This thesis is a case study of the pro-poor part of the Information Communication Technology (ICT) policy in South Africa. It utilizes government policy documents and interviews with bureaucrats to understand the pro-poor intentions of the ICT policy and the tools and structures chosen to achieve them. It found that the intentions were to provide ICT services to poor communities in telecentres, households and schools. To achieve this three implementation channels were established: (i) the telecentre channel, which consisted of a Universal Service and Access Agency in charge of paying subsidies from a Universal Service And Access Fund to establish telecentres; (ii) the market channel, which consisted of an independent regulator in charge of issuing and enforcing universal service obligations to ICT companies to provide access to ICT in households and schools; and (iii) and the education channel, which consisted of a structure within the provincial Department of Education aiming to provide ICT to schools.
This thesis then uses data collected from fieldwork in three poor communities in KwaZuluNatal to assess and explain the results of the three channels on the ground. Despite political backing and significant funding the policy was only partially successful. The telecentre and market channels were largely unsuccessful, while the education channel was relatively successful. It is argued that the variation in results can, to a large extent, be explained by differences in the tools and structures that were utilized by the channels. The two less successful channels were both centralized at the national level and relied on third parties – which they found difficult to control – for implementation at the local level. The education channel, on the other hand, relied on direct delivery through a decentralized bureaucracy. This thesis provides tentative but empirically founded support to the argument that direct tools, which rely on hierarchic bureaucracies, tend to be easier to implement and more effective than indirect tools relying on third parties. This might be particularly true in context of South Africa, where state institutions lack the capacity to manage indirect tools that rely on complex implementation structures.