Working Paper: NBER ID: w23085
Authors: Benjamin B. Lockwood; Dmitry Taubinsky
Abstract: A common objection to “sin taxes”—corrective taxes on goods like cigarettes, alcohol, and sugary drinks, which are believed to be over-consumed—is that they fall disproportionately on low-income consumers. This paper studies the interaction between corrective and redistributive motives in a general optimal taxation framework. On the one hand, redistributive concerns amplify the corrective benefits of a sin tax when sin good consumption is concentrated on the poor, even when bias and demand elasticities are constant across incomes. On the other hand, a sin tax can generate regressivity costs, raising more revenue from the poor than from the rich. Sin tax regressivity can be offset by targeted transfers or income tax reforms if differences in sin good consumption are driven by income effects, but not if they are driven by preference heterogeneity, and not if the indirect incentives the sin tax generates for labor supply decisions are not salient. The price elasticity of demand determines the extent to which corrective benefits versus regressivity costs determine the size of the optimal tax. We implement our optimal tax formulas in a calibrated model of sugar-sweetened beverage consumption for a range of parameter values suggested by empirical work.
Keywords: sin taxes; optimal taxation; behavioral economics; income effects; preference heterogeneity
JEL Codes: H0; I18; I3; K32; K34
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
Cause | Effect |
---|---|
redistributive concerns (D63) | corrective benefits of sin taxes (H23) |
increased consumption of sin goods by poorer individuals (D12) | larger welfare gains from corrective taxation (H23) |
income effects (H31) | offset sin tax regressivity (H23) |
price elasticity of demand (D12) | optimal tax size (H21) |
low elasticity (D11) | regressivity costs dominate (H23) |
high elasticity (H30) | corrective benefits dominate (J32) |
underlying consumption profile (D12) | optimal tax size (H21) |
preference heterogeneity and income effects (D11) | optimal tax size (H21) |