Rethinking project management

“the challenge for research…. is precisely the perceived weakness of the discipline's theoretical base”

Peter Morris’ conclusion to his keynote speech at the first Project Management Institute Management (PMI) Research Conference in Paris (Morris, 2002, p. 53) echoes a widespread perception of researchers and reflective practitioners in the project management field. Disciplines – in both the academic and professional senses – mature through the development of a coherent body of ideas that deepens understanding and enables predictive propositions. Yet, this coherent body of ideas – let us call it a theory – is widely perceived to be immature in the discipline of project management. The perceived weaknesses of the project management discipline such as its under-valuation by senior executives or its reduction to a single chapter of an operations management text (e.g., Slack, Chambers, & Johnston, 2003) in MBA courses are seen as symptoms of this underlying immaturity. This perceived under-evaluation has been the inspiration for much of PMI‘s own research program in recent years – see, for instance, work by Ibbs & Reginato (2002) and Thomas, Delisle, & Jugdev (2002).

This keynote address will attempt to address this immaturity and thereby propose one possible way of theorizing project management. First, it will review the current state of the accepted canon of project management theory, suggesting that it contains many important elements that articulate the distinctiveness of project management as a discipline. Second, it will review the development of organization theory over the last 40 years, showing how this already contains the building blocks of a mature theory of project management, and build on the work of Jay Galbraith to push the debate on. Third, the paper will identify one of the large gaps in both existing PM theory and PM-focused organization theory in terms of the dynamics of inter-firm relationships in the project coalition, and suggest how it might be filled in a manner consistent with the first two elements through building on the work of Oliver Williamson. In so doing, it will attempt to articulate the outlines of a theory of project organizations as information-processing systems.

The Development of Project Management Theory

Mainstream project management theory grew from the attempt to understand the new organizational forms that were implemented for the development of cold war weapons systems after 1945, and matured within NASA for the Apollo mission. Perhaps the first formal statement was Gaddis’ (1959) characterization of the project manager as the “man in between” the demands of technologists and the demands of senior management responsible for delivering the required asset to the agreed schedule, budget, and specification. Cleland and King (first published in 1968) drew largely on the experiences of the same projects to articulate and illustrate their theory of project organization as a dynamic system. Their approach builds on the theory of an organization as an open system which makes and implements decisions through a process of planning, feedback and control. In their analysis, they identify many of the key elements of project management theory that remain with us today (1983, part 4):

• the project organization is a temporary matrix coordinated by the project management team;
• projects move through a distinctive life-cycle from inception to completion;
• the project management process is inherently one of planning, and then control against that plan, through feedback from implementation;
• projects organizations deliver against objectives set outside the project organization in terms of budget, schedule and specification;
• tools and techniques such as the work breakdown structure and critical path method play an enabling role in the planning and control process.

For Cleland and King, project management was the process by which the outcome of decisions taken through systems analysis were implemented, and its inherent complexity reflected the complexity of the problems for which it was providing a solution. An important underlying assumption of systems analysis was that the phenomenon under analysis could be modelled “as a representation of a system which is used to predict the effect of changes in certain aspects of the system on the performance of the system” (1983, p. 133). However, as Cleland and King recognised, many aspects of the project organization system could not be measured, and so in terms of project implementation, systems analysis was, in effect, limited to the control of schedule and budget. It is these systems analysis aspects of project management that have dominated subsequent research in project management. But as a much more recent advocate of the application of systems dynamics techniques to project management has argued, project modelling is not the same as project management, but an aid to effective project management (Williams, 2002, p. 1).

Perhaps the major limitation of the systems analysis approach, which has been the predominant theory in project management for the past 30 years, is that it is fundamentally deterministic. That is to say, it assumes that the scope of the project is completely knowable in advance; that an appropriate plan can be developed to deliver that scope; and that the problem of control is simply to keep the project delivery to plan. Although there is a discussion of decision-making under uncertainty (1983, chapter 5), it presumes that the complete range of possible outcomes can be specified in advance, and similar comments apply to the PERT approach to schedule planning.

A further limitation of the systems analysis approach is that it pays little attention to a very important feature of most project organizations – that they are coalitions of independent firms and are chartered through legally-binding commercial contracts. Cleland and King pay no attention to this point, while many texts in project management do little better. This is a remarkable oversight because:

• all projects are launched at the behest of a client, and that client is typically a separate organization from the asset supplier, except in new product development projects;
• many projects involve the use of subcontractors as technology suppliers to the principal contractor.

The systems analysis theory of project management articulated by Cleland and King has become a paradigm (Kuhn, 1970) in two very important senses. First, it has provided a framework for subsequent research in project management over the last 40 years; second, it has been distilled into, for instance, the PMI Project Management Body of Knowledge and the PRINCE 2 methodology, thereby shaping practice as well as theory. This is not to argue that the framework has not developed – it has, but along the lines of what Kuhn called “normal science,” which takes premises of the paradigm as given, and engages in incremental puzzle-solving research within the paradigm. The analysis of Kloppenborg and Opfer for the first PMI Research Conference (2002) shows how research has focused on planning and control, particularly with respect to time, cost and quality, on through the 1990s, and how commercial issues have been virtually ignored. Morris (1994) argues that the most significant new development during that period has been the growth of interest in risk management. Even here, the research followed the systems analysis paradigm. The emphasis was on the development and application of tools and techniques such as Monte Carlo analysis to well-defined problems – Chapman and Ward (2003) provide an authoritative review - rather than grappling with the uncertainty facing project decision-makers.

If there is now a widespread perception that there are fundamental problems with the systems analysis paradigm of project management – see, for instance, Morris (1994); and Giard & Midler (1993) - there is also a perception that little progress has been made in developing a new paradigm that incorporates the lasting merits of that paradigm, while providing a theory that provides us with the intellectual tools to address the issues identified above. There has been a growing number of models of the project management (see Morris, 1994; Morris, 2001; and Turner, 1999 for reviews), and while these can be thought of as notes for a new theory, they are necessarily of such a summary nature that they cannot provide the sort of conceptual orientation that the discipline – academics and practitioners – arguably needs. With the aim of providing such conceptual orientation, the argument will now turn to a body of literature that developed in parallel with the systems analysis paradigm in project management – what can be called project-orientated organization theory.

The Development of Project-orientated Organization Theory

As management practice developed into a well-articulated theory of management during the first half of the 20th century (e.g., Fayol, 1925; Barnard, 1938; Gulick & Urwick, 1937) a consensus developed on the “one best way” to manage effectively and efficiently. This perspective was increasingly challenged during the 1960s by the contingency theorists who argued that the most effective way to manage depended on the activity being managed. For project management theory, this work had some important implications, because it identified:

• the distinctiveness of the unit and small batch approach to production (Woodward, first published 1958) compared to the large batch and mass production traditionally associated with manufacturing;
• the distinctiveness of organic forms of organizing where innovation was required compared to the more mechanistic forms where it was less important (Burns & Stalker, first published 1961)
• the importance of the variability in the market faced by the firm in the most effective mode of organizing (Lawrence & Lorsch, 1967).

A common theme running through these contributions is the greater importance of coordination through non-administrative means when management decision-makers face greater uncertainties. Woodward (1980, chapters 8 and 9) stresses the importance of personal interaction among personnel from the commercial, engineering, and manufacturing functions in unit and small-batch production, compared to the more systematized procedures in large-batch and mass production. Typically, this was either coordinated by the chief executive himself, or facilitated by the commercial engineers responsible liaison with the client, and there is no indication of any project form of organization in her case studies. Burns and Stalker did find evidence of formal coordination roles in their cases studies of defence electronics firms, and identified the role of the project engineer in “taking care of troubles arising with test, drawing office and development” (1994, p. 170). Lawrence & Lorsch (1967) focus considerable attention on the role of integrators between different functions, and show that the higher the levels of uncertainty faced by the organization, the more sophisticated the coordination arrangements. However, most of the coordination activities that they discuss were performed either by liaison committees, or individual liaison staff. The analysis of coordination roles was formalized by Galbraith (1977), which drew to an important extent on his research (1970) in Boeing. He identified and classified a whole set of coordination mechanisms and placed the role of the project manager at the pinnacle of this suite of organizational arrangements for solving the coordination problem. Mintzberg (1979) relied heavily on these authors for his analysis of the adhocratic nature of project organization.

In an important sense, Mintzberg's book represented a landmark in the analysis of organization design, synthesising and summarising what we know about how organizations are structured. Research attention moved elsewhere, particularly to issues related to strategy, and more recently to innovation. It was from the latter research theme that an academic interest in project management was rekindled in the early 1990s. Perhaps the first, and arguably still most important, contribution was Clark and Fujimoto's (1991) analysis of new product development in the global motor industry. Complementing the operational manufacturing focus of the seminal The Machine that Changed the World (Womack, Jones, & Roos, 1990), Clark & Fujimoto (1991, chapter 8) identified the growing importance of project management techniques in shortening development cycle schedules and budgets. In particular, they identified the importance of the “heavyweight product manager” in ensuring effective project delivery to “job 1.” Called, variously, the shusa at Toyota (Womack et al., 1990) or program manager at Ford (Walton, 1997), this new role has equal status with the resource-base managers and represents a significant enhancement in responsibility compared to the traditional lightweight, liaison-focused project manager. Perhaps most importantly, heavyweight product managers have responsibility for interfacing with the market. Midler (1995) provides further evidence of this development in the car industry in his work on Renault, while later work on Toyota (Cusumano & Nobeoka, 1998) identified the importance of managing the total portfolio of development projects in an integrated manner.

Others working more broadly across the manufacturing sector have identified a similar shift. The Massachusetts Institute of Technology (MIT) study of five manufacturing firms (Bowen, Clark, Holloway, & Wheelwright, 1994) identifies the crucial role of “heavyweight project leaders” in effective new product development, and, through the development of successful products, corporate renewal. Heavyweight project managers are distinguished from their lightweight counterparts by total responsibility for the process of project delivery, and responsibility for interfacing with the client or customer, thereby achieving both “internal” and “external integration” in new product development. Womack & Jones’ discussion of lean thinking – while principally focused on operations management rather than project management – identifies the development of the role of the “project engineer” at Pratt and Whitney during the 1930s as an precursor to the lean approach (1995, p. 155). Winch's (1994) analysis of 15 United Kingdom (UK) metalworking firms also identifies the emergence of “program managers” who reported directly to the managing director, as distinct from project managers who reported within functions such as engineering and manufacturing.

As the innovation and new product development literature rediscovered the importance of project management, particularly in its new heavyweight form, innovation researchers concerned with the capital goods sector were also rediscovering it. Recent research – particularly on the flight simulation industry (Miller, Hobday, Leroux-Demers, & Olleros, 1995; Rosenkopf & Tushman, 1998) - has argued that there is a qualitatively different type of innovation process in the industries which manufacture what they characterise as “complex product systems” (CoPS), as opposed to the volume sectors such as autos researched by those concerned with new product development discussed above. Examples of CoPS include rail signalling systems; jet propulsion systems; power generation equipment, and bespoke information systems. The organization of production in complex systems industries is characterised by:

• its orientation around a project,
• the contribution of temporary coalitions of firms to production,
• the heavy involvement of the client in the process,
• and, most notably, the failure of the industry to restructure to reflect the dynamics of the product life cycle (Utterback, 1994) familiar in the industries where the new product development model of project organization is the norm.

The early work by Miller and Hobday has stimulated a large body of work, which has begun to explore the development of project organization in the capital goods sector (e.g., Hughes, 1998; Miller & Lessard, 2000; Prencipe, Davies, & Hobday, 2003). A major conclusion is that the core competencies of CoPS suppliers lie in the combination of project management capabilities and systems integration capabilities, rather than leadership in specific technologies, and that learning from projects is the key management challenge for CoPS supplier firms (Hobday, 2000; Lampel, 2001).

Building on Galbraith: The Project as an Information Processing System

Galbraith's starting point is that all organizations are, in essence, information processing systems. In order to function, they must monitor their environment, take decisions, communicate their intentions, and ensure that what they intend to happen does happen. In manufacturing organizations, these information flows generate and control flows of materials as well, but many service organizations are purely devoted to managing flows of information. Information flows are the heart of the business process in all organizations. These information flows are directed and enabled by the structure of the organization, and the problem of management is the problem of continually shaping processes by manipulating the structure – what has been called the tectonic approach to organization (Winch, 1994).

The fundamental problem in the management of information is uncertainty; in other words, the lack of all the information required to take a decision at a given point in time. Galbraith (1977, p. 38) formalizes the definition of uncertainty as the difference between the information required for a decision, and the information available at the point in time when the decision is to be made. Uncertainty is, by definition, unmeasurable, but it is possible to have a subjective feel for how much information is missing at the decision-point.

This uncertainty has two sources (Winch, 2002):

• complexity, or the condition where the information is, in principle, available but is too costly or time-consuming to collect and analyse;
• and predictability, or the condition where the past is not a reliable guide to the future. The future is, by definition, unknowable, but past experience is a valuable, if not infallible, guide to the future in many situations.

Where enough data are available to assign meaningful probabilities to the information required, then we can talk of risk rather than uncertainty (Knight, 1921).

Exhibit 1 The Project Process as the Dynamic Reduction of Uncertainty through Time
source: Winch (2002)

At the inception stages of a project, uncertainty is very high - the asset of the future is little more than an idea and possibly a few sketches. How high depends upon a number of factors such as the extent to which the asset is a copy of ones already existing; the extent to which standardized components and solutions can be used; and the extent of the requirement for new technologies to solve the particular problems posed by the project. This may be thought of as the level of mission uncertainty inherent in the project. As the project moves through the life cycle, uncertainty is reduced as more information becomes available – ambiguities in design are resolved; technology trials are completed; regulatory approval is obtained; component suppliers provide their shop drawings, and contractors successfully complete their tasks. The level of uncertainty at a particular point in the project life cycle relative to earlier and later points in the project life-cycle may be thought of as the level of dynamic uncertainty on the project. This framework is illustrated in exhibit 1, which shows how uncertainty is progressively reduced through time until all the information required for the project is available at completion and embodied in the asset created. The area to the left of the s-curve represents information still to be acquired, that is, uncertainty; that to the right represents what is known that is, certainty.

This perspective on project organizations as information processing systems that progressively reduce uncertainty through time can help us bind many of the disparate aspects of project practice into a coherent framework. The work of Shenhar and his colleagues (see, especially, Shenhar & Dvir, 1995; Dvir, Lipovetsky, Shenhar, & Tishler, 1998) has already demonstrated the quality of insights that this perspective can yield in distinguishing between different types of projects and identifying critical success factors, but it is of more general applicability in developing a theory of project management. For instance:

• it provides insight into how project organizations change through the project life-cycle. The sorts of project organization and project management competencies that are appropriate for early-stage project definition are very different from those required for effective project delivery. For instance, managing Research and Development (R&D) scientists or design engineers is a very different task from managing assembly fitters during erection – the difference is essentially the level of uncertainty faced in task execution;
• it helps us to identify where the tools and techniques of the systems analysis approach can be most effectively deployed. They all share the need for quantitative data input for their analytic power to be released; it follows that they are most useful in the later stages of the project life-cycle once uncertainties have been reduced to risks through the acquisition of adequate information;
• it identifies the vital nature of project leadership, because leadership is, in essence, a requirement of decision-making under uncertainty (Barnard, 1938). Under certainty, decision-makers can follow prescribed procedures manuals; under uncertainty, judgement is required.
• it provides a common basis for the consideration of the commercial issues to which we now turn.

Building on Williamson: Projects and the Mobilization of Resource Bases

Projects mobilize capital and human resources. The capital that finances the process comes from the client and its financiers. The human resources that enable the progressive reduction of uncertainty through time are supplied by the firms on the supply side, which act as resource bases for the project. Resource bases are the permanent – in the sense of not having a pre-determined life - containers of the skills and technologies required for the definition and execution of the project. The resources may be the capability to design and manufacture sophisticated components at the high-technology end of the spectrum, or the ability to mobilize large gangs of labor at the other. These resource bases may be individual firms acting as contractors on the project, or they may be internal to the client organization and organized as specialist departments. However constituted, they are fundamentally different from the temporary project organization which they nurture. These groupings of resource bases are often called the project team. However, the numbers of people involved are, in practice, too large to be meaningfully called a team. Moreover, all these different resource bases have different interests. We can more usefully think of these groupings of resource bases mobilized on the project as the project coalition, which comes together around shared objectives so that each member can meet its individual objectives. The term “project team” is then reserved for those actually charged with the task of coordination. One of the main reasons why interests differ is that most resource bases will be supplying resources to more than one project at once, and can find themselves juggling resources between projects. We can, therefore, most usefully think of projects as coalitions of resource bases, and resource-base firms and functional departments as participating in portfolios of projects.

Exhibit 2 Determining Appropriate Transaction Governance Mode
Source: Winch (2001) figure 1

The conceptual framework for the analysis in this section is derived from institutional economics, particularly the work of Oliver Williamson (see, especially, 1975; 1985). Williamson's basic proposition is that total costs of supply are derived from two main components – production costs and transaction costs. Production costs are well understood and, in essence, involve the efficient transformation of inputs into outputs, where prices are used to signal the most efficient choice of technology. Transaction costs are the costs of coordinating any complex production process, and occur whenever a good or service crosses a “technologically separable interface.” An important feature of this definition – which is often overlooked – is that it is neutral between the transaction being between two departments within the same organization, or between two different organizations. The problem of the governance of transactions and their associated costs occurs whichever option is taken, and the formal problem of the governance framework is the comparative evaluation of the internal and external supply of resources for the project. Thus, we can talk of internal contracts (Nahkla & Soler, 1996) within the framework. More recent research in the context of project organizations has also convincingly argued that transaction costs and production costs are not necessarily independent (Turner and Simister, 2001). For instance, incentive arrangements in the project governance arrangements can be used to economize on production costs by motivating suppliers to find efficiency savings which can then be shared with the client – see Scott (2001) for the North Sea oil and gas industry case.

Williamson argues that there are two basic options for coordinating – or governing – transactions. A market transaction is where independent buyers and sellers meet in the market to negotiate the price for supply of a good or service in a spot contract – prices are set by what Adam Smith called the “invisible hand” of the market. An hierarchical transaction is where the transaction is governed internally by administrative means – prices are determined by what Alfred Chandler called the “visible hand” of management through an authority relation. In between these two polar forms of transaction governance lie a wide variety of mixed form of relational contracts. It is these relational contracts which will form the focus of the discussion here, as the pure market and pure hierarchy have relatively limited application in project coalitions (Stinchcombe & Heimer, 1985).

What determines the most efficient governance mode on a project? Williamson argued that there were three main characteristics of transactions which influenced the choice of how they are governed – uncertainty, frequency, and asset specificity. Uncertainty – whether mission or dynamic - affects transactions because it creates bounded rationality for decision-makers. This bounded rationality makes writing a complete and unambiguous contract between the parties impossible due to uncertainty regarding the precise conditions under which the contract will be executed, and also makes it impossible to measure fully the performance of the contract. Asset specificity is the condition where either the buyer or supplier is limited in their choice of transaction partner due to the specific nature of the resources to be supplied. This asset specificity may be pre-contract, in which case the problem is one of monopoly or monopsony in the market, or it may be generated post-contract because contract-specific investments are made by one or both parties. This generates the possibility of opportunism on the part of one of the parties as they exploit the other's disadvantage. This opportunism often takes the form of withholding information from the other party. Frequency affects transaction governance because one-off transactions provide no opportunity to learn about the other party, while repeated transactions allow learning about the behavior of the other party, and hence the generation of trust, where trust is the confidence that the parties to the transaction will not take advantage of asset specificities to behave opportunistically (Lyons & Mehta, 1997) – either by withholding information or seeking monopoly rents. Thus the most appropriate choice of transaction governance mode can be thought of as occupying a three-dimensional space in the manner indicated in exhibit 2.

This approach can help us understand various aspects of project organization:

• It explains the considerable differences in the type of contracts written for the supply of highly uncertain research and development, compared to the relatively certain contracts for the supply of components. The latter are much more likely to be on a reimbursable basis, while the latter are more likely to be fixed price – see Winch (2002) for an extended analysis in the context of construction projects;
• Just as the principal problem in the theory of the firm is why firms exist, so, arguably, the principal problem in the theory of project management is why projects exist. One answer to this question could be that project organizations economize on both production transaction costs by clearly articulating buyer/supplier relationships in the context of complex organizational forms - Turner & Müller (2003) provide one argument along these lines.
• It addresses important issues in the development of project plans related to the “hold-up problem” where resource bases that are contracted to execute tasks that lie on the critical path can behave opportunistically to exert higher prices for their work – see the study by Masten, Meehan, & Snyder, (1991) of shipbuilding.
• It explains why project managers tolerate poorly performing resource-bases because the switching costs of removing them from the project are high compared to the costs of their non-performance – see Winch (2002) for an analysis.

The Implications of the Approach

The conceptual framework proposed here suggests that projects are temporary organizations that create new – or extensively refurbished - assets through the dynamic reduction of uncertainty through time. The role of the project management team is to coordinate the definition of the project mission through negotiations with all relevant stakeholders and to co-ordinate the execution of all the tasks required for the project by the resource bases in the interests of the client (i.e., the provider of capital). These resource bases are mobilized through coalitions of firms and functional departments operating as the members of a temporary network. The project management toolkit is an essential aid to this process, especially during the later stages of the project life cycle.

This approach has several implications for our conventional understanding of project management:

• The management of uncertainty is project management; absent uncertainty and management becomes administration. From this perspective, project risk management is the management of uncertainty once enough information has been acquired to assign meaningful values for probabilities and impacts of occurrence. Projects can be most usefully classified in terms of the levels of uncertainty that they face in terms of their mission. Different levels of uncertainty will have profound implications for the most effective way of managing the project (De Meyer, Loch, & Pich, l 2002) as illustrated in exhibit 3, which is a development of their work.
  

  Exhibit 3 Project Mission Uncertainty and Project Management Approaches

• The most appropriate project management style will change during the life-cycle of the project as the nature of information processing changes as more information is acquired. The matrix swing (Morris, 1994) in project organization structure is the organizational manifestation of this phenomenon, but it is also implicit in exhibit 3 as unknown unknowns become known unknowns and then known knowns with the dynamic reduction of uncertainty through time. The development of A Guide to the Project Management Body of Knowledge (PMBOK^® Guide) and PRINCE 2 has given the impression that there is one best way to manage a project, and just as the contingency approach to organization challenged the movement for the professionalization of general management in the 1950s (Woodward, 1980, chapter 12), this contingency approach to project management has important implications for the foundations of the PMBOK^® Guide.
• The issue of the relationships between resource bases is pervasive in all aspects of project management. Indeed, the project management team coordinating the project can only get tasks executed through contracts – be they internal or external to their organization. Arguably, a deep understanding of the project governance framework developed above is a vital competence for any project manager.
• Much of the recent work attempting to move beyond the systems analysis paradigm has stressed the importance of articulating an alternative and distinctive conceptual framework – for instance, Söderlund's (2004) most recent articulation of the Scandinavian school; Thomas (2000) on sense-making; and Winter & Checkland (2003) on soft systems analysis. Within the perspective articulated here, these alternatives are most fruitfully seen as antithesis to the systems analysis paradigm's thesis – we still need to continue the work in search of a synthesis of the tools and techniques of the systems analysis paradigm within a more comprehensive theory of project organization. In Kuhnian terms, the new paradigm that emerges from a successful scientific revolution must be able to explain better the phenomena addressed by the old, rather than present an alternative perspective.
• The inherently temporary nature of the project organization – while a touchstone of the systems analysis paradigm of project management research – presents novel challenges for mainstream organization theory and the theory of the firm. Some of the issues here have begun to be addressed by, for instance, Bryman, Bresnen, Beardsworth, Ford, & Keil, (1987); deFillippi & Arthur (1998); Lundin & Söderholm (1995) and Reve (1990), but much more work remains to be done here.
• The information processing approach to project management theory is articulated at the most generic level of project organization. There is considerable merit in identifying sub-processes within which contingencies vary both within and between projects. One attempt to do this is provided by Winch (2000), with a detailed application to construction projects in Winch (2002). The five generic information processes which articulate complementary facets of overall project management process are:

  ▪ defining the project mission –captures all the issues around investment appraisal, stakeholder management, the politics of internal client decision-making, and co-learning between client and contractor as client desire is tempered by technical possibility and regulatory constraint.

  ▪ mobilizing the resource-base – here the principal/agent problem of selecting and motivating competent suppliers thorough the project chain (Winch, 2001) is addressed at the level of principal contracts, sub-contracts, and the employment relationship.

  ▪ riding the project life-cycle – is the heart of the execution phase of the project where the traditional tools are mobilized alongside newer ones, such as critical chain and probabilistic risk assessment.

  ▪ leading the project coalition – is the process by which the project mission is infused (Barnard, 1938) throughout the project coalition through the appropriate choice of project organization design and the creation of a project culture.

  ▪ maintaining the resource base – is the often overlooked process of ensuring the capabilities of project-based firms to define and deliver, renewed through the full development of human resource potential and innovative new technology development. Professional development and competencies are an important part of this process.

Issues for Further Research

If this sketch for a theory of project organization as an information processing system can be accepted as a starting point for further work, then it would be useful to indicate some of the areas most in need of attention:

• The relationship between the providers of human resources on the supply side and the providers of capital on the demand side need much greater elaboration. There is a variety of terminology used to define the demand-side actor depending on the nation and sector under discussion – common terms include “client,” “employer.” and “owner.” These terms typically mask complex organizations with important differences of interest between promoters advocating the project and those having to finance it, such as bankers, shareholders, and taxpayers. There are two aspects to this problem. The first, internal to the demand side and typically formulated as the alignment of project objectives with business objectives, has received remarkably little research attention (Morris, 2002). Yet the most recent research shows how project promoters’ incentives are frequently poorly aligned with project financiers’ incentives (see below). The second is the struggle over margin (Porter, 1985) between the client side and the supply side where clients want to reduce their costs and contractors their profits. These are both aspects of the problem of the alignment of incentives within the project coalition, and we have little understanding of how to align all the incentives dyads in a project coalition so that they are pulling in the same direction. The growing importance of project finance techniques in providing the capital for asset delivery is making this issue of even greater importance.
• Traditionally, project management and systems engineering have been seen as parallel - even competing – paradigms (Johnson, 1997). Yet recent research on systems integration on complex projects (Hughes & Hughes, 2000; Prencipe et al., 2003) has shown that the potential for them to be considered as complementary processes in the creation of complex assets.
• The execution bias of the systems analysis paradigm has meant that the issues around the front-end definition of the project mission have tended to be under-emphasised. There are two main aspects to this problem. The first is helping the beneficial operator of the asset created by the project to articulate the operational requirements for the asset. These processes are known variously as briefing, requirements capture, whole life-costing, and the like. They are essentially processes of co-design as client desire and technical possibility are articulated and re-articulated to develop a business case in collaboration with the internal stakeholders. The second is then negotiating that business case with external stakeholders such as regulators and local losers - see Winch (2004) for an analytic framework.
• The research into project management competencies has taken the PMBOK® Guide as a reference point (Morris, 2001), but if the analysis of the limitations of the systems dynamics paradigm has weight, then these can only provide a very limited basis for PM competencies, and a much broader orientation is required based on more mainstream management skills. Moreover, the implementation of systems analysis-derived approaches in organizations can appear as authoritarian rather than supportive of professional development (Hodgson, 2002).
• A fundamental tenet of the systems analysis paradigm is control against schedule and budget objectives, yet there is growing evidence that in many cases these objectives are merely bait to attract investment capital. Project promoters tend to give themselves the benefit of the doubt -- to put it benignly -- when setting budgets and schedules. This is well documented in the work of Flyvbjerg and his colleagues for European mega projects (2003), and by Lovallo & Kahneman (2003) for the US corporate sector.
• If it is accepted that the fundamental challenge in project management is decision-making under uncertainty, then we need to know much more about how project managers make decisions - interestingly, the decision analysis aspects of Cleland and King's book do not seem to have been taken forward as enthusiastically as the more organizational systems analysis aspects. While there has been a considerable amount of research on the cognitive aspects of decision-making in a project context – see Vick (2002) for a valuable review – most of it assumes that at least the probabilities are known. We have little idea how project managers actually make the decisions that move the project through the life cycle when the deterministic suite of tools can offer little assistance.

Concluding Thoughts

This keynote address has attempted to move forward the debate on project management theory by:

• identifying the importance of the systems analysis paradigm for normal science in project management research;
• identifying some of the symptoms of paradigm crisis in project management research;
• reviewing a parallel literature which was called project-orientated organization theory;
• identified some important conceptual advances applicable to the analysis of project organizations derived principally from the work of Galbraith and Williamson;
• sketching what a new theory of project management might look like in terms of project organizations as information processing systems.

This can only be the start of the development of a project management theory to synthesize the contributions of the systems analysis paradigm and its critics. Perhaps such an endeavor in the ever-more-complex context of project management is futile. I suggest that it is not, and that even if that goal is not attained, the journey will have been worthwhile.