Working Paper: NBER ID: w18180
Authors: Henry Saffer; Dhaval Dave; Michael Grossman
Abstract: The behavioral economic model presented in this paper argues that the effect of advertising and price differ by past consumption levels. The model predicts that advertising is more effective in reducing consumption at high past consumption levels but less effective at low past consumption levels. Conversely, the model predicts that higher prices are effective in reducing consumption at low past consumption levels but less effective at high past consumption levels. Unlike the models used in most prior studies, this model predicts that the effects of policy on average consumption and on the upper end of the distribution are different. \n\nBoth FMM and Quantile models were estimated. The results from these regressions show that heavy drinkers are more responsive to advertising and less responsive to price than are moderate drinkers. The empirical evidence also supports the assumption that education is a proxy for self-regulation. The key conclusions are that restrictions on advertising are targeted at heavy drinkers and are an underutilized alcohol control policy. Higher excise taxes on alcohol reduce consumption by moderate drinkers and are of less importance in reducing heavy consumption.
Keywords: Alcohol Consumption; Behavioral Economics; Advertising; Price Elasticity
JEL Codes: D03; I18
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| advertising (M37) | alcohol consumption (L66) |
| price (D41) | alcohol consumption (L66) |
| education (I29) | alcohol consumption (L66) |