The proposed research will be one of the first in-depth investigations of the effects of an extreme form of neighborhood violence, specifically homicide exposure, on adolescent pregnancy in the United States. The analysis will examine how homicide exposure affects the risk of pregnancy and will further identify underlying mechanisms including changes in aspirations for and expectations about the future, perceived behavioral control, relationship dynamics, and sexual behavior and contraceptive use. Understanding the relationship between homicide exposure and pregnancy is a critical step toward understanding the effects of neighborhood context more broadly, and is especially important to consider during late adolescence, when unintended pregnancy rates are highest. At present, no research examines the specific relationship between homicide exposure and adolescent pregnancy. Moreover, the few studies that do consider the relationship between neighborhood violence, broadly construed, and pregnancy have been limited to cross-sectional data sources and data sources collected at wide intervals with imprecise measures of violence exposure. These data limitations may result in the underestimation of the effects of violence and have prevented scholars from investigating potential psychological, relational, and behavioral mechanisms. In contrast, new data from the Relationship Dynamics and Social Life (RDSL) study consists of weekly data on the pregnancy status and sexual and reproductive behaviors of a racially and socioeconomically diverse sample of late adolescent women residing in a city with one of the highest murder rates in the United States. These unique features of the RDSL, combined with administrative records on homicides, allows the proposed study to address previously unanswered questions regarding whether and how adolescent women?s pregnancy risk changes in response to their homicide exposure.
The specific aims of this project will be to: (1) assess how young women?s aspirations, expectations, and perceived behavioral control evolve alongside their exposure to homicide (2) examine how young women?s relationship dynamics, sexual activity, and contraceptive use, evolve with homicide exposure and (3) investigate how the risk of adolescent pregnancy changes following homicide exposure. To achieve these aims, the proposed study will integrate geo-referenced administrative data on the date and location of homicides with time-varying information from RDSL panel data on young women?s aspirations for and expectations about future education, employment, and family formation; relationship dynamics; sexual activity and contraceptive use; and pregnancy status. Fixed effects regressions will be used to assess how each of these outcomes differs before and after women are exposed to homicides occurring within a 2.5 year period. By utilizing a fixed effects strategy that holds constant all immutable characteristics of women, the study will isolate the effects of homicide exposure from potential confounders such as poverty and race.
The proposed study will provide new insights into whether and how exposure to neighborhood violence, homicides specifically, affects the risk of adolescent pregnancy. The findings have the potential to advance research and theory on reproductive behavior and social stratification by isolating the effects of homicide exposure from characteristics of women and their circumstances that could otherwise confound the estimated relationship, facilitating a more detailed understanding of a potentially important determinant of pregnancy during a period of the life course marked by high unintended pregnancy rates. The findings will additionally inform outreach strategies and public health campaigns that are designed to target late adolescent and young adult populations at particularly high risk of early pregnancy and relationship volatility.
