Limited resources are often spent on poorly-selected projects that fail to achieve benefits concomitant with their cost. The result can be under-used assets and poor service delivery at a higher cost than necessary. These systematic problems result from:
• Poor planning and coordination-good sector and cross-sector planning and coordination is needed to ensure that the 'best' projects-those that represent value for money, enable integrated regional development, and provide customers with the services they desire-are consistently selected. Without sound plans, responsible agencies will not have the full view of potential projects that could be implemented and will not know the sequence in which to implement the projects to achieve the best value for money, and cross-sector coordination will be weak. Box 1.4: Mumbai Water-Example of Poor Planning in Infrastructure provides an example of how weak infrastructure planning can mean projects fail to achieve value for money. McKinsey [#179] report on infrastructure investment-identifying $57 trillion globally required between 2013 and 2030-notes that scaling up best practice could save an average of $1 trillion a year in infrastructure costs during that period
• Flawed analysis-the analysis underpinning project selection is often flawed, so projects that appeared to be cost-benefit justified turn out not to be so in practice. Benefits are often over-estimated, resulting in projects that are larger or more complex than is justified by demand for services, while costs are often under-estimated. The United Kingdom Government's Green Book on project assessment [#238, pages 29-30] acknowledges this as a systematic problem and highlights the need to correct for 'optimism bias' in project analysis. UK Treasury supplementary guidance on optimism bias [#239] presented evidence on the extent of optimism bias dating from the early 2000s-although more recent evidence from the UK notes that public procurement practices have since improved-see for example [#242, #243] and [#243]. A global series of studies of large transport projects by Flyvbjerg [#101, #102, #103] found that costs are systematically under-estimated, and benefits often over-estimated:
- A study of 258 transport projects found that actual costs were on average 28 percent higher than planned costs-and 65 percent higher on average for projects outside Europe and North America
- A study of 25 rail projects found traffic was heavily over-estimated, at over twice actual traffic, on average. The accuracy of traffic forecasts for 183 road projects was also found to be highly variable, but without a tendency to over-estimate.
• Politics or personal gain interfering with the project selection process; increasing costs, or diverting funds to less beneficial projects. An IMF analysis of corruption in public investment in infrastructure found corruption tends to create a bias towards capital spending projects, and increase their size and complexity-reducing the productivity of that investment [#225].
These factors often feed into each other. For example, weak analysis or poor planning can enable badly-chosen projects to be pushed through for political or personal gain, as described in the World Bank's sourcebook on deterring corruption in the water sector [#279, Chapter 6]. Flyvbjerg's studies also emphasize, with examples, that costs and benefits can be deliberately misrepresented, to push through projects for political or organizational reasons [#101].
Box 1.4: Mumbai Water-Example of Poor Planning in Infrastructure The experience of the Municipal Corporation of Greater Mumbai provides an example of weak planning in the water sector. The Corporation was looking for ways to improve the efficiency of its operations. Mumbai is short of water, with supply rationed to around four to six hours a day in most parts of the city. Corporation planners were working on new schemes to transport water from hundreds of kilometers outside the city. Consultants engaged through the World Bank analyzed the cost of achieving a 24 hour water supply in one ward (K-East) entirely with new supply, and compared this with the cost of achieving 24 hour water supply through improving the distribution system to reduce leakage and theft. The consultants estimated that the cost of distribution improvements would be one sixth or less of the cost of bulk supply increments, for the same level of service improvements. The size of the discrepancy suggests that the Municipal Corporations' planning had been biased toward large projects. |