ABSTRACT
The case of this chapter is the Global Environment Facility (GEF) or more precisely the GEF evaluation offi ce. The GEF is an international funding mechanism aimed at protecting the global environment. The GEF is an example of a so-called Global and Regional Partnership Program (GRPP). These programs exhibit the following characteristics (IEG, 2007: xvi):
Multi-level interventions require evaluative evidence for the same reasons as single intervention evaluative evidence. First of all, organizations and especially GRPPs are accountable for their actions not only at the level of discrete interventions but also, and especially so giving the particular institutional and management characteristics of GRPPs, at portfolio and program level. The questions “are we doing things right” and even more so “are we doing the right things” require systematic evidence at program level supported by evidence at the level of discrete intervention activities in order to be answered. Second, learning about effectiveness across interventions is of crucial interest to GRPP stakeholders. In general, single intervention evaluations are limited in terms of their knowledge generation potential (Cook and Campbell, 1979). More specifi cally, there are limitations in terms of the internal validity (i.e., is there a causal relationship between changes in goal variables and intervention activities) as well as the external validity (i.e., to what extent is this relationship valid for a larger group of interventions or contexts). Consequently, there is a need to clearly qualify the fi ndings generated by these evaluations, and correspondingly, to be cautious about the potential to generate knowledge about the performance and effectiveness of particular types of interventions. To different degrees the available multi-level evaluation approaches are helpful in reducing these threats to validity in order to generate robust evidence on what works and why across interventions.
