Scandinavian Working Papers in Economics

Working Papers,
University of Aarhus, Aarhus School of Business, Department of Economics


Philip DeCicca , Donald Kenkel () and Alan Mathios ()
Additional contact information
Philip DeCicca: Department of Policy Analysis & Management, Postal: Cornell University
Donald Kenkel: Department of Policy Analysis & Management, Postal: Cornell University
Alan Mathios: Department of Policy Analysis & Management, Postal: Cornell University

Abstract: This paper re-examines the empirical support for predictions that proposed cigarette tax or price increases will substantially reduce youth smoking. Part of the support for these predictions comes from evidence that higher taxes reduce aggregate tobacco sales and adult smoking rates. But taxes may have much different impacts on youth starting behavior than on adult quitting behavior. We use a panel microdata set, the National Education Longitudinal Survey of 1988 (NELS:88), that spans a period when many states increased taxes on cigarettes.We are able to study the impact of taxes and prices on smoking behavior during exactly the period in adolescence when most smokers start their habits. Cross-sectional models of 12th grade smoking based on the NELS:88 data yield estimated price elasticities ranging from -0.29 to -0.98,similar to previous studies. But when we exploit the longitudinal nature of the data our results suggest that cigarette taxes or prices are not important determinants of smoking initiation. We find weak or nonexistent tax and price effects in models of the onset of smoking between 8th and 12th grade, models of the onset into heavy smoking between 8th and 12th grade, and discrete time hazard models that include state fixed effects. Our estimates create doubt about the strength of the response of youth smoking to higher taxes or prices, and suggest that alternative policy approaches to preventing youth smoking deserve serious attention. We also provide a new perspective on the relationship between smoking and schooling. We find that students with better tests scores are less likely to smoke, and that eventual dropouts are already more likely to smoke in 8th grade. Possible explanations for these patterns include individual heterogeneity in: the rate of time preference; tastes for deviancy; parental investment in smoking prevention as an aspect of child quality; and optimal lifetime plans for health and education human capital investment.

Keywords: Cigarette smoking; taxation

JEL-codes: D11

46 pages, January 1, 2000

Full text files

00-3.pdf PDF-file 

Download statistics

Questions (including download problems) about the papers in this series should be directed to Helle Vinbaek Stenholt ()
Report other problems with accessing this service to Sune Karlsson ().

This page generated on 2018-01-23 23:30:43.