Budget support, conditionality and poverty

This paper examines the effectiveness of budget support aid as an anti-poverty instrument. We argue that a major determinant of this effectiveness is the element of trust – or `social capital´, as it may be seen – which builds up between representatives of the donor and recipient. Thus we model the conditionality processes attending budget support aid, not purely in the conventional way as a non-cooperative two-person game, but rather as a non-cooperative game which may mutate into a collaborative equilibrium if sufficient trust between the negotiating parties builds up. Whether or not this happens is, we argue, fundamental to the effectiveness of conditionality, and of budget support aid. This then requires us to enquire into the determinants of trust, which - we empirically demonstrate - derive from the experience of the negotiating parties with one another, from the incentives they are able to provide to trust one another and from the processes within which their negotiations are conducted. The model is tested against two samples: extensively against a broad sample of all African countries undergoing budget support operations and intensively against a narrow sample of Ethiopia, Uganda, Malawi and Zambia. The statistical analysis suggests that trust has in practice been achieved not only through a positive `social history´ but by the transmission of forward-looking `signals´ or `bona fides´ concerning fundamentals: high pro-poor expenditure, low military expenditure, and low corruption show a positive relationship with growing trust (measured in terms of freedom from programme interruptions). Where these signals are present, budget support aid is in general growing, and slippage on overt conditionality is in general forgiven; but there are exceptions to this trend, as our case-study analysis demonstrates . A proactive stance in defence of a pro-poor strategy is positive for trust, as are certain procedural reforms including the presence of an IMF resident mission and frequent face-to-face meetings between negotiators for donor and recipient. High trust generates stability of aid, and stability of aid, in conjunction with its level and its targeting, significantly influences growth and poverty outcomes.