Social Network Effects in Contraceptive Behavior Among Adolescents Mir M. Ali, PhD,* Aliaksandr Amialchuk, PhD,* Debra S. Dwyer, PhD† ABSTRACT: Objective: To quantify empirically the role of peer social networks in contraceptive behavior among adolescents. Method: Using longitudinal data from a nationally representative sample of adolescents, the authors use a multivariate structural model with school-level fixed effects to account for the problems of contextual effects, correlated effects, and peer selection to reduce the potential impact of biases from the estimates of peer influence. The peer group measures are drawn not only from the nominations of close friends but also from classmates. Contraception use among the peer groups was constructed using the peers’ own reports of their contraceptive behavior. Results: Controlling for parental characteristics and other demo- graphic variables, the authors find that a 10% increase in the proportion of classmates who use contraception increases the likelihood of individual contraception use by approximately 5%. They also find evidence that the influence of close friends diminishes after accounting for unobserved environmental confounders. Conclu- sion: The findings of this study support the findings in the literature that peer effects are important determi- nants of contraception use even after controlling for potential biases in the data. Effective policy aimed at increasing contraception use among adolescents should consider these peer effects. (J Dev Behav Pediatr 32:000 –000, 2011) Index terms: adolescent contraception use, peer measurements, peer influence. The majority of pregnancies among adolescents in the United States are unintended. 1 The use of contraception has been identified as a primary factor in reducing the number of unintended pregnancies. 2,3 However, adoles- cent contraception remains a complex issue for health care providers and policy makers because sexual activity often precedes the ability to make responsible decisions related to sex. 4 Although the use of contraception among adolescents has been increasing, 5 only 28% of sexually active adolescents use any method of contra- ception, while 62% of all age groups use contraceptives. 6 Research suggests that effective use of contraceptives reduces the medical, socioeconomic, and health costs associated with adverse outcomes from unprotected sex- ual intercourse. 7–9 Determinants of adolescent contraceptive behavior include factors like demographic characteristics (e.g., age, ethnicity, and income), parental characteristics (e.g., parental education and relationship with parents), and other individual characteristics (e.g., attitude toward contraception and pregnancy, age at sexual debut, and number of sexual partners). Another proposed determi- nant of contraception use is the receipt of advice about contraceptive methods from members of one’s social network, such as family and friends. 2,10 Considerable research has been devoted to determining the impor- tance of social networks in influencing adolescents’ risky health behaviors and outcomes. 11–16 In the case of ado- lescent contraception use, the role of social networks or peer effects is unclear, 17 although there is some evi- dence to suggest that behavioral choices by adolescents are partly determined by how acceptable their peers believe the behavior to be. 18 For example, individuals’ self-reported perception of the amount of contraception use among members of their social networks has been identified as an important predictor of contraceptive behavior, 2,10 and friends and family members are listed as the primary source of contraceptive knowledge. 19 Regardless of the mechanism through which social in- fluences affect adolescent contraception use, from a policy perspective, whether the social network effect operates and, if it does, its size are of interest because peer effects may serve to amplify the effects of interven- tions. 11–13 Therefore, it is important to understand the From the *Department of Economics, University of Toledo, Toledo, OH; †School of Health, Technology & Management and Department of Economics, Stony Brook University, Stony Brook, NY. Received May, 2011; accepted July, 2011. This research uses data from Add Health, a program project designed by J. Richard Udry, Peter S. Bearman and Kathleen Mullan Harris, and funded by a grant P01-HD31921 from the National Institute of Child Health and Human Development, with cooperative funding from 17 other agencies. Special ac- knowledgment is due to Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design. Persons interested in obtaining data files from Add Health should contact Add Health, Carolina Population Center, 123 W. Franklin Street, Chapel Hill, NC 27516-2524 (addhealth@unc.edu). The views expressed in this study are those of the authors and do not necessarily reflect the views of the Food and Drug Administration. Disclosure: The authors declare no conflict of interest. Address for reprints: Mir M. Ali, PhD, Office of Regulations, Policy and Social Science, Food and Drug Administration, College Park, MD 20740; e-mail: mir.ali@fda.hhs.gov. Copyright © 2011 Lippincott Williams & Wilkins Original Article Vol. 32, No. 8, October 2011 www.jdbp.org | 1