International Aid to Tanzania – with some comparisons from Ghana and Uganda

This paper has three principal objectives. First, to review the level of Official Development Assistance (ODA) to Tanzania over the last two to three decades, and to place this into an economic context. This review includes some comparisons with the experience of Ghana and Uganda. Second, to discuss three major issues for the Tanzanian aid: the position of ODA as budget support, corruption, and alignment with the principles of the Paris Declaration on Aid Effectiveness. Third, to review the literature on the Tanzanian aid experience, including a range of official evaluation reports produced by the Tanzanian government and by the donor community. The conclusions, broadly, are that ODA has been at a sustained high level for most of the period reviewed, funding a significant amount of government development expenditure, and that economic growth has been strong, with poverty reduction ‘flat-lining’ in Tanzania but being significant in Ghana and Uganda. Experience with budget support in Tanzania has been mixed, corruption continues as a major concern, and improvements to public finance management have been difficult to achieve. In this context governance adjustments come slowly, requiring patience on the part of both recipient governments and the ODA donor community.

Choosing on Influence

Interaction, the act of mutual influence between two or more individuals, is an essential part of daily life and economic decisions. Yet, micro-foundations of interaction are unexplored. This paper presents a first attempt to this purpose. We study a decision procedure for interacting agents. According to our model, interaction occurs since individuals seek influence for those issues that they cannot solve on their own. Following a choice-theoretic approach, we provide simple properties that aid to detect interacting individuals. In this case, revealed preference analysis not only grants the underlying preferences but also the influence acquired. Our baseline model is based on two interacting individuals, though we extend the analysis to multi-individual environments.