Time series analyses reveal differential impact of the Alcohol Warning Label by drinking level

Abstract

This paper uses time series analysis to examine the impact of the Alcoholic Beverage Warning Label on the attitudes and drinking behavior of 3,572 African American pregnant women. The results indicate that knowledge of the Warning Label increased 3 months after the implementation of the law. Drinking during pregnancy declined after a 7 month lag only among non-risk drinkers, although birthweight showed no trends. This analysis illustrates the importance of controlling for timing, history and behavior of subgroups when studying the impact of a natural experiment.