The validity of survey data for measuring family planning service impact in a rural locality of northern Ghana

This paper examines a decade of prospective panel survey data recording proximate fertility determinants and demographic surveillance data in a rural locality of northern Ghana where a quasi-experiment tested the impact of four contraceptive service provision strategies on fertility.
Although treatments explain variance in observed fertility, analyses also reveal that fertility implied by proximate determinant responses exceeds observed fertility in areas with convenient doorstep family planning services. Analysis of longitudinal surveillance data suggests that community exposure to these doorstep services is associated with reproductive change. Results are consistent with the hypothesis that denial is most pronounced where observed fertility decline was greatest. The gap between fertility expected from proximate determinants and fertility observed from surveillance increases over time. This finding challenges the widely held assumption that survey appraisal of contraceptive use represents a valid endpoint for studies of the reproductive impact of community-based family planning services.