Relative Prices and Quality Substitution in the Chinese Cigarette Market: A Discrete Choice Analysis Using ITC China Data

Relative Prices and Quality Substitution in the Chinese Cigarette Market:

A Discrete Choice Analysis Using ITC China Data

Justin S. White¹; Jing Li²; Teh-wei Hu²; Geoffrey T. Fong³; Yuan Jiang⁴;

 

¹ Stanford University, USA; ² University of California at Berkeley, USA; ³ University of Waterloo, Canada; ⁴ Center for Disease Control and Prevention, China

Abstract

This study explores consumer choice among quality grades of cigarettes in China.  We test a theoretical prediction that consumers will substitute toward high-quality brands as the relative price of low-quality brands increases, introducing a new discrete choice methodology into the existing literature.  We derive a mixed multinomial logit model that allows for random taste variation and random utility maximization of brand choice.  We apply the model to three waves of longitudinal micro-data on smokers from six Chinese capitals.  Using within-city variation in cigarette prices across quality grades or tiers, we identify the impact of cigarette prices on brand choice, controlling for person-specific attributes.  We further identify the socio-demographic sub-groups that are most likely to engage in within-product substitution.  We find that relative price increases induce a large share of smokers to trade up toward greater quality as predicted, but also an equally substantial share of smokers to trade down toward cheaper cigarettes.  Low-income and less educated smokers are more likely to violate the theoretical prediction.  We hypothesize that an income effect and mental accounting best explain our findings.  We also describe the implications of our findings for tobacco taxation policy in China.

 

Keywords: discrete choice analysis, mixed logit, cigarette prices, excise taxes, quality, China

JEL: H26, I12, I18,