Published online by Cambridge University Press: 27 September 2011
Contraceptive prevalence rates and estimates of continuation rates are derived from unique longitudinal data on post-partum behaviour collected in the Cebu region of the Philippines. Continuation rates vary by base-line and time-varying socioeconomic characteristics for certain contraceptive methods and for using no contraception. Calendar rhythm users have a much lower relative continuation rate than has been found in cross-sectional samples. For women who use a contraceptive method, breast-feeding does not appear to influence contraceptive continuation, but for those who use no methods, breast-feeding appears to substitute for other forms of contraception. Profiles by socioeconomic characteristics of those who continue to use each type of contraceptive method and who use no method could be used to direct family planning programmes towards population sub-groups. The results suggest that for women wishing to postpone a pregnancy, a family planning programme aimed at the young and less educated could effectively increase contraceptive use and continuation.