Center for Family and Demographic Research
Distal Determinants of Race-Ethnic Variation in Unintended Fertility
Karen Benjamin Guzzo
Sarah Hayford (Arizona State University)
Jennifer Barber (University of Michigan)
Yasamin Kusunoki (University of Michigan)
Unintended fertility in the United States has remained stable since 1994, at half of all pregnancies and a third of all births, with young adults having the highest rates (Mosher, Jones, & Abma 2012). Continued high levels of unintended fertility in the U.S. are one of the “persistent empirical puzzles” of family demography (Bachrach et al. 2007: 2), especially since the reduction of unintended fertility has been a public health goal since the 1980s (U.S. DHHS 1980, 1990, 2000, 2010). An equally persistent problem is that unintended fertility is substantially higher among minorities than non-Hispanic whites (Finer & Zolna 2011), which may exacerbate health disparities given the negative association between unintended fertility and maternal and child health and well-being (Logan et al 2007). Although the proximate determinants of unintended fertility are clear (failure to use effective contraception and carrying unintended pregnancies to term), the underlying causes of these behaviors, and of race-ethnic-nativity differences, are not well understood. Extending Coale’s (1973) “ready, willing, and able” fertility model, this project proposes two key distal determinants – reproductive knowledge and fertility motivation – that may explain unintended fertility and race-ethnic-nativity differences. By using three complementary datasets (the Relationship Dynamics and Social Life study (“RDSL”; a population-based sample of women aged 18-19 residing in a Michigan county in 2008/09, followed weekly for 2.5 years), the National Longitudinal Study of Adolescent Health (“Add Health;” youths enrolled in grades 7-12 in 1995 and reinterviewed in 1996, 2001/02, and 2007/08), and the 2009 National Survey of Reproductive and Contraceptive Knowledge (“Fog Zone”; national sample of unmarried men and women aged 18-29), this project moves beyond prior descriptive work to identify precursors of risky sexual behavior and unintended fertility, hypothesizing that minorities have lower reproductive knowledge and motivation to prevent pregnancy, that disparities widen over time, and that these distal determinants are differentially associated with outcomes across groups. Using factor and latent class analyses (guided by integrative data analysis techniques), the project will create equivalent measures of the distal determinants across surveys and race-ethnicity-nativity in Aim 1. In Aim 2, this work will examine the stability of these constructs as young adults learn and gain new experiences that may affect their reproductive knowledge and fertility motivation using cross-lag regression models, leveraging the longitudinal nature of the RDSL and Add Health, and examine whether race-ethnic-nativity disparities widen during the transition to adulthood. In Aim 3, this work will establish the validity of the distal determinants by linking them to the behavioral determinants and fertility itself in the RDSL and Add Health using event history models, testing whether the determinants are differentially associated with outcomes across groups. The ultimate goal is to create reliable and valid measures to identify those most at risk for unintended fertility, which can then be applied to future research and interventions.