This study uses the 1993 National Demographic Survey(NDS) of the Philippines to examine sources of variation in contraceptive use-effectiveness. Focusing on pills, the analysis presented here indicates substantial performance differences among users supplied by different outlets. Evidence is presented that some of the observed differential is due to differences in clientele. However, once observable user characteristics are controlled, substantial variation by supply source in use effectiveness
remains. The paper concludes with an illustrative comparison of cost-effectiveness analyses on various sources of pill supply. Use of an outcome-based measure of effect is shown to generate different cost effectiveness conclusions than obtain from couple-years of protection, an input-based measure.
