ABSTRACT:
The field of organization design is changing rapidly, reflecting contributions from managers and diverse organizational consultants. One of the most recent developments is in the design of non-routine systems. Explores non-routine systems design from a socio-technical systems (STS) perspective. Includes a brief discussion of non-routine systems and an examination of alternative methods for diagnosing and redesigning organizations composed of knowledge workers. The results of two STS case studies provide new learnings and point to special design principles for non-routine systems. Concludes with implications for both managers and STS consultants interested in management of knowledge workers.
Introduction
Competitive trends of the global business environment are causing executives to rethink traditional design configurations and work design. One of the key challenges for managers and organizational scholars in the 1990s is to ensure an adequate business response to global competition. Early indications are that the response includes movement from mechanistic to organic forms of structure, from functional to interdisciplinary project work, from individual worker to team-based structures and towards self-management and experimental network organizations. Increasingly, these innovations are being utilized with knowledge workers and non-routine organizations.
The importance of modifying organization designs to fit and support new business competitive strategies is a familiar theme in management. Increasingly, management and organization consultants are being asked to participate in improvement efforts that focus on organization redesign. A variety of models, design principles and change processes can be found in the management literature.
A recent comparison of organization design perspectives identified five "comprehensive" orientations (Stebbins et al., 1994). Each perspective offers a guiding model for viewing organization design, provides a set of design principles and outlines a process that managers and employees can use to study and redesign the enterprise.
All five perspectives have an established research record of applications within organizations, including non-routine systems. The five schools of thought are: organizational hologram theory - an integration and advanced development of the organization audit and analysis design orientation (Mackenzie and Holder, 1990); managing beyond the quick fix - an integrated application of multivariate analysis, participation and structure (MAPS) technology or organizational change (Kilmann, 1989); information processing theory (Galbraith, 1977; Nadler and Tushman, 1988); socio-technical systems (STS) theory and it various perspectives (Hanna, 1988; Kolodny and Stjernberg, 1986; Pasmore, 1988; Taylor and Felten, 1993); and, finally, self-design (Mohrman and Cummings, 1989). The most active dialogue on design and non-routine work can be found in the STS literature.
The potential application of socio-technical systems theory and principles to non-routine work settings was first suggested in the early 1970s (Emery and Trist, 1972; Herbst, 1974). However, enhancements of STS theory and diagnostic methods were required before practical applications could be tried. A broadening of STS theory to include open systems scanning, market and strategic matters (Hanna, 1988; Shani and Elliott, 1989; Taylor and Felten, 1993), adjustments in technical system analysis methods to accommodate non-linear throughput and knowledge work (Shani et al., 1992; Taylor et al., 1986), and identification of key issues in non-routine work situations and "deliberations" on the key issues (Pava, 1983) are illustrative of the enhancements. Combined with the work of Pasmore et al., and new attention to the high involvement of knowledge workers in the design process (Pasmore and Purser, 1993), we can now be optimistic about extending STS design to the white-collar world.
This article explores non-routine systems design from a STS perspective. After an initial discussion of non-routine organizations, a STS redesign process for knowledge work is presented. Two cases of STS application - one for a chemical company's R&D division and the other for a teaching hospital - are explored, followed by a discussion of special issues and the identification of new design principles. A discussion of implications for managers and STS consultants concludes the article.
Non-routine organizations
It has been often repeated that we are in an era of post-industrial society, with a new emphasis on computers, telecommunications and other information technology innovations. New information technologies, such as groupware, change the nature and dynamics of the work organization. Organizations depend on more non-linear transformation processes in core work areas, and have more departments that can be classified as non-routine. Non-routine organizational units face a high number of exceptions (unexpected situations, with frequent problems) in the course of carrying out work, and problem resolution is complex or unknown at the outset (Perrow, 1967). The combination of high task variety and complex or unanalysable conversion processes are the hallmarks of non-routine work. Examples of non-routine tasks and organizations include basic research, new product development, strategic planning, medical diagnosis and software development.
At the most basic level, non-routine organizations are composed of a social sub-system (the people with knowledge, skills, attitudes - all that is human), a technical sub-system (the inputs and the technology which converts inputs into outputs - or product-in-becoming) and an environment sub-system (including customers, competitors and a host of other outside forces). Organization design seeks to pull the three sub-systems together through a better strategy, conversion process, structural configuration and organizational support processes.
STS design process
Several STS redesign models and design processes can be used with non-routine organizations. For simplicity, we will adopt the Taylor and Felten (1993) four-phase design process, but will substitute specific tools and elements found to be useful with knowledge workers. The phases include discovery, systems understanding or diagnosis, ideal organization and implementation.
As shown discovery involves education about the socio-technical systems' paradigm, as well as project planning, management commitment and other start-up issues. Systems understanding involves creation of redesign steering and working groups, planning and implementation of diagnostic efforts such as an open-system scan of the environment, separate technical system and social system audits, data consolidation and feedback to various constituents. The ideal organization phase covers efforts to involve people in the creation of the ideal design and testing how well the new design advances the organization's purpose, controls key variances, contributes to quality of work life and meets other design criteria. Finally, implementation covers efforts to gain widespread involvement in the implementation of all aspects of the ideal strategic design, operational redesign, evaluation and continuing adjustment.
Beyond the basic STS design process, the list depicts special design elements associated with non-routine work. At the discovery phase, much more attention is given to involvement of the knowledge workers than of executives who normally sanction and fund typical STS programmes (Pasmore and Purser, 1993). Programme design must be done in collaboration with knowledge workers, who are typically quite autonomous. Pasmore suggests bringing all the employees together at one time to explore STS concepts and to determine what the knowledge workers value. Multiple meetings may be needed to gain consensus that redesign is worthwhile. Participation of this form at the front-end greatly complicates usual discovery activities, but leads to a stronger sense of partnership throughout the programme.
Significant changes are also evident during the second phase - systems understanding. During routine organization STS redesign, technical sub-system analysis focuses on the transformation process or interplay between input-throughput-output components and the human agent. Technical variance analysis provides a view of the transformation process, as depicted in a variance matrix. In many forms of non-routine work, especially in basic research and the early phases of projects, it is difficult to chart the component steps. Because of the non-linear and uncertain conversion process inherent in knowledge work, additional technical analysis tools and processes are needed. To date, the most promising analytical tool has been the study of "deliberations". As in phase one, widespread involvement of knowledge workers is needed to identify deliberations worth studying, and to conduct the deliberations' analysis itself.
Deliberations are reflective and communicative behaviours concerning a particular topic or issue. They are patterns of exchange and communication among people to address the topic, and may take the shape of formal meetings or informal desk-side discussions. Deliberations are not the same as discrete decisions or project milestones in that they encompass informal human interactions and the ebb and flow of information related to the topic over time (Purser et al., 1992). Completion of knowledge work tasks involves multiple deliberations. Deliberation topics can be ranked in importance, and the more important deliberations can be fully analysed. Analysing deliberations involves scrutiny of structured and unstructured deliberation forums. The analysis can be performed by STS consultants, or knowledge workers can be trained to conduct it themselves.
While the methods of analysing deliberations are relatively new, they show promise for understanding non-routine conversion processes. Deliberation forum structures and processes can be analysed in each work situation. Variables include forum size, membership (managers and other formally-designated players, informal players, customers, etc.), openness in sharing information versus competition among forum members, spirit of inquiry versus advocacy, knowledge of the task and technical transformation process, and time pressure. All potentially affect the quality of deliberations that take place. Additional complications include the stage of development (assuming a project exists), and number of information gaps or unknowns. Table I provides a brief glossary of terms associated with deliberation analysis.
Table I. Glossary of Terms For Deliberations Analysis
The social system audit is also enhanced by the deliberation analysis method. Focus is on the knowledge utilized during deliberations, which is in part determined by member needs, personal influence, professional disciplines and perspectives and other individual and group variables. The analysis identifies knowledge adopted and discarded, along with barriers to learning. Results of the deliberation analysis are used at the redesign or ideal organization phase to better align knowledge with influence in decision making, and to identify improved formal structures for integration of knowledge work. The remaining design activities in phases three and four of the list follow familiar STS patterns. Again, they are highly participative, following the notion that high involvement leads to high commitment to redesign solutions.
Case studies
A brief overview of the organizations and projects The first case involved the R & D division of a large high technology firm which manufactures both consumer and industrial products. The company suffered from increased domestic and foreign competition, with declining market share. High product development costs and delays led the new CEO to seek changes in the way products were conceived, managed and introduced. A design team was established to work full-time over a one-year period to co-ordinate the STS programme.
The second case involved redesign of a university-based medical rehabilitation hospital. The hospital faced moderate competition in the marketplace, and was mainly preoccupied with internal issues related to quality of health services and teamwork among different health care providers. A parallel learning structure (Bushe and Shani, 1991) was created to guide the redesign project.
Comparative analysis
The two organizations can be classified as non-routine systems. The tasks pursued were highly technical and dependent on a continuing influx of knowledge. The chemical company case covered highly-educated scientists, technicians and managers, while the hospital case involved medical school faculty, other health-care providers and hospital support-department personnel. High specialization was a factor in both cases, along with the need for continuing education for new knowledge and skills. Table II provides a snapshot of the organizations using the STS framework.
The environment sub-systems showed considerable variation, with the chemical firm facing many competitors and other external factors that changed rapidly (turbulent environment), while the hospital faced less competition but high scrutiny by outside accreditation agencies. Uncertainty was high and moderately high for the two organizations, respectively. Outside pressures were catalysts for deliberations on regaining market share in the chemical firm, while the audit activities of the accrediting agencies were catalysts for deliberations about quality of care and improved co-ordination among providers.
Deliberation topics in the chemical firm included basic research and product development issues of a technical nature, acquisition of knowledge, internal dissemination of knowledge and knowledge application. Deliberation forums in the chemical firm ranged from unstructured to semi-structured, in that informal meetings were the norm within technical discipline and project work units, while most significant topics were covered in large monthly review meetings, where diverse expertise could be shared. In addition to the stable forum membership, other project and functional experts attended when their projects were on the agenda.
Analysis of the social system in the chemical firm revealed a host of problems related to knowledge utilized. While the large forum format was appreciated by the scientists, certain barriers to full utilization of knowledge were identified by divisional employees. Barriers included lack of product system knowledge in managerial ranks, inadequate discussion and premature decisions on technical research approaches to be used, personal biases of scientists, and other factors shown in Table II. In brief, hierarchical arrangements and decision-making processes were not always well-aligned with knowledge. Analysis of the technical and social systems in the hospital produced a different scenario partly through greater emphasis on the hierarchy and formal hospital procedures. Deliberation topics concerned technical discussions about medical treatments and alternative ways to provide health-care delivery. On a total systems level, an administrative forum composed of hospital department heads and directors of clinical specialties dealt with everything from hospital policies and procedures to professional development and training. This forum was also the main channel for dissemination of information, and attendees typically held their own staff meetings following the hospital-wide sessions. While the administrative forum was judged to be effective by participants, some problems related to full participation and utilization of knowledge were identified.
At the medical care level, multi-disciplinary groups composed of faculty, students and clinical team providers conducted bedside sessions for treatment and teaching purposes. Interviews identified multiple problems in the clinical teams (past attempts to develop a standard approach to team functioning had failed). Problems included negative attitudes towards working in interdisciplinary teams, poor team processes and lack of shared expectations. In brief, many team members were concerned only about their technical contribution to the patients' care, rather than a group effort for rehabilitation. This issue stimulated efforts towards system redesign.
Business system redesign - the hospital
Medical rehabilitation work-processes in a teaching hospital are complex, and treatments for patients are problematic, particularly when multiple maladies are involved. The hospital had followed a pattern of assigning providers on an individual basis, but there were controversies about which specialists ought to be represented as well as about the capabilities of individual providers. For example, some administrators felt that a mental health provider ought to be assigned to every patient. Also, there had been attempts to assign non-physicians as team leaders for certain patients and this was not accepted by the physicians involved. Information from the diagnostic phases of the study increased top management's resolve to improve clinical team health-care delivery.
Table II. STS Comparison of Two Non-Routine Organizations
(A) STS Design Elements for R&D Division Chemical Firm
The design study group offered three options for improvement, ranging from hospital-wide trans-formation, based on a new team approach, to modest experimentation. Top management chose the most conservative option, believing that it would take a scientifically-controlled experiment to convince physicians of the interdisciplinary team approach advantages. Budgetary, cultural and political forces blocked extension of the programme to other providers. While the change programme did not move beyond the experimental phases, the parallel learning structure continued to sponsor diagnostic, design and training activities during succeeding years. The parallel learning structure was institutionalized and temporary groups were formed to create and implement new projects. Hospital management learned how to create and run improvement groups, co-existing with the regular hierarchical groups.
Business redesign - the chemical firm's R&D division
From the outset of the STS redesign project, upper division management hoped to develop and move new products to market more quickly. Management felt that both structural and procedural changes were needed to accomplish this objective. Following the diagnostic phase of the project, the design team recommended that an ideal product development organization would contain "integration groups", separate from the hierarchy. The integration groups would be composed of people who had relevant expertise and who had skills to integrate the activities of different work units throughout the system. The idea was to better align decision making with technical and product system knowledge. Integration groups were to set goals, allocate resources and involve subject-matter experts in deliberations and decision making. They were to maintain close relationships with management and to be included in management decision making. Integration groups were to be formed around research topics and short-term projects. Longer term learning groups were also formed around acquisition, dissemination and application of knowledge. The integration groups were implemented as planned, eventually displacing the existing 40-member forum and changing the existing process for gaining input to management decision making.
Learning from the case studies
It is noteworthy that both STS redesign projects produced new deliberation forums to supplement the formal hierarchy. In the R&D division case, the structural change was adoption of multiple integrator groups linked to the existing hierarchy. The advantage over the existing forum was that work could be done more quickly in parallel with stronger ties to scientific expertise and product system knowledge. The structure also showed promise for alleviating existing barriers, such as quality of deliberation processes. After the change, forums were composed of people with relevant expertise who also had clear interest in advancement of the topic, in contrast to use of a single large forum where interests varied and competing experts and strong personalities often dominated discussions. After the change, promising ideas could surface earlier and be examined in a more friendly setting. Longitudinal results of the changes have not been reported.
Similar changes were initiated in the hospital case. The parallel learning structure became the steering group for diagnostic, training and other improvement programmes occurring in the hospital. Since the steering group included medical providers, representatives of several management levels and support employees, multiple perspectives on problems and opportunities were offered (Stebbins and Shani, 1988). With approval from top management and the existing administrative group, the steering committee had freedom to identify and launch new programmes. Early reports from administrators indicate that the new forums provide a responsive mechanism for early identification and resolution of cross-departmental problems. No longitudinal results are available.
Design principles
The cases summarized in this article, along with the body of literature on management of non-routine organizations, point to a number of special design principles. The new principles are viewed as supplementing existing STS theory by providing guidelines for use with knowledge-worker populations. The principles can be grouped into three categories: principles related to macro-STS design theories; principles concerning what the new business design should accomplish; and, finally, principles connected with initiating and implementing STS design programmes (design process).
STS design theory
Principles to help evaluate new business designs
Design process
Conclusion
In classical STS theory, organization design is centred on the core production process. Pava's (1983) pioneering work on deliberations helped us see that the core production process in non-routine organizations is not always clearly identifiable and that new diagnostic methods and redesign tools were needed. Pasmore and Gurley extended the shift away from traditional STS variance analysis by expanding the deliberations analysis method to cover how knowledge is generatedand used. Recent case studies, including those reported here, both confirm the value of this focus and provide tools for analysing and redesigning organizations in other non-routine settings. We expect that further work on deliberations will produce fruitful and interesting results in the 1990s and beyond.
The design principles proposed in this article reflect case evidence and the authors' consulting experience with systems engineering, semi-conductor, academic, medical and scientific research and development organizations. Early indications are that the most significant modifications to STS theory relate to principles that can be used to evaluate new designs for knowledge workers and principles to guide the redesign process itself. Clearly, redesign projects for non-routine organizations will be increasingly participative; this may also mean that they will be increasingly conflict-ridden. We have also noticed a trend towards adoption of parallel learning structures and temporary forums in non-routine organizations and believe that these mechanisms show high promise for better integration of technical knowledge and business needs. One key to success may be managerial willingness to support multiple structural configurations and forums, and to align them with managerial decision making and other fundamental organizational processes.
References
Bushe, G.R. and Shani, A.B. (1991), Parallel Learning Structures: Increasing Innovation in Bureaucracies, Addison-Wesley, Reading, MA.
Emery, F.E., and Trist, E.L. (1972), Towards a Social Ecology, Plenum Press, New York, NY.
Galbraith, J.R. (1977), Organization Design, Addison-Wesley, Reading, MA.
Hanna, D. (1988), Designing Organizations For High Performance, Addison-Wesley, Reading, MA.
Herbst, P. (1974), Sociotechnical Design: Strategies in Multidisciplinary Research, Tavistock Publications, London.
Kilmann, R.H. (1989), Managing Beyond the Quick Fix, Jossey-Bass, San Francisco, CA.
Kolodny, H. and Stjernberg, T. (1986), "The change process of innovative work designs", The Journal of Applied Behavioural Sciences, Vol. 22 No. 3, pp. 287-301.
Mackenzie, K.D. and Holder, G.W. (1990), The Organizational Hologram, Kluwer Academic Press, Boston, MA.
Mohrman, S.A. and Cummings, T.G. (1989), Self-Designing Organizations: Learning How To Create High Performance, Addison-Wesley, Reading, MA.
Nadler, D. and Tushman, M. (1988), Strategic Organization Design, Scott, Foresman, Glenview, IL.
Pasmore, W.A. (1988), Designing Effective Organizations: The Sociotechnical Systems Perspective, Wiley, New York, NY.
Pasmore, W.A. and Gurley, K. (1991), "Enhancing R&D across functional areas", in Kilmann, R.H. (Ed.), Making Organizations Competitive, Jossey-Bass, San Francisco, CA.
Pasmore, W.A. and Purser, R.E. (1993), "Designing work systems for knowledge workers", Journal For Quality and Participation, July-August, pp. 78-84.
Pasmore, W.A., Sherwood, J.J. and Srivastva, S. (1978), "Social relations and organizational performance: a sociotechnical systems approach", in Pasmore, W.A. and Sherwood, J.J. (Eds), Social Systems: A Sourcebook, University Associates, La Jolla, CA.
Pava, C.H. (1983), Managing New Office Technology: An Organizational Strategy, Free Press, New York, NY.
Perrow, C. (1967), "A framework for the comparative analysis of organizations", American Sociological Review, Vol. 32, pp. 194-208.
Purser, R.E., Pasmore, W.A. and Tenkasi, R. (1992), "The influence of deliberations on learning in new product development teams", Journal of Engineering and Technology Management, Vol. 9 No. 1, pp. 1-28.
Shani, A.B. and Elliott, O. (1989), "Sociotechnical system design in transition", in Sikes, W. Drexler, A. and Grant, J. (Eds), The Emerging Practice Of Organization Development, University Associates, La Jolla, CA, pp. 187-98.
Shani, A.B., Grant, R. and Krishnan, R. (1992), "Advanced manufacturing systems and organizational choice: a sociotechnical systems approach", California Management Review, Vol. 34 No. 4, pp. 91-111.
Stebbins, M.W., Sena, J.A. and Shani, A.B. (1994), "Information technology and organization design", Journal of Information Technology, Vol. 9 No. 3.
Stebbins, M.W. and Shani, A.B. (1988), "Communications forum interventions: a longitudinal case study", Leadership and Organization Development Journal, Vol. 9 No. 5, pp. 3-9.
Taylor, J.C. and Felten, D.F. (1993), Performance by Design: Sociotechnical Systems in North America, Prentice-Hall, Englewood Cliffs, NJ.
Taylor, J.C., Gustavson, P.W. and Carter, W.S. (1986), "Integrating the social and technical systems of organizations", in Davis, D.D. (Ed.), Managing Technical Innovation, Jossey-Bass, San Francisco, CA.