Do integrated population, health and environment programs work? Gathering evidence using a quasi-experimental design.

Eckhard Kleinau, Analysis, Information Management & Communications Activity (AIM)
Odile Randriamananjara, U.S. Agency for International Development (USAID)
Fred Rosensweig, Training Resources Group, Inc

This article proposes the use of quasi-experimental or, if possible, randomized controlled designs to evaluate ongoing and future population, health and environment programs. Using the example of a population, health and environment program in Madagascar, the authors show how a quasi-experimental evaluation design can be built into the program design and implemented as the program matures. Results show that the integrated PHE program has achieved measurable impact over a three-year period. Twenty-nine out of 44 key indicators resulted in higher outcomes in integration than in non-integration communities. For only two indicators non-integration communities showed better results, although these could have occurred by chance alone. Fertility rates are high along Madagascar’s forest corridors, which are largely cut off from essential services such as health care and education. Contraceptive Prevalence Rate (CPR) reached 17 percent in integrated communities in 2004 compared to 8 percent in communities without integration.