Furthermore, although there is a balance of IEs between lower-middle and upper-middle-income countries, less than a fifth of IFC IEs correspond to lowincome countries.
World Bank Group Impact Evaluation Methods Recent World Bank Group IEs have tended to use experimental designs more than quasi-experimental designs. Although the majority of completed World Bank IEs (77 percent) use quasi-experimental methods, 81 percent of ongoing evaluations use experimental methods.29 As can be seen in figure 2.7, the number of IEs using experimental methods has sharply increased since 2007. In particular, 83 percent of all IEs initiated between 2009 and 2010 use randomization, compared with 57 percent in 2005–06 and 20 percent in 2000-04. The higher rate of randomization among recent World Bank IEs is due to the explicit focus of DIME on prospective evaluations (either on the basis of random assignment or random phase-in). Similarly, the SIEF selection guidelines also prioritize randomized IEs.30 However, because the IE database
Figure 2.7
Impact Evaluations at the World Bank Using Experimental versus Quasi-Experimental Design by Initiation Year
60 Quasi-experimetal
Experimental
Number of World Bank IEs
50 40 30 20 10 0
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
IE initiation year Source: IEG. Note: Based on 381 of 396 World Bank IEs initiated between 2000 and 2010. For 15 ongoing IEs, the IE design was not specified in the DIME database. For three additional ongoing IEs, the initiation year could not be determined, so they are excluded from the count of 396 IEs referenced above. The information on IE design for completed IEs is drawn from IE reports; the source of information for design of ongoing World Bank IEs is the DIME database. IE = impact evaluation.
30
World Bank Group Impact Evaluations: Relevance and Effectiveness