This study contributes to the debate on aid-effectiveness by analysing a prominent, under-researched policy question: do bilateral grants and loans distort trade in different ways?

Focusing on this ‘gap’ in policy analysis, the study first reviews the theoretical arguments and after finding little evidence embarks on a fresh empirical investigation into the possible links between levels of concessionality and trade distorting effects of aid.

Phase One of the study (Clay and Turner, 2007) found no direct evidence and few a priori reasons to expect the choice of aid instrument, grant or loan, would have differential trade distorting implications. Rather, as aid policy analysts and the exiguous literature suggest, it is whether or not aid is tied that is likely to determine the differential trade distortion effects.

Formal tying is self-evidently a distortion, except where the funder is the most competitive source. Some de facto tying is intentional. Some distortions may occur that are unperceived, possibly because they are unintended outcomes of donor-recipient relationships and information asymmetries.

In Phase Two three follow-up research activities were undertaken: first, a global econometric analysis of the relationships between levels of bilateral aid flows and exports from DAC member countries (Massa and te Velde, 2009) and second, an exploratory country study of Ghana looking at use of instruments, aid modalities, tying practices and links to possible intended and unintended distorting outcomes (Clay and others, 2009a). As members of the ODI have also been acting as core group for the OECD-DAC-PDE thematic study on the untying of aid, it was agreed that the Ghana study would also act as the pilot for a set of country recipient studies undertaken in Phase II of the OECD study. Accordingly the third activity, the sectoral and project-level methodology developed in the Ghana study has been used in these recipient country studies. Accordingly the methodology for sectoral level analysis of aid disbursement, sourcing and cost-effectiveness developed during the Ghana study is being used for a set of investigations focused primarily on the water and sanitation sector in these studies. The commitment of donors to cooperate with the OECD study link has considerably facilitated this investigation, and its report in December 2009 will provide further evidence on this previously relatively under-researched subject.

The consistent finding of the whole study is that there is no significant difference as between grant and loan ODA in terms of trading distorting effects. In view of a widespread view that bilateral loans are in practice more distorting than grant aid, that is an interesting result. Nevertheless it is also stressed that this finding should be regarded as provisional. The global econometric analysis was constrained by serious data limitations and the country study by time and resource. But encouragingly the findings of the literature review in Phase One now complemented by the Phase Two global econometric results and the Ghana country study are highly consistent and a justification for the eclectic approach is adopted. Total aid flows and sourcing, as reflecting formal and informal de facto tying practices are the main aid-related sources of trade distortion. The choice of instrument (grant or loan) is per se not an influence on trade.

The provisional findings point to the need for further research, especially at a country and sub-sectoral level, of the relationships between use of instruments (extended if possible to include FDI and export credits), tying practices, sourcing of goods and services and aid effectiveness.