make basic changes in the content and organization of work if production technology is not altered at the same time. This difficulty is worsened by the growing tendency to purchase technology in the form of “turn-key” packages. Another difficulty is that the accompanying training only comprises general servicing of equipment; it does not allow for changes to be made nor does it provide a view of the process as a whole.31
These factors tended to confine trade union work-quality demands to resistance to deskilling and displacement. But this was characterized as a “defensive strategy” that was ultimately bound to be futile as “production is increasingly based on scientific and technological expertise” and as demand for “practical experience of trade skills” declines. Technology thus comes to be perceived “as adverse or just plain evil,” eroding the positive attitude to technology “traditionally held by the trade union movement.” Trade unions had to instead pursue an “offensive strategy” aimed at developing technological and training alternatives more consistent with work-quality demands.
Instead of defending the status quo an offensive strategy can be developed for another type of technology that improves the quality of work and products, [and] that is not inflexible but can be dynamically changed at individual workplaces as the workers develop their skills. . . . Training [therefore has] to be developed to provide workers . . . with the means to adapt and use the system so that it meets their demands for the work and the product.32
To pursue such a strategy, it was necessary not only for the unions to have access to relevant technological and work-organization expertise but also to develop methods for bringing the expertise to bear on the unions’ concerns. UTOPIA was conceived as an effort to develop such methods in the course of developing specific alternative technologies and associated work organization and training. Thus, it was both “a technology development project and a social-scientific experiment to understand the conditions for technology development” responsive to union goals.
In its design UTOPIA sought to cope with the fundamental constraint on unions’ capacity for an offensive strategy posed by the costs of developing production technology, generally entailing investments so large as to make it impossible for unions to undertake such development on their own. Yet this was not equally true for all aspects of technology, such as the software crucial to computer technology. “The development costs of programs,” the prospectus pointed out, “are considerably less, generally speaking, than of hardware.” Accordingly, focusing on software offered a possibility for overcoming the cost obstacle. It was seen as offering additional strategic advantages as well. “Developing alternative software also enables a dynamic technology to be created [which] can be further developed at local level if alternative training programs are simultaneously developed.” Unaltered hardware would still impose limits to the trade union goals that could be realized, but those limits would be more precisely identified in the process of software development, making it possible to specify demands for subsequent hardware development. Thus, software was viewed “as a foot in the door for trade union technological development.”
Finally, software for text and image processing in newspapers seemed an especially promising application on which to work. This was clearly a next stage in the technological change that was rapidly transforming the printing industry. The change that had already occurred gave workers in the industry strong stakes in gaining some leverage on the process of change—their numbers were declining, many of their traditional skills were no longer needed, and “knowledge of typography and layout, the basic trade skill” threatened to disappear, along with the “quality of graphical products.” At the same time, conditions for developing trade union alternatives were especially favorable in several respects. The technology agreements established a favorable climate, and the graphics workers’ technical sophistication put them in a good position to participate. Moreover, the several thousand companies in the Nordic countries formed a significant potential market for alternative text and image processing technologies that might be developed under the joint auspices of the Nordic graphics unions.
This market potential is what prompted Liber’s entry into the picture. Liber had set out to capitalize on its comprehensive knowledge as a user of printing equipment to enter the field of producing it. To that end, it organized the TIPS project, which if cofinanced with the Nordic Industrial Fund, an R&D support agency formed by the Nordic governments. A subsidiary, Liber System AB, was set up to run the development project. Various aspects of the work were carried out by Imtec, a spin-off company from Linkoping Technical University that became a Liber subsidiary; Typlan, a subsidiary of Nokia, a large Finnish conglomerate; and the Graphics Laboratory at Helsinki University. Liber’s strategy was to establish itself as a major supplier of the next generation of newspaper production technology in the Nordic market, which would give it a basis for entering the international market. Acceptability of the new technology by the Nordic graphics unions would clearly be an advantage, if not a necessary condition, for achieving its Nordic market goal. The prospects of acceptability would certainly be enhanced if the new technology met the unions’ work-quality demands. Union as well as newspaper concerns were incorporated into the TIPS project through a reference group, but UTOPIA offered TIPS the opportunity of integrating union concerns into the development project more directly and organically. Liber accordingly had a strong interest in establishing a link between the two projects.
UTOPIA obviously had much to gain from cooperation with TIPS as well. Such cooperation promised an opportunity to formulate, test, and revise the system specifications that UTOPIA was designed to produce in the context of a concrete production technology development project of the kind it could never expect to carry out on its own. Given these mutual interests, an agreement on the terms of cooperation was readily reached. UTOPIA committed specified resources to continuously develop, test, and revise its specifications of requirements and suggest ways of meeting them in the technical development process, and to develop training materials and organization for the system being developed. TIPS likewise committed specified resources to try out alternative solutions aimed at meeting UTOPIA’S specifications and evaluate the system in terms of them. UTOPIA was not bound to endorse the particular technical solutions adopted by TIPS, which the latter remained free to choose in the light of the commercial, economic, and time constraints under which it operated. The agreement also provided for the use of early prototypes and purchase of early series installations on favorable terms by the various institutes involved in UTOPIA and other Nordic printing trades training institutions.
Implementation of the agreement was contingent on additional funding of UTOPIA’S participation. At the same time, the agreement improved the prospects for financing UTOPIA. Financing had turned out to be an intractable problem, even on the smaller scale required for software development. The difficulties encountered by the project in this connection underscore the fundamental issue of government support for development of technological alternatives that the unions had been stressing.
NGU and ALC provided the initial funding for UTOPIA. This sufficed for the planning stage of the project, but additional funding was needed to go on with the actual development work. The computer scientists required for that had been identified at the Institute for Numerical Analysis and Computer Science Group (Numerisk analys och datologi, or NADA) at the Royal College of Technology (Kungliga Tekniska Hogskolan, or KTH) in Stockholm, and the Computer Science Department (Datalogisk Afdelning, DAIMI) at Aarhus University in Denmark. One possible source of funding for their participation was the ASF, mentioned earlier. Run by a tripartite board including employer and union representatives, ASF was established by the government to support research on the work environment and was ALC’s principal source of support. Typically, as noted already, ASF supported research on the effects of existing technology on the work environment, in the broader sense of work quality as well as health and safety, rather than on the development of new technology with work-environment implications. Although ASF was at the time planning the new program of research on work-organization issues associated with the introduction of computer technology, it was not clear that it would support development of new technology. Even if it did, the support would not be available for some time. Moreover, as LO argued, the fact that there already existed an agency for funding technical research and development, STU, made it seem the appropriate source for funding of the technical as opposed to work-organization aspects. So funding was sought for NADA’s participation from STU and for DAIMI’s from the Danish counterpart of STU.
Although STU had put a high priority on research in computer technology, repeated efforts to secure funding from the section of STU responsible for such research were unavailing. UTOPIA was evidently a new kind of project for STU that did not fit precisely into the categories of projects it was accustomed to handling. These typically were conventional R&D projects undertaken by business firms or technical universities, from which STU staff members usually came. To them, R&D projects for which unions were responsible could apparently not be regarded as serious or even legitimate, particularly if conducted by researchers tainted by the reputation for radicalism acquired by ALC. In any case, rather than responding imaginatively to this new kind of project, STU’s computer technology staff seemed almost to go out of its way to find in the existing project category definitions reasons for rejecting UTOPIA’S proposals.
The prospects for a favorable response from STU were much improved by UTOPIA’S association with a vendor having a commercial stake in the successful outcome of the development project, for the absence of such an association had been one of the grounds for rejecting earlier proposals (although the computer science aspect of the project for which support had been sought could readily have been interpreted as falling into a “knowledge development” category for which such a requirement was not set). With the backing of LO, which had been pressing STU to accept responsibility for supporting union projects, a new proposal was submitted, but this time to STU’s technology procurement section rather than its computer section. Although UTOPIA did not fit into the procurement category any more precisely than into the other ones, the proposal was approved. Whether this was because the staff now involved was more imaginative or less biased against unions, or because the top administration felt that a positive response to LO’s persistent pressure was desirable (or all of these reasons) is not clear. Whatever the reasons for it, the decision provided UTOPIA with the funds it needed to go on. But the terms on which it was made did not resolve the underlying policy issue of public support for autonomous trade union development of alternative technologies. Thus, having been defined as a technology procurement project, UTOPIA was unable to get further STU funding to continue its association with TIPS once it reached the first pilot installation phase at the end of 1984.
Operationalizing Work-Quality Goals: The Tool Perspective
As a technology procurement project, UTOPIA’S main task was to specify the requirements to be met by the system being developed. The method for formulating the specifications that was worked out is one of UTOPIA’S most important results. As defined in the project, the problem was whether “the new technology could be designed as a tool for skilled printing workers.” From this “tool perspective,” it was necessary first to identify the functions the new technology had to perform in order to enable the workers to implement judgments based on their knowledge of what constitutes good quality in the product and then to specify the characteristics and capabilities required to perform the functions. The tool perspective can best be described by illustrating its application.33
In the case of newspaper page makeup, the essence of the skilled workers’ competence was conceived as lying in their understanding of the visual ingredients of readability. This is what enables them to give “typographical form” to the “journalistic model” of a page produced by the editorial work of choosing content, producing text, distributing it among pages, and so forth. To give the journalistic model typographical form requires judgments concerning the positioning of different kinds of material on a page—main text, pictures, headlines, rules, white space,—based on the skilled workers’ distinctive training and experience, including the resulting “tacit knowledge” that enables them to achieve high quality with the infinitely varied material with which they work.
The tools with which to implement those judgments are provided by the technology of page makeup. For centuries the tools were those of lead (hot metal) technology. Only in the late 1960s were they replaced by the tools of paper pasteup technology. This resulted from the introduction of photo typesetters and computer word processing, producing the material to be made up into pages in the form of paper or film rather than lead. Now, paper technology is already being replaced by technology that produces the material entirely in the form of digitalized data, enabling the material to be positioned on pages with computers and then serving as instructions for the machines that make the plates with which the printing is done.
The problem for UTOPIA, then, was how computer technology could be made to serve as a tool as did lead and paper technology, enabling workers to carry out at least the same operations on the material as they could with conventional technology. To propose the capabilities and characteristics that the computer technology therefore had to have, the computer specialists had to learn from the graphics workers what the operations were. To evaluate the adequacy of proposed capabilities and characteristics for carrying out the operations, in turn, the graphics workers had to learn from the computer specialists what the possibilities and limits of computer technology were, at its current or near-future state of development.
Such a mutual learning process between workers with particular occupational skills and scientists with the relevant technological and work-organization skills is itself essential to the conduct of trade union technology development projects. Developing the capacity for engaging in such a process has frequently proven difficult, however. By trial and error UTOPIA achieved it to a considerable extent. “More or less formalized description methods ranging from scenarios to data flow” were attempted initially but “did not function very well as a vehicle of communication” for the workers. A mock-up to simulate computerized page makeup proved to work much better. With it, the workers could go through the page makeup process step by step. At each step corresponding images were drawn on simulated display screens, and the operations and interactive devices required to implement them identified and recorded by the computer specialists. Crude at first, the mock-up was later made more realistic in appearance, paper simulation of display screens was replaced by back-screen projection, and a “real computer workstation with a high-resolution screen and a tablet with a puck” was used “to experiment with and illustrate aspects which were difficult to simulate with the mock-up such as coordination between puck movements and screen image changes.”
With these and other components UTOPIA built up an experimental and research milieu or “technology laboratory.” This proved to be an effective instrument not only for collaboration between workers and technical specialists within UTOPIA but also for other purposes. One was for training in vocational colleges and retraining programs. This presupposes that the graphics workers’ occupation (if not the number of jobs) would be preserved through the development of computer technology that called for typographical and layout skills and that there would also be scope for further development in the workplace by the workers themselves, for whom programming and related understanding therefore became part of their essential job skills. Another use to which the instrument was put was the formulation and evaluation of proposals in negotiations over local union-management contracts and over purchase contracts with suppliers of equipment and software.
In the process of translating page makeup operations into specifications for computerized systems, both the limitations and potential of available computer technology were underscored. Display screen technology was revealed to fall far short of requirements. To be able to see what a page looks like when materials are positioned in a particular way, workers need to have the whole page before them “in a resolution good enough to facilitate effortless reading of the body text.” The capabilities and characteristics that the display screen accordingly had to have were:
• it is large enough to hold the whole page or spread in natural size plus a little extra room for a work area;
• it has a resolution of a quality that makes the body text easy to read in its true size and in all details;
• it is non-light emitting, which means that the display should have a variable degree of density, just like the printed page.
However, “not even . . . the most advanced available display technology” could meet these requirements for a “satisfactory representation of the page and its material as a basis for the makeup person’s professional evaluation and processing.” There were no screens large enough to display a whole page at full size, but there were none with sufficient resolution to permit the body text to be readable when a page was reduced sufficiently for it all to fit on the screen. Text had to be black on light background to approximate the appearance of the printed product, but the resulting amount of light emission strained eyes so much as to impair evaluation. Light emission could be avoided with liquid crystal display technology, but this was not yet developed to the point where it offered sufficient size, resolution, and speed. The magnitude of the gaps between the requirements and available technology, at least when the project was carried out, was indeed large, with processors estimated as being some one hundred times too slow and screen resolution ten times too low.
Nevertheless, other features of computer technology offered many advantages over the earlier technologies. The greatest was freedom in making changes of any kind. They could be made more easily, more rapidly, and at any time, right up to the last minute. Multiple alternatives could be tried without consuming much time, and pages could be made up without having to wait for all material to be in. In these respects computer technology promised to provide more powerful tools than either lead or paper. In addition, it could recapture an advantage of lead that had been lost by the shift to paper: precision in positioning so that all material can be exactly rectilinear, spacing uniform where desired, and so on. For the potential of computer technology to be realized, however, it was necessary to find some way to deal with the display screen limitations as well as to provide the programs and interaction tools to select and implement the needed operations, so designed as to permit modification by the workers as they learned of new needs and possibilities.
To work out ways of meeting these requirements, a “user model” for conceptualizing the process of making up a page on a display screen was devised. The model postulated a “table” on which a page ground, menus, material lists, status information, and a work area were placed. Given the limited screen size, a capability for viewing the whole page and any portions of it at various levels of reduction or magnification, along with the other items on the table, was conceived as a set of “lenses” at the worker’s disposal. To look at the same portion with different magnifications at the same time, different lenses could be used simultaneously (e.g., to look at one portion in its position within the whole page through one lens and to look at it in higher magnification in the work space through another, which apparently might be on an adjacent screen). All operations could be implemented on the material, regardless of the lenses through which it was viewed.
Working with this user model, the project team specified the additional capabilities required to make operations on the material easy and natural. These included fixed scales rather than continuously variable reduction and enlargement to facilitate rapid and precise shifts among lenses, the availability of gravitational field pointing for automatic precision positioning, and the like. They also included the various interaction tools required to select and implement all the operations. In contrast with display screen technology, available interaction tools such as those for pointing to material like an electronic pen directed at the screen or a puck (“mouse”) moved on a tablet, a keyboard, and function keys were evidently deemed adequate, if the software permitting them to be used in required ways could be provided. Thus, the software had to permit the simultaneous use of several tools, continuous feedback via cursors and status information, and retrieval of preceding statuses, whether immediate or remote.
Pending developments that overcame the limitations in available and prospective technology and leaving aside the idea that page makeup should not be computerized until the limitations are overcome, UTOPIA thus concentrated on designing a system that was as good as possible from the standpoint of work and product quality, given the limitations. That meant formulating specifications designed to satisfy work-environment and work-organization requirements as well as those concerned with the page makeup process itself. Of course, the latter inescapably involved work-environment considerations, as in connection with the eyestrain associated with display technology deficiencies. Using the technology laboratory to simulate the computerized work process, UTOPIA compiled a substantial list of other requirements for ergonomically satisfactory equipment and procedures embracing the entire work station as well as the computer technology itself.
But the work environment in the broader sense shades into work organization. To facilitate thinking about work organization, the simulation technique was supplemented by a “construction box” of functions or tasks involved in page makeup with which to visualize alternative ways of relating the functions to each other. In the graphics workers’ case, as in others, a fundamental union goal was to diminish the social division of labor by widening the range of tasks within any production process on which each worker works and to provide the conditions for “a more democratic work organization where a collective of autonomous workers can be responsible for and distribute multiple work tasks.” In addition to the workers involved in page makeup on which we have focused, those involved in image processing are also engaged in the page preparation process. A choice can be made between a division of labor in that process that assigns these two sets of tasks to two separate, specialized groups of workers or one that allows the same group of workers to alternate between them. To be able to opt for the latter, it was necessary for the technology to facilitate rotation among tasks. This led to a variety of specifications for computerized page preparation technology. One of them was a sufficient degree of standardization between the page makeup and image processing hardware and software to minimize the need to learn different operational skills to do both kinds of work, while allowing sufficient flexibility to allow adaptation of operations to individual work styles, capabilities, and handicaps. Another was administrative software permitting easy retrieval of work in progress at the stage at which it was interrupted. Along a different dimension of job enlargement were requirements concerning user maintenance and repair.
Work-organization issues bearing on system design became controversial and difficult to deal with when they impinged on the tasks of separate occupations organized in separate unions, however, particularly when they concerned the consequences of computerization for the graphics workers and journalists. As in other industries where technological change is reducing the demand for labor while changing the labor process, workers in the affected occupations are pitted against each other in competition for the remaining jobs, and unions are plunged into conflict with each other as jurisdictional demarcations based on previous technologies are eroded. As elsewhere (but not everywhere), the newspaper industry in Sweden has been beset by disputes between graphic workers and journalists over who will do what in the chain of processes from “manuscript to plate” which computerization makes increasingly possible to integrate. Journalists feared loss of influence over layout if computerized page makeup workstations were staffed by graphics workers, while also anticipating new employment opportunities if journalists got the right to staff them. But if journalists did so, then the graphics workers, in turn, feared the loss of a key function of their occupation and with it the loss of the associated jobs.
UTOPIA’S premise was that these processes could be designed in such a way as to permit both “graphic workers’ and journalists’ occupational skills, identity and pride to live on.” One way to do it would be to simply duplicate traditional work organization with new technology. But this approach, it was pointed out, made it difficult to exploit the possibilities for flexibility—e.g., eliminating restrictions on alteration or introduction of new text imposed by traditional page makeup technology—opened up by the new technology. So a model work organization that would permit journalists and graphics workers to cooperate on material accessible to both through an integrated computerized system was suggested. This model was largely based on experience in a small Norwegian newspaper at which graphics workers and journalists worked side by side and communicated directly with each other about the journalists’ intentions and the graphics workers’ efforts to give them typographical expression. The possibility of realizing UTOPIA’S premise in practice was thereby given some confirmation. Yet, the tensions between the Swedish graphics workers’ and journalists’ unions precluded the joint development of that model from becoming a part of the UTOPIA project.
Thus, journalists did not participate in the formulation of specifications constituting the core of UTOPIA’S first stage, neither in the project itself nor in the reference group to which it reported, although there were journalist representatives in the TIPS reference group. Opposition by the journalists was also a major factor preventing implementation of what had been intended as UTOPIA’S second stage. This was to occur in connection with the first pilot installation of the TIPS system. The UTOPIA project group was to evaluate the system in the light of the specifications that had been formulated, leading in turn to further revision of the specifications, and to develop an experimental work organization for utilizing the system in which graphics workers and journalists would both participate.
Although the site chosen for the first pilot installation was a Stockholm afternoon daily owned by the LO, Aftonbladet, it proved impossible to undertake the work-organization experiment. The unions of graphics workers and journalists (the latter is a TCO union) could not agree on staffing rights or on how to carry out the experiment. In the absence of agreement by the unions, Aftonbladet’s management could proceed without such an experiment, which it preferred not to have. And in the absence of a desire for it by the purchaser of the first pilot installation, STU was unwilling to press for such an experiment, which it evidently did not see as part of the technology procurement project it was supporting. As a result, the second and final stage of UTOPIA was scaled down drastically. It was confined to observation and documentation of what happened when TIPS was tried out at the newspaper, without any systematic effort to realize the potential for work and product quality that UTOPIA’S participation in its development had incorporated into it. With this to be carried out in 1986, UTOPIA will have come to an end.
The UTOPIA Project’s Implications
All of UTOPIA’S results are accordingly not in—and the full story of what it has done so far has not been secured at the time of writing—but enough is known to permit some tentative conclusions, not only about the UTOPIA project itself but also about the issues of trade union technology policy on which it sheds light.
To begin with, the project has had some immediate, practical results. The set of specifications it developed, preliminary as they are, are serving several purposes. They have entered into the design of the TIPS system, although precisely how is not indicated by the available information. They are also serving as the basis for negotiations between local unions and management at several newspapers concerning the purchase and utilization of computer technology and, in turn, for negotiations with suppliers (including suppliers other than Liber). The technology laboratory and construction box developed for the purpose of formulating the specifications are themselves important results, providing techniques that are being used in adapting the specifications for local negotiations and for training in the vocational college.
More broadly, UTOPIA serves as a valuable demonstration project, providing a rich fund of experience bearing on the problems and possibilities of developing a trade union strategy for influencing technological development. On the positive side, its most important contribution has been its demonstration of how workers can participate actively in technological development, incorporating their skills, including tacit knowledge built up through long experience, directly into the design process. This provides a credible countermodel to the conventional approach to technological development that dispenses with workers by relying on experts alone to incorporate operations equivalent or alternative to those carried out by workers into the design process on the basis of theoretical knowledge and formal description.
The lessons drawn from this experience have been articulated by contrasting the “tool perspective” from which UTOPIA worked and the “systems perspective” that prevails in the design and construction of computer technology. Although such technology is readily conceived in systems terms, those terms are seen as fostering a conceptualization of applications in ways that have negative work-quality consequences. Thus, in systems terms applications are viewed “from the top of the organization,” the organization is viewed as a “structure, whose important aspects may—and should—be formally described,” and workers’ jobs are reduced to “algorithmic procedures” so that workers and computers are viewed “as information processing systems on which the described data processing has to be distributed. This conceptualization is said to facilitate, even if it does not cause, a tendency for applications to reduce the “need for experience and skill,” allow “less control over and understanding of the production process,” increase the division of labor, and diminish “planning as part of the job.” By instead taking the skills that workers display in practice as the point of departure, and by conceptualizing computer applications as tools for exercising and enlarging those skills, the tool perspective is aimed at designing applications that more adequately meet union work-quality objectives. UTOPIA thus demonstrated the feasibility of using the tool perspective in developing a specific computer application.
The claim is not that the tool perspective provides a model for achieving union work-quality objectives in all applications, but that it is one alternative to a systems perspective, most relevant in situations where “some kind of material is refined by skilled workers.” The range of such situations is wide, including many kinds of office work as well as industrial work, such as that of machinists. However, in other situations, such as those in which there is “more formalized (or formalizable) data processing with long sequences of predeterminable operations,” the need to develop other alternatives is called for. In that connection, UTOPIA offers an example only at the most general level of combining workers with technical and work-organization specialists in developing alternative technologies.
Major obstacles to doing so are also illustrated by UTOPIA’S experience. One concerns the resources required. As the LO documents cited earlier point out, R&D resources are heavily concentrated in large private firms, particularly multinationals based outside of Sweden, with additional resources scattered among other firms, universities, and governments so that the resources are directed toward the objectives of those institutions. Unions, on the other hand, do not have the revenues from sales or taxes enabling them to direct R&D toward the distinctive work-quality objectives for which they are the most significant voice. They must therefore rely primarily on their ability to influence the other institutions’ definitions of the problems on which they bring R&D to bear. This influence can take various forms, including direct participation in problem definition in the other institutions, shaping the demand for technology by setting standards through collective bargaining, and by collaboration with suppliers (as in the UTOPIA project.) But all of these forms on influence are contingent on at least sufficient command of technical resources to be able to autonomously operationalize union objectives as specific technical problems and to evaluate the solutions that result from R&D efforts.
In the last analysis, as LO argued, unions cannot have that minimal command over R&D resources without public funding, and this will not be available except insofar as the unions’ work-quality objectives become incorporated into public policy, as have the other kinds of objectives that have been incorporated into policy guiding public funding of R&D, such as industrial competitiveness, defense, health, and education. Since the extent to which this happens ultimately depends on the political strength of the unions, this might at least have been expected to happen in Sweden, where that strength is probably greater than in any other country. Work quality has indeed become an objective of public policy in Sweden, as the legislation of the 1970s on the work environment and extension of collective bargaining to workplace issues indicates. But the limits of the policy commitment are clearly demonstrated by the difficulties UTOPIA encountered in securing funding from STU, the government’s principal agency for supporting technological R&D.
Although STU finally did provide the funds needed for UTOPIA’S first stage, it did so by treating it as a type of project with which it dealt all along, technology procurement, rather than recognizing union capacity to bring R&D to bear on work-quality problems they define as a new type of technological development project entitled to support. A redefinition of STU’s mission in this way is unlikely to occur unless the government does so explicitly, and this does not seem to have happened. To be sure, STU has financed some union participation in development projects, such as one concerned with computerized medical information systems, and also some work by central LO and TCO staffs on the kinds of projects for which unions seek support. But the significant shift in research policy needed to enable unions to play an autonomous role in technological development, which LO called for in 1981, has not yet occurred.34 Until it does, UTOPIA is likely to remain an isolated demonstration of what could be a “new Scandinavian model” of technological development rather than an important step in the realization of that potential.
The unions’ effectiveness in bringing about such a shift in national research policy has certainly not been enhanced by divisions among them created by conflicting stakes in technological change. This is another major obstacle to achieving union influence over technological change that is underscored by UTOPIA’S experience. Its inability to secure collaboration between graphics workers and journalists reflected the fact that in the newspaper industry, as in other industries where technological change is reducing the demand for labor while changing the labor process, workers in the affected occupations are pitted against each other in competition for the remaining jobs, and unions are plunged into conflict with each other as jurisdictional demarcations based on previous technologies are eroded. The specific consequence in UTOPIA’S case, as we saw, was that the conflict blocked implementation of the project’s second stage as it was originally conceived. If the two unions had been able to agree that a work-organization experiment in the pilot installation should be conducted, it is doubtful that the newspaper management could have refused to agree to it. Under those circumstances, it is in turn doubtful that STU would have cut off funding for UTOPIA’S participation.
Similar conflicts arise between other occupational groups and the unions that represent them, such as machinists represented by the metalworkers’ union and technicians such as designers and programmers represented by the private-sector industrial salaried workers’ union. The blurring of occupational lines distinguishing the latter two unions that results from technological change has led some union officials and some employers to propose a merger between them. But the obstacles to that are formidable, not least the political obstacles posed by the fact that the metalworkers are linked to the Social Democratic party and the salaried workers’ union is not. Short of merger, there may be some scope for collaboration between them, for there is broad similarity in the work-quality objectives espoused by the two unions, and there are individual work sites where they are engaged in joint efforts to influence the introduction of new technology. Yet, the conflicts over staffing of new technology that requires fewer workers are likely to be too great to permit the unions to develop a common strategy for influencing the development of the new technology.
What prevents a common strategy at the local and industry levels makes it very difficult to pursue a common strategy at the national policy level. The necessity of pursuing a “multilevel strategy,” in which union efforts at these different levels reinforce each other, is in fact one of the main lessons drawn from the UTOPIA project by its members. But the need for it to be pursued in common by all the affected unions is evidently as important a lesson. For, as a UTOPIA document concludes, the “lack of union cooperation—not the technology, the newspaper proprietors, nor the suppliers—can tragically enough be the decisive factor preventing the utopia from becoming a reality.”35
Notes
1. This essay is derived in part from a report prepared for the Swedish National Board for Technical Development (Styrelsen for Teknisk Utveckling, STU), under a contract to the Massachusetts Institute for Technology Center for Policy Alternatives. The author is grateful to the many STU and union staff members and others who generously gave their time for interviews.
2. Harry C. Katz and Charles F. Sabel, “Industrial Relations and Industrial Adjustment: The World Car Industry,” Industrial Relations 24 (3) (Fall 1985); 295–315.
3. Michael J. Piore and Charles F. Sabel, The Second Industrial Divide (New York: Basic Books, 1984).
4. See Robert E. Cole, “The Macropolitics of Organizational Change,” ch. 2 in this volume.
5. Statistiska årsbok 1985 (Statistical Yearbook) (Stockholm: Statistiska centralbyrån, 1985), 203.
6. Conflict has been especially marked between printers and journalists, as noted later in this essay. Less intense but perhaps more important for the future of the labor movement are the rival stakes of the LO metalworkers union and TCO private-sector clerical and technical workers union (Svenska Industritjänstemannaförbundet, SIF).
7. The following account, particularly of phases 1 through 3, is based on Andrew Martin, “Trade Unions in Sweden: Strategic Responses to Change and Crisis,” in Peter Gourevitch, Andrew Martin, and George Ross, eds., Unions and Economic Crisis: Britain, West Germany, and Sweden, 191–359 (London: George Allen & Unwin, 1984).
8. LO, The Trade Union Movement and Full Employment (Stockholm: LO, 1953), English translation of LO, Fackföreningsrorelsen och den fulla sysselsättningen (Stockholm: LO, 1951).
9. T. L. Johnston, ed. and transl., Economic Expansion and Structural Change (London: George Allen & Unwin, 1963). This is a translation of LO, Samordnad näringspolitik (Stockholm: LO, 1961).
10. Rudolf Meidner, addressing the 1961 LO Congress. LO, Kongressprotokoll (Stockholm: LO, n.d.), 336. All translations from Swedish, including document titles, are by the author unless otherwise indicated.
11. S. D. Anderman, ed. and transl., Trade Unions and Technological Change (London: George Allen & Unwin, 1967), 122. This is an English translation of LO, Fackforeningsrörelsen och den tekniska utvecklingen (Stockholm: LO, 1966).
12. Ibid., 134–35.
13. Program for aktiv näringspolitik. Näringspolitiska kommittens slutrapport (Program for an Active Industrial Policy. The Industrial Policy Committee’s Final Report) (Stockholm: Socialdemokraterna, 1968), 5–6, 13–14.
14. Anderman, Trade Unions and Technological Change, 258.
15. This and the quotations in the preceding paragraph are from LO, Fackföreningsrörelsen och arbetsmiljon (The Trade Union Movement and the Work Environment) (Stockholm: Prisma, 1971), 29–35.
16. For further discussion of the employers’ position, including SAF’s encouragement of change in work organization on management terms, see Cole’s chapter 2 in this volume, as well as Martin, “Trade Unions in Sweden,” 254–58.
17. LO, Arbetsmiljon (The Work Environment) (Stockholm: Prisma, 1976), 85–94, 126–28, 148–49.
18. LO, Fackföreningsrörelsen och forskningen (The Trade Union Movement and Research) (Stockholm: Tidens Förlag, 1981), 9.
19. Ibid., 18–19, 72–74.
20. Ibid., 21–22.
21. Ibid., 19–20, 56–72, 74–98.
22. LO, Facklig datapolitik (Union Computer Policy) (Stockholm: Tidens Förlag, 1981), 9–13. The decline in Sweden’s advantageous position in the international economy, which had been under way for some time, was sharply aggravated by the impact of the first oil price increase and subsequent international recession. The impact was delayed by policies designed to “bridge over” the recession but that may well have intensified the impact when it could no longer be averted, which was after the Social Democrats were replaced in office by the bourgeois parties. The structural crisis that emerged then put a greater premium than ever on structural and technological change to restore the Swedish economy’s competitiveness. For a fuller discussion, see Martin, “Trade Unions in Sweden,” 288–310.
23. LO, Facklig datapolitik, 105–106.
24. Ibid., 106–108, 50–51.
25. Ibid., 51–56, 92–101, 108–109.
26. Ibid., 56–61, 101–104.
27. Ibid., 12.
28. Ibid., 117–19.
29. The text of the private sector agreement is available in English as Agreement on Efficiency and Participation: SAF-LO-PTK (Stockholm: Swedish Employers Confederation, 1982). Additional sources for the discussion of recent developments include press accounts, interviews, and materials generously supplied in response to mail and telephone requests.
30. The following account of the UTOPIA project is based on interviews with the project leader and others at the Center for Working Life (ALC), union participants, and STU staff members, as well as extensive documentation issued during the course of the project, including the project’s information bulletin, Graffiti, and more detailed reports on specific aspects of the project.
31. The UTOPIA Project; English translation of the UTOPIA research program (Stockholm: Center for Working Life, 1981), 12–13, 7.
32. Ibid., 13, 27–28.
33. An English language discussion of the tool perspective and its application to newspaper makeup work is provided in Pelle Ehn and Morten Kyng, “A Tool Perspective on Design of Interactive Computer Support for Skilled Workers,” Computer Science Department, Aarhus University, Aarhus, Denmark, DAIMI PB-190, January 1985.
34. Since this was written, STU has taken a further step by agreeing to LO’s proposal to establish a standing committee of STU and union representatives dealing with the whole range of projects in a large new program on computer R&D recently authorized by the government. In addition to providing information on the projects, the committee is supposed to facilitate collaboration between the unions and the researchers involved and to initiate new projects.
35. Graffiti 7 (Dec. 1984): 9. The UTOPIA experience is set within the perspective of a multilevel trade union strategy for technological change in a book by one of the participating researchers, Åke Sandberg, Technological Change and Codetermination in Sweden (Philadelphia: Temple University Press, forthcoming).