| Abstract: | Using state-level panel data for the USA spanning three decades, this research estimates the demand for cigarettes. The main contribution lies in studying the effects of cigarette advertising disaggregated across five qualitatively different groups. Results show cigarette demand to be near unit elastic, the income effects to be generally insignificant and border price effects and habit effects to be significant. Regarding advertising effects, aggregate cigarette advertising has a negative effect on smoking. Important differences across advertising media emerge when cigarette advertising is disaggregated. The effects of public entertainment and Internet cigarette advertising are stronger than those of other media. Anti-smoking messages accompanying print cigarette advertising seem relatively more effective. Implications for smoking control policy are discussed. |
