Selection effects in an analysis of contraceptive discontinuation in Morocco: an application of a multiprocess multilevel model

An important area of demographic research in developing countries has been the contribution of family planning programmes to increasing contraceptive prevalence and fertility decline. More recently, as contraceptive prevalence increases, attention has extended to other measures of programme success such as contraceptive continuation. Studies of programme impact on fertility and contraceptive prevalence typically involve examining the effects of community-level indicators of programme availability and accessibility. However, to examine the impact of a family planning programme on contraceptive discontinuation it is necessary to compare discontinuation rates across different types of facility. To do so, individual-level data on where couples obtain their contraceptive supplies is required. Such data were collected for the first time in Morocco in 1995 under the Demographic and Health Surveys (DHS) programme. The aim of this paper is to estimate the effect of source of contraceptive supply on pill discontinuation in Morocco using these DHS data