The vast majority of state capital spending employs traditional procurement methods whereby the state internally manages or outsources design, development and project management usually as separate contracts. The most common method is an input specified asset and/or service procured by lowest price tender. There is a significant body of evidence that points to the failure of traditional procurement for complex construction and infrastructure projects. The larger and more complex the task, the greater the risk that projects will not be delivered efficiently or in a cost effective manner. This adversely affects the quantity and the quality of public service delivery. Recent evidence from Europe and Australia suggests that when governments use lowest price tender methods, around 70% of projects are late and a similar percentage are over budget.12 This does not necessarily mean that procurement by tender is the problem. Evidence suggests that the poor performance of traditional procurement is a consequence of government agency failure in the preparation of the business case, inter-agency friction, the separation of the design, construct and operations elements of the project and an input specification that is either incomplete at the time of tender or is subject to ongoing change during the early stages of the project (Flyvbjerg 2003; NAO 2003, 2004a). It is not unusual for a combination of an incomplete design and the lowest price tender for a project to end up the more costly form of procurement. This is evident from a long list of traditionally procured projects and Federation Square in Melbourne, the Scottish Parliament building in Edinburgh and the Opera House in Sydney are obvious examples. Less well known examples include the Southampton Oceanographic Laboratory, Guy's Hospital Stage III, the New British Library and Quarry House in the United Kingdom (NAO 2003, 2005).
Poor procurement performance involves large sums of public monies and is a form of public failure. Public failure occurs when governments fail to allocate resources efficiently or the social cost of a state intervention exceeds its benefit (Winston 2007; Regan 2008). Additional contributing factors include optimism bias (the overestimation of benefits and underestimation of costs), low levels of design and construction innovation and, little regard to lifecycle costing and the risks associated with the long-term management of complex assets such as hospitals, corrective service institutions, public utilities and telecommunications systems. Steps are being taken to improve traditional procurement and many of the improvements are flowing from lessons learnt with alternative procurement methods including alliance contracting, public private partnerships, outsourcing and the build own operate transfer arrangements. It is improvements to the science of public procurement that is the central issue here and not the benefits and disbenefits of specific procurement methods.13
________________________________________________________________________________________________________________________
12 NAO 2003. The 2002 Mott McDonald report found that traditionally procured standard building projects (those not requiring special design considerations) had taken between 1-4% -longer to complete than expected at business case stage, before contract award. Non-standard buildings (those involving special design considerations, for example, specialist hospitals, innovative prisons, high technology facilities other unique buildings or refurbishment projects) had taken between 2-39% longer. The 1999 report, Benchmarking the Government Client, found that construction programmes overran by an average of 13% compared to the tender stage (NAO 2003, p. 3, n. 2).
13 NAO 2005 Improving Public Services Through Better Construction; Egan 1998 Rethinking Construction; NAO 2001 Modernising Construction.