Can Changes in Sexual Behaviors among High School Students Explain the Decline in Teen Pregnancy Rates in the 1990s?

John S. Santelli, U.S. Centers for Disease Control and Prevention (CDC)
Joyce C. Abma, National Center for Health Statistics, Centers for Disease Control and Prevention
Stephanie Ventura, National Center for Health Statistics, Centers for Disease Control and Prevention
Laura Lindberg, Urban Institute

Although teen pregnancy rates declined in the 1990s, no single national data set can explain these changes. We used behavior data from Youth Risk Behavior Survey and method-specific, contraceptive failure rates from the National Survey of Family Growth in attempting to understand pregnancy rates for 1991-1997. This method found an excellent fit to data for black teens, a partial fit for Hispanics, and a poor fit for whites. Sensitivity analyses considering recent sexual activity, behavioral data from 14-16 year old, 1999 behavioral data, or first birth rates to 15-17 year olds did not improve understanding. Limitations to this method include pregnancies to non-students, insufficient power given small changes in behaviors, possible temporal improvements in CFR, changing frequency of intercourse, and possible temporal change in reporting bias. Using several national data sources can compensate for individual limitations and provide a better explanation of adolescent pregnancy trends than each data source alone.