The sensitivity of measures of unintended pregnancy to question timing: Evidence from Malawi

Unintended pregnancy is an important demographic measure used to justify and assess the success of family planning programs. There are three commonly used techniques to measure unwanted and mistimed pregnancies, all of which rely on the simplifying assumption that fertility preferences are relatively stable. Accumulated research from a variety of contexts, however, indicates that key measures of fertility preference, including ideal family size, the desire for another child, and the desired timing of next pregnancy, change frequently over time, particularly as women’s life circumstances evolve. This paper uses seven waves of panel data, each spaced four months apart, from young women in southern Malawi to examine the prevalence of change in fertility preferences and the implications for measures of unwanted and mistimed pregnancy. Specifically, we estimate the prevalence of unwanted and unintended pregnancy using seven different measures, three of which allow for preferences to change over time. The paper concludes with a discussion of the differences in estimates and the possible implications for measures of unintendedness in other contexts.