Nearly half of all pregnancies worldwide are unintended (41%).1 Among young women, this number is even higher. By the age of 20, one in three women in the United States will experience a pregnancy,1 and over 80 percent of these will be unintended.2,3 Ensuring that a woman is able to make the right choice for herself about whether to carry an unintended pregnancy to term is fundamental to her health and wellbeing. A woman’s decision to continue or terminate an unintended pregnancy has ramifications that affect her health, her educational, professional and personal aspirations, as well as the health and wellbeing of her children. 4-7 However, methodological limitations in our ability to measure unintended pregnancy and abortion, and our ability to study their causes, limit the effectiveness of interventions designed to improve women’s health.
The first chapter of my dissertation introduces a novel methodological tool for the measurement of sensitive and stigmatized events: the list experiment. Validation studies suggest that the degree of underreporting of self-reported abortions is high. In countries where abortion is illegal, underreporting may be even greater. But without accurate estimates of the size of the population affected, effective policy and programs cannot be developed or targeted. This chapter describes results from a study of women of reproductive age in Liberia in 2013. To measure abortion prevalence, each woman was read two lists: A) a list of non-sensitive items, and B) a list of correlated non-sensitive items with abortion added. The sensitive item, abortion, was randomly added to either List A or List B for each respondent. The respondent reported a simple count of the options on each list that she had experienced, without indicating which options. Difference in means calculations between the average counts for each list were then averaged to provide an estimate of the population proportion that has had an abortion.
The second chapter of my dissertation extends the work of the first chapter. I implement two multivariable regression estimators with the list experiment data to understand how age and education vary with history of abortion. We find that education and abortion are inversely associated, after accounting for age. The hope is to encourage other epidemiologists to utilize newly developed tools for multivariable regression estimation with list experiment data.
The third and final chapter of the dissertation moves from measurement to analysis. The aim of this chapter is to introduce a causal inference framework to the family planning literature. I examine whether social support is causally linked to the incidence of undesired pregnancy among approximately 1,000 young women in Michigan. Using multivariable logistic regression, and an extension using standardization, I calculate relative and absolute estimates of the incidence of undesired pregnancy under two levels of social support.
As a body of work, my dissertation introduces a novel measurement tool to the field toward the goal of more accurate measurement – a first principle of epidemiology. It also offers a roadmap for how to approach family planning questions with a causal inference framework, to bring new rigor to the field, and improve our understandings of the complex determinants of unplanned pregnancy and abortion.