Book Chapter
Financing Global and Regional Public Goods through ODA?

The present level of ODA falls short of the amount needed to finance the Millennium Development Goals (MDGs). The figure of additional $50 billion per year, roughly the present total of ODA spent by DAC donors, is often quoted (e.g. by the Zedillo Report); it results from the sum of the fight against communicable diseases ($ 7-10 billion), primary schooling ($10 billion), infant and maternal mortality ($12 billion) and halving world poverty ($20 billion). The scarcity of public resources raises the importance of investing in international public goods as the cost of lifting one person out of income poverty, for example through agricultural research and global trade expansion, is estimated to be much lower than the cost of the same impact through traditional aid to poor countries. This raises important issues for donor strategies, in particular principles of aid allocation, which this paper aims to address. First, should aid be partly earmarked towards international public goods? Second, what is the impact on aid to the poorest countries and on traditional aid projects if ODA is allocated towards deleting the underprovision of international public goods?Based on definitions for international, global and regional public goods, data on grant commitments from the OECD Creditor Reporting System will be attributed to these categories(roughly 15 per cent for global and 15 per cent for regional public goods) as well as to other, traditional, ODA (the remaining 70 per cent). The paper then presents a highly stylised model of public goods, adapted to the special donor-recipient relationship, to highlight the trade-off between ODA earmarked for the provision of international public goods and reform ownership in recipient countries through free use of resources. In the empirical analysis, the paper quantifies in a panel data analysis for the period 1997-2001 the offset coefficients for the impact of higher ODA commitments on both aid to the poorest countries and on traditional aid. While the hypothesis of extreme crowding-out is strongly rejected, the average offset coefficient between GPG-related ODA and traditional aid is also significantly higher than zero, namely 25 per cent. The paper also shows that there is no significant crowding-out of aid in poor countries caused by the provision of global public goods through ODA.