Document Type

Working Paper

Publication Date



Since the 1960s the proportion of couples practicing contraception has risen rapidly, particularly in the developing world, and the mix of methods is now dominated by modern methods. Despite these trends, the incidence of unintended pregnancy remains high mainly because the number of children desired has declined. Worldwide there are almost as many unintended as intended pregnancies each year (not counting miscarriages, which are excluded in this analysis) and more than half of these unintended pregnancies end in abortion. This study examines the potential role of further increases in contraceptive prevalence and effectiveness in reducing abortion rates. The model used in this analysis links the abortion rate to its direct determinants, including couples’ reproductive preferences, the prevalence and effectiveness of contraceptive practice to implement these preferences, and the probability of an abortion to avoid unintended births when contraception fails or is not used. An assessment of the tradeoff between contraception and abortion yields estimates of the decline in the total abortion rate that would result from an illustrative increase of 10 percentage points in prevalence. This effect varies among societies, primarily because the tendency to use abortion after an unintended pregnancy varies. For example, in a population with an abortion probability of 0.5, a 10 percentage point increase in prevalence would avert approximately 0.45 abortions per woman, assuming contraception is 95 percent effective. If all unintended pregnancies were aborted, this effect would be three times larger. Eliminating all unintended pregnancies and subsequent abortions would require a rise in contraceptive prevalence to the level at which all fecund women who do not wish to become pregnant practice 100 percent effective contraception. A procedure is provided for estimating this “perfect” level of contraceptive prevalence.