The aim of this project is to gain a better insight into the delivery of public services in the developing world. Public good provision is at the core of the infrastructure that supports an economy’s drive for growth. Water and sanitation foster health, rural roads and electrification provide the infrastructure necessary for small and medium enterprises to flourish, and drainage systems and dams directly increase the productivity of agriculture.
As a starting point, the study uses evaluations by the Nigerian Government of the fates of over 6000 projects and programs in the 2006 and 2007 government budgets. These evaluations showed us how roughly 40% of public projects are never implemented. Many of those projects that are completed are of questionable quality. On the other hand, this implies that 60% of projects are completed in a country infamous for corruption and inefficiency. Some organisations of government completed 100% of their projects to a level that independent experts believed to be of a satisfactory quality or higher.
The focus of our research was why some organisations were more effective than others in delivering these goods, and what determined which projects were efficiently completed. The projects under study range from water and sanitation infrastructure to primary health centres and HIV/AIDS programs. Each class of project was implemented by a variety of organisations. In this research, we concentrated on the degree to which organisations were centrally managed. Some organisations, such as the central ministries of a sector (such as the Ministry of Health), are the most centralised tiers of government. Others, such as regional hospitals, are the most decentralised. We found that more decentralised tiers of government were more effective at implementing the basic public projects studied. This corroborates many of the existing studies on decentralisation. However, we also believed that there was a need to take the research further. A simple comparison of average completion at the different tiers assumes that the projects being compared are the same. The early literature on decentralisation pointed out that there may be an innocent reason as to why this might not be so – different tiers of government may be more effective at implementing different types of project.
The International Growth Centre has funded a set of engineers to assess different aspects of the technical ‘complexities’ of every project under study. By using multiple (independent) engineers, (secretly) having them code a sample of projects twice, and by comparing with the assessments of a British engineer, we have tried to gain as accurate a data set as possible. This data provides a platform to investigate the ‘economics of complexity’. We found that more centralised tiers do in fact implement projects that are more complex, and that this provides an important explanation as to why centralised tiers perform worse. Implementing more complex projects is harder and projects naturally fail more of the time. Different tiers also specialised in different types of complexity. The concepts of complexity, measurement tools, and validity checks developed in this project can be used for a far wider range of analyses than those presented here.