Government reforms have led to the adoption of business models, and the pursuit of lower costs has encouraged widespread computerisation of many government processes. These reforms have helped to make government more efficient but have not always made it more accountable or improved communication between citizens. Efficiency has been the main priority for government applications of information technology. The principles guiding democratic application of information technology for citizen participation will not be identical to those appropriate for efficient service delivery. The differences may be roughly explained by the distinction between instrumental and developmental applications. Instrumental approaches allow completion of an immediate or short term task, perhaps with great efficiency. Developmental approaches foster iterative reflection on what the task should be. This work provides evidence that only the latter perspective is fully compatible with democratic process.

This study focuses on the requirements of democratic policy processes as an activity specific to government. No business models exist for democratic policy formulation. This includes policy relating to information technology. This is an important task at a time when corporations are often contracted to supply and manage information technology on behalf of governments. Developmental perspectives may be seen as irrelevant or too cumbersome, if efficiency is the main focus. Because information technology is nowhere a local industry, it was necessary to look at the broadest level influencing the design and distribution of information technology. This is, of course, the global level. Governments, however, exist at multiple levels, each using information technology according to their needs, resources, and abilities. Enterprises which require relationships of power and authority over the group may also be described as exercising a form of government. This brings an obligation to ask how that government should be structured (Dahl 1989:329).

The increasing popularity of the term ‘governance’, with its more general meaning of control, authority, or management, reflects recognition that such systems do not only exist as structures within geographic or legal jurisdictions. Governance exists within organisations, both large and small. It also exists in transnational forms, both formal and defacto. Information technology has become an essential component in every stage of governance, whether democratic or not. Despite enormous variation in scale and function, patterns emerge in uses of information technology for governance which are influenced by global structures and values. For example, minimisation of risk and customer focus have been guiding principles that have helped shape applications of information technology at all levels. This research examines these patterns, and argues that the systems operating at different levels are connected. It is necessary to understand the factors driving these systems to develop a model for the use of information technology in democratic policy processes.

This study uses the insights and concepts of complexity theory to offer a fresh perspective on how democratic policy processes relate to interactive technology. The perspective spans several scales of governance, but the detailed focus is on the organisational level. Grounding in evidence from diverse literatures and case studies provides an integrating approach which is both conceptual and practical. The public sector agency is seen as a site where the application of advanced technology for developmental, democratic interaction would potentially be most viable, and also most critical. This stems from the triple role of the public sector agency. In Australia, the public sector has been a pace-setter for best practice in management and human resources. Public sector agencies also develop and manage the policies that impact on all citizens, and thus have special accountabilities in democratic societies. The third aspect is particularly important for the argument developed here: the developmental impact on staff of changing internal communication processes and how this affects them as citizens.

The theoretical arguments developed in this thesis extend the application of complexity theory, as discussed below, to human systems in three principal ways. Firstly, this thesis proposes a link between democracy and information technology use in complex systems. Interactive technology can help re-pluralise democratic systems. This provides a practical and theoretical framework for discussions of electronic democracy. Interactive technologies are shown to be capable of shifting policy deliberation away from elite actors, by adding diversity and quantity. One value of such a framework is its overt requirement for normative communication protocols to ensure these benefits. The assertion that nonlinear dynamics expresses the democratic ethic has been made elsewhere (Kiel 1994:219), and the researcher builds on this to suggest that interactive technologies can be structured to enhance this potential.

The researcher’s second extension of complexity theory is the identification of relationships across scales in the way information technology is used in governance. These patterns are described as ‘attractors’ in complexity theory, and the literature reviews of Part I show that these are essentially self-similar fractal patterns. Chaotic attractors graphically illustrate both inherent variation and underlying order. Figure 1 illustrates the attractor for the weather system.

In the current research, instrumental applications dominate at each level, but strong indications of alternative, developmental patterns also appear. This recognition provides both a conceptual framework for considering the impacts of globalisation on public administration, and also a way to approach change. Because the levels are linked, action at any level affects the others, through complex feedback patterns.

The researcher’s third extension of complexity theory is the centrality of values, as both ethics and desired outcomes, in shaping the uses for information technology at all levels. In relation to applications of information technology in governance, a growing polarisation between globalising and democratising values is identified. If values are the drivers of behavioural attractors, then attention to this parameter suggests a critical point of leverage. Management values have been described as the source of the deepest order in organisations (Kiel 1994:218), and the current work suggests that values drive not just organisational uses of interactive technology, but also global applications. It will be shown that the globalising values driving the attractors for the uses of information technology on each level lead to instrumental applications that do not maximally enhance the complex interactivity which is essential for democratic process. Figure 2 offers an illustration of this dichotomy, or bifurcation. Again the value of this hypothesis is that assertion of democratic values at any point in the multi-layered system can contribute to a redirection which increases learning and adaptive feedback.

This thesis also contributes to the scant work on government uses of information technology and electronic democracy in Australia. Four case studies of computer-mediated communication, each involving different levels of governance, provide the evidence. The three minor case studies are of single interactive technologies: a list in a national agency, an international list, and a local experiment in electronic democracy. Some uses of the technology support democratic process and others do not; the evidence from these case studies is complex and sometimes contradictory. The largest case study is broadly based on interactive systems in Australia’s Department of Finance and Administration over two years. Interactive is here defined as the facility for two-way, open-ended communications. This means more than the ability to respond to a form. Interactivity is inherently creative and unpredictable. The Finance study presents data on organisational issues and user perspectives relating to the degree of participation in internal policy matters. This is analysed in relation to their uses of computer-mediated communication.

This may be the only study of an Australian government agency relating public sector reforms to the global system of technology. It may also be the only current study examining computerisation and industrial democracy in a public sector agency. The major case study uses informant data from all levels of the organisation to show how these elements interact. It provides evidence that the patterns of interactive technology use observed at national and supra-national levels repeat at the organisational level. It also documents, perhaps for the first time, how the sea change in public sector culture which has accompanied the reforms of recent decades impacts on the organisation and individuals. These impacts are inseparable from organisational uses of computer-mediated communication. The analysis developed here is consistent with the literature applying complexity theory to management. As well as extending the theoretical framework in the ways outlined above, it contributes to that literature by suggesting communication protocols for the use of technology which can assist the adaptation of organisations in situations of high environmental complexity and instability. These general communication protocols have implications for other areas where information technology applies to democratic forms of governance, by offering a view beyond the dominant service delivery model.

Part I presents the theoretical background for a broad systemic analysis of the role of interactive technologies in democratic policy processes. The remainder of this chapter discusses three key sets of concepts which inform the analysis. These are the selected definitions of democracy and policy, and the basic elements of complexity theory. Then, an outline is given of how this study relates to other areas of research. The approach taken here is unusually integrative and draws on a number of areas which are themselves interdisciplinary. Chapters 2, 3 and 4 provide the core of the theoretical background, and each contains a relevant literature review. These three chapters discuss information technology in three different contexts where democratic perspectives can occur. These correspond to the three scales of analysis: global, national, and organisational. Each presents a broad context, then current changes and pressures, a review of selected literature, and finally a perspective related to the core theories of this thesis. Each chapter proposes a set of hypotheses on the uses of interactive technology in systems of governance, broadly considered. Some of these find corroboration within other chapters, particularly in relation to the links between scales.

Chapter 2 considers the evidence that technology is socially determined. It also looks at the emerging information society in the international context of converging technologies and corporate activities, and reviews the literature on electronic democracy. Chapter 3 looks at the elements of public sector reform, and how these relate to the ways governments have used information technology. It analyses some Australian technology policy documents and projects from the perspective of democratic interactivity, and identifies an underlying model. Chapter 4 examines current management theory in relation to organisational participation, and the position of the individual within these systems. This is linked to a review of the literature on organisational computer-mediated communication. All three chapters are linked by the three theoretical perspectives which provide a framework for discussing democratic policy processes and complex systems.

Part II presents the empirical evidence. Chapter 5 formulates the theoretical framework more closely, and describes how a qualitative methodology was used to test the hypotheses. Chapter 6 is the major case study, and Chapter 7 contains the three minor case studies. Conclusions and recommendations are in Chapter 8. The Appendices supplement the research; in particular Appendix A lists existing efforts to outline rights and responsibilities in a global information society.

The theoretical position of this research rests on generic concepts of democracy and policy, and concepts from complexity theory. The triangulation of these concepts is used as a foundation which the following three chapters develop at different levels of analysis. Evidence from the case studies also helped shape the theory.

First, the concept of democracy is drawn from Dahl: ‘a unique process of making collective and binding decisions’ (1989:5). A necessary assumption is his Strong Principle of Equality: all the members of the association are adequately qualified to participate on an equal footing with the others in the process of governing the association (Dahl 1989:31). His criteria for democratic process could apply to associations at any level: effective participation, voting equality at the decisive stage, enlightened understanding, and control of the agenda. His citizen body is everyone satisfying the presumption that they are better judges of their own interests than others would be (1989:108). This framework has been chosen because of its generality. Five other aspects of Dahl’s analysis relate closely to the current research, which considers democratic uses of information technology in a global context. These are: his recognition that scale or size is a challenge for democracy; the dangers of control by intellectual elites; the role of democracy at the organisational level; the increasing power of trans-national corporate activities; and the potential of information technology to assist a modern transformation of democracy. Underpinning Dahl’s arguments is a belief in a collective, interrelated, socially defined common-wealth, rather than an isolated and individualised pursuit of benefits. He justifies political equality and democratic process as ‘the most reliable means for protecting and advancing the good and interests of all the persons subject to collective decisions’. His ‘implicit and real concern’ is freedom, human development, and human worth (1989:323). These place him clearly within a strong liberal tradition, with underlying values which see government as a set of practices to achieve the greatest good for the greatest number.

Dahl suggested that there have so far been two transformations of democratic practice, both related to issues of scale. The early assembly democracy of the Athenian state had an exclusive citizen body, but was small enough to make most decisions collectively. There was no need for a separate administrative arm, as individuals would take on various tasks as they arose. Participation was fairly universal among these citizens, the issues to be dealt with were limited, and relevant information was transparently available to the group. In that less technically developed age, information transparency, complexity and access were not critical issues. This form of democracy is still feasible within small associations.

Dahl’s second transformation of democracy was the representative structures of the nation state. Suffrage gradually became universal, and the necessary delegation of administration gave rise to bureaucracies and the institutionalisation of accompanying information flows. Mechanisms necessary to satisfy the criteria for democratic process on this larger scale included elected officials, institutions necessary to satisfy include free and fair elections, inclusive suffrage, the right to run for office, voting equality, effective participation, enlightened understanding, control of the agenda, and inclusion freedom of expression, alternative information, and rights of association (Dahl 1989:220, 222). These all require technical support and the involvement of specialists. In principle, a form of these institutions and rights could also apply within organisations. The effectiveness of this model is now compromised by the scope and complexity of issues which often transcend national borders (Dahl 1989). These challenges to pluralism are also the challenges of globalisation, including the difficulties of participation and consensus building in diverse mass societies. The problems now facing national governments repeat a pattern seen in the second transformation:

Dahl believed that inequalities in knowledge pose a greater threat to democracy than inequalities of resources or problems of trans-border influence (Dahl 1989:333). The specialist skills and knowledge of intellectual elites, including public bureaucracies, executive offices, legislatures, political parties, universities, research institutions, media, lobbying organisations, advisory groups, business firms, labour unions and law firms grant them an influence based on their ability to persuade, rather than outright coercion. He was particularly concerned with the public policy specialists who tend to be privileged and detached from the control of the demos (Dahl 1989:333-5). Dahl saw the democratisation of the workplace and democratic uses of information technology (which he called telecommunications) as essential for a third transformation of democracy appropriate for our age which would also address the problem of elites. Writing before the rise of the Internet as a major communication channel, he saw that it was technically possible to make appropriate information about the political agenda easily and universally accessible to all citizens and to make it possible for citizens to place questions on agenda, and participate in discussions with experts, policymakers, and fellow citizens. He also saw that ‘without a conscious and deliberate effort to use the new technology of telecommunication in behalf of democracy, it may well be used in ways harmful to democracy.’ (Dahl 1989:339).

Other writers have indicated the need for a reinvigoration of democracy at the local and workplace levels, and the role of the media in entrenching a market ideology, notably contributors to Hirst and Khilnani (1996). In that volume Wright noted that the private governance of corporations should not be divorced from the public governance of states, and mentioned on-line information flows as one of the avenues aimed at enhancing those democratic components of representation, accountability, participation, and openness. The issues of unaccountable elites, the inadequacy of 18^th century institutions for a complex globalised world and the need to harness new technologies for democratic ends may be seen as challenges to maintaining sufficient pluralism or diversity in the democratic system. Figure 8 in Chapter 2 illustrates how a leading electronic democracy activist conceptualised how interactive technologies might foster communication and collaboration between sectors.

Throughout this thesis, reference to the need for access to information, participation, and accountability, and the right to express a view and set agenda issues link back to the criteria given above for democratic process. Dahl’s argument for industrial democracy is considered in Chapter 4. His criteria for democratic participation also relates to Considine’s concept of public policy, which is the second theoretical underpinning for this research.

This model has been chosen for its generality, and also because of the emphasis placed on values, participation, policy as innovation and communication, and the developmental aspect of policy making. Considine’s definition of policy is applicable to any level of organisation, by deleting the word ‘public’: ‘the continuing work done by groups of policy actors who use available public institutions to articulate and express the things they value’ (1994:4). This emphasises all communications between actors along the way to major decision points and commitments. At any level of structure, analysis of the policy actors, their values, and their mechanisms of accountability will aid understanding. The interest here is on how groups of actors influence the policies on uses of information technology in governance, and how this fits into a democratic framework.

Considine, like Dahl, assumed a framework of legitimacy, values and rights. His actor networks lie between institutional power and individual citizens, and arise from key social groups such as parties, corporations, unions, professions, and citizens, with patterns of interdependence between them. They work through complex and changing alliances to influence outcomes towards their preferred values. Considine recognised dual aspects of policy making. It always has a material, instrumental, output oriented side. Beyond that, there is a developmental or intellectual side, where process is also an outcome, and a site of continual struggle between groups of actors and their values (1994:130). Although Considine did not consider the role of technology in these processes, Chapter 2 will show that this dual analysis resembles closely Zuboff’s (1988) ‘informating’ quality of technology. The developmental and the instrumental aspects of policy making are interrelated, not separate variables which can be tracked and modified independently. Democracy has a similar duality: protective versus developmental. Protective democracy deals with issues having practical or immediate import, and corresponds to the instrumental aspect of policy, whereas developmental democracy has longer term implications for citizens’ ability to shape their political environment. It is also a condition for learning, and involves awareness of being both subject and object of governance, through what have been called ‘reflective’ rights (Habermas 1992).

This duality is a central concept for the current research. When governments consider information technology, instrumental values such as efficiency and accuracy, cost-effectiveness and elimination of unnecessary duplication are undeniably important. At the same time, the communication and interactivity dimension of information technology has implications for participation in the policy process that reflectively determines its uses. The theory chapters consider this potential at different scales, and in different disciplines. The literature on learning organisations uses the terminology ‘single’ and ‘double loop’ learning in a similar context. For Considine, the key to the meaning of policy systems was the values concealed in the actions, as ‘values explain policy and invest all its ordinary practices with meaning.’ Values are ‘titanic and ubiquitous’ (1994:48-49). Ubiquitous features are often ignored, or accepted as unchallenged assumptions. The analysis of information technology policy systems in the following chapters will seek to identify the values underpinning the actions of actor-networks at each level.

The centrality of participation in Considine’s model of policy relates directly to democratic process. Given the dual role of policy making, it is mistaken to assume that only outcomes count, or that emphasis on consultation or participation is expensive and counterproductive. Considine saw participation as a primary structure for all policy development and implementation. It facilitates rational deliberation, creates and communicates moral principles, and expresses personal and group affects and needs. Participation can generate the social capital from which ‘all central democratic objectives spring, including legitimacy, co-operation and innovation’ (1994:130). Thus, he argued all policy making must be based on the widest possible level of participation, consistent with an effective process of decision making. The benefits are not just improved outcomes (instrumental value), but developmental values, including improvements in the capacity of the system or community, through better knowledge, understanding, solidarity, trust and sympathy (1994:130-131). Participation has risks for the powerful, and ‘public institutions will frequently oppose and subvert participatory systems’ (1994:143-144).

Considine’s actor networks are fully active on both levels of policy: they represent fixed interests for their group, but also absorb, reflect and develop ideas. They both act and are acted upon. No one group has unlimited authority or power, but acts in a system of changing internal and external factors. Thus, public policy is a process of innovation and learning, and a form of communication. Actor networks are ‘knowledge communities’ which ‘rely upon lateral forms of organisation rather than traditional forms of hierarchy’ (1994:267). This model relates closely to the complex adaptive systems described below. Considine saw an urgent need for both the public and private sectors to find new forms of direct accountability to the public (1994:232) which can eliminate the ‘stovepipe’ effects of traditional hierarchical decision making (1994:263).

New accountability mechanisms are part of the search for new democratic protocols which complement representative structures. Decision making, power and accountability in hierarchical systems flow upwards only. Senior managers are required to both lead and devolve, while also obeying their political masters, public or private. This linear model precludes responsivity to feedback and information from below. The following chapters illustrate how interactive technologies are already being used to create lateral communication channels that cut across traditional structures at many levels. Complexity theory helps to show how opening the system to broader inputs is adaptive, but also introduces the ‘messiness’ of democracy. Interactive technology can both contribute to and help manage this disorder.

The third major theoretical underpinning of this thesis is complexity theory as applied to human systems. Stacey’s (1996) work on complex systems and organisational dynamics reveals that organisations, and indeed, all human systems may be thought of as complex adaptive nonlinear systems. That is, the relationship of the components to each other changes over time. It is not possible to assume that a particular input will always have the same output. In such systems, feedback creates important characteristics within the system. Negative feedback operates like a regulator to bring systems back to a point of equilibrium. Positive feedback amplifies and pushes a system further from stability. Time is the critical variable, as often phasing is the sole determinant of whether feedback will be amplifying or stabilising. Economic examples include the emergence of the QUERTY keyboard and VHS as a standard for video machines, although neither necessarily offered the best technical solution (Waldrop 1994). In human networks, these processes are commonly understood as ‘self-fulfilling prophesies’ or ‘vicious circles’. In these nonlinear systems innovation and creativity can occur at the ‘edge of chaos’, or when instability threatens collapse of the system. This is also where complex, adaptive or ‘double loop’ learning occurs, or where a system shifts into a new state of stability. Such systems are inherently unpredictable, but their instability tends to be ‘bounded’ or operate within limits. This inherent unpredictability offers an alternative perspective to the Newtonian, linear and predictable world which was the basis for Weberian clock-work bureaucracies (Kiel 1994:12). One implication is that such systems are not subject to managerial control. Management has a different role in such situations, which are the norm in today’s unstable business and government environment. Rather than seeking stability and maintaining an illusion of predicability or control, it becomes more important to encourage the diversity and self-organisation which will provide creative adaptive solutions. This tends to happen most productively in the ‘shadow culture’ of informal interactions immune to regulation (Stacey 1996).

Nonlinear systems are also extremely sensitive to minuscule changes in initial conditions. That is, a subtle difference in a single parameter can lead to totally divergent outcomes. Paradoxically, nonlinear systems also regulate themselves by tending to return to their basic pattern of behaviour when disturbed. This mode locking has been an important discovery of complexity theory in physiological systems (Gleick 1988:292-3), and has parallels in, for example, bureaucratic systems that resist change. Interactive systems, such as computer-mediated communication, are also feedback systems: they can create surprises, even while showing identifiable patterns. Complexity theory shows that rather than suppressing the resulting disorder and instability as undesirable risks, these processes can be accepted as inherent in all forms of evolution, adaptation and renewal. In this sense, the ‘messiness’ of democracy is not only ‘natural’, but important for survival at every level of human systems, including the organisational.

The patterns which emerge in nonlinear systems are called ‘attractors’, because the system is ‘attracted’ to this state. They have also been described as ‘the deep order in work’ (Kiel 1994:103), because they fall within limiting boundaries. These patterns are determined by the system, but are not predictable in detail, any more than one can predict exactly what a day at the office will be like, although most people can safely assume it will not include a visiting circus. Figure 3 illustrates an attractor created using data from a public administration context

These attractors provide the emergent properties of the system, not deducible from the components alone. The abstracted pattern, or attractor, is the union of all transformations of itself (Barnsley 1998). Thus, highly interactive computerised systems may produce patterns which satisfy the criteria for democratic process, but not necessarily. Unlike physical systems, human systems are subject to ethical influences. Discovery of how these impact on the complex relationships created by information technology and human structures for governance would have both predictive and practical applications.

Observations made by Kiel (1994) in his analysis of complexity in government have also informed the current research. He saw the nonlinear dynamic as a true paradigm shift, away from Weberian, hierarchical models. In this new paradigm, disorder and even bounded instability is an opportunity, rather than a threat (p. 15). He suggested the possibility of renewal through transformation, but only if risks are encouraged. These are not the wild and reckless risks of the anarchic, but rather those necessary to explore new possibilities. He said freedom is more important than control (p. 205), and process is more important than structure (p 209). Kiel’s assertion that highly energised organisations are more responsive to change (p. 213) has implications for agencies which have experienced increasing enervation in recent years. He asserts that an organisation’s deepest order is found in management values (p. 218). Therefore, a commitment to service, a desire to excel, a drive to improve, and a dedication to having open organisations that let democracy flourish are essential management values in a world of increasing complexity (p. 218). In this way, public agencies can be a stabilising force in an unstable world (p. 223), supported by a new information ethic in public agencies that extends beyond the government’s fiduciary responsibility to maintaining information for citizens (p. 216). Kiel’s conclusions, supported by mapping of attractors based on work process and employee performance data, provide early indications that mathematical analysis of public sector processes can yield new ways of conceptualising and discussing these dynamic activities.

Notable for its lack of attention to information technology is the public policy literature (Henman 1996). Within Australia, two popular current texts on public policy make no mention of the growing importance of information technology for this field (Considine 1994, Davis et al 1988). Zifcak’s (1994) analysis of public sector reforms in England and Australia also ignored the mushrooming role of information technology. Australian public policy on information technology does not seem to have attracted the attention of the public policy community, and it seems no previous study has analysed major information technology policy documents as they relate to citizen participation or transparency in government. The literature on public sector reform is often overtly concerned with democratic process, but these rarely reflect on either technology or the impacts at the organisational level. Williams (1991) is an exception, but he did not address the policy process. Perhaps the most closely related study of information technology in the policy process is Gualtieri (1998), who looked at several countries and found there had been little impact. He found wide agreement that new communication technologies have had a positive impact on delivery of government services and programs. However, he found no evidence that decision making had either improved or become more democratic. Rather, the policy process itself had become more complex, with little interest from either citizens or policy makers in enhancing participation. The current study confirms his findings for Australia and links them to internal culture and events in a public sector agency.

Writers in areas such as computer science or technology and society have pointed out the importance of information systems to government, and the many issues involved (Rosenberg 1997, Lyon 1988, Branscomb 1994). However, they have tended to focus on computerisation for government administrative systems rather than for public policy development. Margetts’ (1996) study of democratic accountability and information technology in the UK government and Henman’s (1996) examination of public policy and computerisation in Australia’s Department of Social Security will be considered in Chapter 3. Both provided evidence which helped develop the current perspective on information technology, by emphasising its globalising (Margetts) and controlling (Henman) patterns. Morrigan (1997) examined the Australian Department of Taxation from a learning organisation perspective, but did not relate this to industrial democracy or information technology. However, she placed the constraints she found on ‘knowledgability’ in a globalisation context, thus indirectly revealing similar processes in a different public sector workplace to the major case study of this thesis.

The computer literature is primarily technical, with minor attention to social impacts or political process. The social impact of computers is a marginalised area academically and ethical perspectives on information technology are also largely absent (Lyon 1988:153). In particular, few attempts have been made to take an integrated view of the uses of information technology across scales of application. At the organisational level there is an extensive business literature on the implementation of information technology in relation to change management, business process reengineering and other aspects of instrumental efficiency. At the global end of the scale are many analyses of convergent media, but economic assessments of technology have not generally considered the social impacts (Preissl 1997). On the other hand, communication studies have focussed on social outcomes, equity and sustainability. The mass media and its implications for citizen communication has provided much of the impetus for discussions of electronic democracy. The extensive literature on organisational computer-mediated communication has considered the democratising effects of interactive technologies in some depth. Zuboff’s (1988) work on the transformative effects of computerisation on workplace power relations presented a key set of insights on the role of technology in wider, more generally political processes. The present study draws connections across scales by looking for the underlying patterns.

Industrial democracy, technologically implemented or not, seems to be less studied in the 1990s than previously. Not has it often been taken seriously as a possible contributor to business success. Semler’s (1994) unusual and successful experiment with industrial democracy is considered in Chapter 4. The perspective of complexity theory offered here, along with the recognition of the potential role of interactive technology in organisations, has clear implications for notions of industrial empowerment and its limits in traditionally structured bureaucratic systems. Some of the literature on industrial democracy has considered technology, particularly the studies in Aungles (1991), and the work on participatory design, such as (Fiorilli 1997). Much of this has been in relation to process or manufacturing work, rather than in the more abstracted realm of policy, or was written before the rise of interactive technologies (for example Bjerknes, Ehn and Kyng 1987). The role of computerisation in the democratisation of internal policy processes is therefore a little-researched area in Australia.

The literature on public sector reform has given little attention to impacts on the organisational level. Nor has the relation of these reforms to globalisation been articulated theoretically as in this thesis. Methodologies of some public sector studies have focussed on higher level staff or relied on theoretical abstractions, thus missing the opportunity to document the value shift which has accompanied the reforms. These will be discussed in the methodology section of Chapter 5.

The literature on democratic theory often gives scant attention to the role of technology in governance (for example Held 1996:322). Dahl’s position, as discussed above, is an exception, but his pre-Internet writing did not develop the ideas in any depth (Dahl 1989:339). Writers on the emerging information society have, however, given extensive consideration to the democratic implications (for example Lyon 1988, Haywood 1995, and Brown 1998). The literature on electronic democracy will be reviewed in Chapter 2. In general, it addresses democracy at the national or regional level. Only rarely does this literature refer to electronic democracy at either the global or organisational level. The researcher knows no studies or theories which integrate these perspectives and levels. An Australian work on deliberative democracy (Uhr 1998) focuses on the role of Parliament and the history of political theory; it does not address the role of new technologies in democratic deliberation.

Evidence from both the literature reviews and the case studies of fractal patterns and attractors in the uses of information technology and the structures which support them are used to construct a global systemic analysis. Exploring these systemic relationships is one of the tasks of the current research. Another task is to explore how the nonlinear systems of information technology might support the ‘democratic ethic’, as Kiel (1994) asserts. The complex learning and self organisation that occurs in systems at the edge of chaos has metaphysical and even mythological dimensions (Briggs and Peat 1989). It also has the simpler meaning that old forms, ideas and values are not replaced without turbulence and sometimes, destruction. The double-loop learning which takes places in complex systems has much in common with developmental democracy and innovative policy processes: feedback and participation are related. The application of complexity theory to the social sciences may be part of a millennial shift to nonlinear thinking. It is used here to extend systems thinking to the wider spheres of globalisation and information technology. Self-organisation, however, is not the same as democracy. Encouraging participation within a pluralistic political system simply spreads the winners and losers more broadly. Underlying defence mechanisms, such as obstruction of implementation, inadequate data for decision making, and control of the agenda still operate (Stacey 1996:399). While no proposal for refining democratic process can ever be complete, this thesis suggests that interactive technology with particular structural properties has a role to play in delivering ‘best practice’ democracy over what is likely to be a unstable start to the next millennium.

Indications of a complexity perspective were voiced more than 30 years ago. Emery and Trist (1965) called for a wider concept of systems, and noted that equilibrium models of classical physics were inappropriate for economic modelling. This prefigured Beniger's work by describing living systems as open systems, capable of attaining various states. Beniger (1986) is considered in Chapter 2. Emery and Trist also noted that turbulent systems were increasing, due to the ‘deepening interdependence between economic and other facets of society’. Emery and Trist highlighted the importance of social values which have the conceptual character of ‘power fields’ and act as injunctions. Others have noted the feedback system between values and technology (Perugini 1996, Beder 1994), or suggested that business is ‘a fountainhead of values’ (Talbott 1997). The merging of democracy and ethics as a consequence of globalisation is another possibility; self-regulation would then encompass the values of accessibility, equitable distribution, cultural identity, social justice, meaningful work, and sustainability (Berleur 1995).

Through the theoretical development and the case study evidence, the researcher explores the importance of values and communication technologies in the self-organisation of new forms of democracy. Together, this perspective is intended to meet the requirements of a practical theory, one which clarifies possibilities for action (Kiel 1994:16-17). Such a theory could contribute to a concept of an information ecology which realises the interconnectedness of many activities previously seen in isolation.

In the public sector workplace, issues of empowerment, computerisation, and impacts of internal decisions on wider social outcomes are mixed together, part of the normal day’s work flow. These are seldom articulated as issues, and rarely analysed as either practical or research areas. For heuristic purposes, the following chapters use a fractal pattern to develop the theoretical argument. The broadest level of analysis is the global: what are the factors affecting democracy in an information age?