By: Sarah K. Cowan, Lawrence Wu, Susanna Makela*, Paula England
Published in: Perspectives on Sexual and Reproductive Health 48 (2016)
Abortion is a frequent medical procedure undergone by diverse women in the United States with profound demographic and political implications yet we do not know how many American women alive today have had abortions. We do not know this basic fact because women under-report their abortion histories in surveys. There are a number of well-established techniques to elicit more accurate survey responses to sensitive items. In this comment, we propose that lifetime prevalence estimates for abortion in the United States could be improved through use of these techniques, in particular the double list experiment. We report on a pilot double list experiment we conducted which shows promising results. We also provide unique formulae for determining the appropriate sample sizes needed to detect that the double list experiment has improved accuracy of estimates over those obtained from asking women directly whether they have had an abortion.