THE CHARACTERISTICS OF CONSTRUCTION PROJECTS

How then do construction projects fit into the overall picture of project organisation? There has been some debate amongst specialists on construction as to whether it is the firm or the project which should be considered as the unit of analysis. However, it is clear from reading Williamson that the most helpful perspective is to take that of the firm (Winch 1989); in particular, analysis should start with the client, and then move backwards down the value chain. In terms of the transaction cost analysis reviewed above, considered from the point of view of the client, construction can be considered to be characterised by low asset specificity, low transaction frequency, and high uncertainty. The market for the vast majority of construction services is highly competitive; indeed, it is considered by some to be too competitive, and clients may deliberately limit tender lists to ensure that poor competitors do not drive out good ones. Clients, even "experienced" ones, procure relatively few buildings each year, and even when they do procure more, this is likely to be a burst of activity which is not sustained year on year. However, construction projects can be considered to suffer from high levels of uncertainty (Winch 1989). In terms of dynamism, they usually suffer from a lack of task information. The site specific nature of construction means that each project has a high number of unique features which need to be resolved through the project life cycle. At the early stages, factors such as the interaction of the project with existing facilities, regulated through urban planning procedures, and the dynamic nature of the client's requirements, can lead to considerable amounts of uncertainty which can only be reduced by negotiation, reworking and abortive work. For larger infrastructure projects, political considerations, particularly budgetary ones, can generate considerable dynamism. This high level of dynamism at the early stages means that it is difficult to smooth programmes of projects so that productive resources can be easily transferred from one to another. Once on site, dynamism generated by the unknowabilities of natural conditions - both geological and climatological - generate further uncertainties. These are reinforced by the temporary nature of the project organisation - many of the actors will not have worked together before, and this creates an organisational dynamism which can easily generate conflict rather than cooperation within the project coalition. As the organisation is temporary, the returns to building long term relationships are inherently limited, and so atmosphere considerations are muted in their effects. This dynamism is compounded by complexity. At the upper end, construction projects are amongst the most complex of production undertakings. While it remains true that construction is largely an assembly industry, that assembly work has to be undertaken on sites which, for the reasons described above, are inherently dynamic. The components to be assembled vary considerably in character and the skills that they require for their assembly. These components range from low tolerance simple components such as bricks, to high tolerance complex components such as advanced engineering services installations. Particular problems derive from the interactions between such components, or more correctly, component systems. There is also a trend through time for this complexity to increase due to technological change and more demanding client requirements. The combination of high dynamism with medium to high complexity on most construction projects generates high levels of uncertainty. These high levels of uncertainty would be enough to drive construction procurement towards hierarchy if it were not for the low asset specificity and, especially, very low transaction frequency. The overwhelming preference by clients is for market governance of construction procurement. This preference is shared by the principal actors within the project organisation. The coalition form, as defined above, tends to be favoured. There is, however, another peculiarity of construction which derives from the time dimension of projects. The market transaction is relational rather than junctural, and it is a capacity to produce rather than a product that is purchased. This relational character of transactions pervades the entire value chain. Its most important result for our purposes here is that once contracts have been exchanged, considerable asset specificities are generated - the transformation is truly fundamental. The costs of replacing a project actor part-way through their service delivery are normally punitive. It is for this reason that trilateral governance of transactions both between project coalition members themselves, and between the client and the project coalition are favoured. This point is illustrated by the prevalence of complex contract forms such as the Joint Contracts Tribunal and Institution of Civil Engineers series in the UK, Code des Marchés Publics in France, the Verdingungsordnung für Bauleistungen in Germany, and FIDEC internationally. Stinchcombe in his analysis of the North Sea oil industry has argued that under such conditions market transaction governance takes on hierarchical forms. He argues (1985 chap 2) that the reasons mainly arise from: difficulties in the prediction of the client's desire for contractor performance; uncertainty about the costs of performance, resulting in the client retaining the right to alter performances through the project life cycle; inability of the client to clearly measure the relative performance of the separate parties. Due to these problems, market contracts are written in such a way that they achieve hierarchical effects by: specifying authority systems to facilitate change providing incentive systems to motivate the project actors using administered pricing systems to handle uncertainties such as bills of quantities providing conflict resolution procedures providing standardised operating procedures It is these features that characterise what has been defined above as trilateral transaction governance. Lorange and Roos assert that no firm will enter a strategic alliance when it can operate alone, due to the greater costs associated with working collaboratively with independent organisations. These are a combination of loss of autonomy, and hence power, and the increased transactions costs inherent in inter-organisational relations. Only where the gains outweigh the costs will collaborative forms be preferred. By extension it may be argued that unbalanced power relations will be preferred by powerful project actors in such collaborations. In other words, coalitions and quasi-firms will be favoured over consortia and joint ventures by those actors who are nodal in the value chain. These nodal actors in construction are the principal designer (architect or civil engineer) and the principal contractor (general contractor or construction manager). Empirically, these power imbalanced forms are generally favoured, particularly by principal contractors, and may range from a strong form such as domestic subcontract in the British system to a weak form such as groupement conjoint in the French system. Atmosphere considerations mean that principal designers tend more towards consortia. In systems such as the UK, such consortia members are often in separate contract with the client, although this is less common in France. However, there is an important set of reasons why at least some of the actors move more firmly towards consortia, or even joint ventures. One reason is that discussed above - the size of project relative to the size of firm. A second is to do with atmosphere considerations - the site-specific nature of production means that the idea of a local firm has a particular resonance. In some systems, such considerations are embodied in local labour requirements by public clients, or technology transfer agreements in international contracting. Such local firms may not have the necessary technology for undertaking the project, while firms that have the technology may not meet the local labour and other requirements. This is resolved by local and non-local firms entering into formal consortia such as groupement mandataire in the French system, and the Arbeitsgemeinschaft in the German. A third reason is the problem of asset specificity. As Clark and Ball (1991) argue, where particular assets are critical to the project principal contractors prefer to enter into a formal consortium with the holders of those assets. This binds that asset holder more tightly to the principal contractor thereby reducing the latter's risk. In other words the consortium is a deliberate means of avoiding opportunistic behaviour that might be generated in the more power imbalanced collaborative forms. Such consortia are sometimes also encouraged by clients with the objective of obtaining smoother running projects. For instance British health authorities tend to favour consortia between general and services contractors because engineering services form such a high proportion of total project cost. A peculiar variety of this is the British nominated subcontract where actors with real equity of power are put into formal imbalance. As might be predicted, this causes considerable frictions.