Budget Support - Rationale For Budget Support

Rationale For Budget Support

The underlying rationale for budget support is variously seen as:

a) Researchers at the Overseas Development Institute (ODI) describe how highly fragmented, ad hoc, small projects were not delivering results and cumulatively may have been undermining aid agencies' own objectives. There is some evidence that the effect of ad hoc, small projects is not systematic, but there is a wide range of individual evaluations of project as well as some more meta-review data such as Wapenhans report (1992) and ‘Assessing Aid’ (1998). Thus the ODI highlights the following benefits:

• High transactions costs from the multiplicity of different reporting and accounting requirements, including tied aid;
• Inefficient spending dictated by donor priorities and procurement arrangements;
• Undermining of state systems through the special staffing arrangements and parallel structures;
• Corrosion of democratic accountability as mechanisms are designed to satisfy donor rather than domestic constituencies;
• Hard to sustain positive impact beyond the short term, with high level of reliance on donor funding which undermined sustainability;
• Corruption, fraud and rent-seeking were also a feature of the management of projects and not overcome by their independence from government control.

b) that building recipient government capacity and accountability to their own populations for service delivery, as the only sustainable way of reducing poverty in the long term. Evidence is scarce, since it is hard to tell what level and type of effect budget support is or even could have on these long term and highly political processes. The findings of the largest review to date are over all positive. For more information, see the 8-country evaluation of GBS (1994-2004), 4-page summary: http://www.oecd.org/dataoecd/16/63/37421292.pdf

In practice the other types of aid instruments are still running concomitantly which means the potential expressed in the underlying rationale is unlikely to be fully tested.