ABSTRACT
The results reported in earlier chapters and summarised in the last chapter suggest that GG does not exert a consistent or substantively important role in the allocation of aid in the 1990s. The results are robust with respect to a number of changes. They do not change dramatically if:
The time-invariant indicators of the extent of corruption and respect for the rule of law from the World Bank data set are replaced by their timevarying counterparts from the international-country risk guide (ICRG) data set. The latter were not used in the main analysis as they are available for fewer countries than the World Bank variables.
