Inequality-Reducing Effects of Tobacco Tax Increase: Accounting for Behavioral Response of Low-, Middle-, and High-Income Households in Serbia

Abstract

While previous research has indicated that increasing tobacco excises is a crucial instrument for lowering tobacco demand, this policy has been criticized for its alleged regressive impact on the poor. However, this critique does not take into account the behavioral response, i.e., decrease in consumption that occurs after excises and prices increase. In this paper, we examine the effect of cigarettes' price increase on tobacco consumption, household expenditures, and tax burdens in three income groups and provide empirical arguments on the regressivity/progressivity effects of tobacco tax increase. Estimated elasticities indicate that all groups decrease their cigarettes demand with increasing prices, with demand decrease stronger for low- than for middle- and high-income households. Results further suggest that increasing tobacco excises (1) decreases tobacco expenditure of low-income households, which increases their productive consumption, such as on food, clothes, etc., and (2) redistributes the tobacco tax burden from low- to high-income households. Therefore, excise increase policies do not have an adverse effect on the position of the low-income households; on the contrary, they lower their cigarettes expenditure and their tax burden, while lower cigarettes consumption has an additional, positive effect on their health, which attenuates future inequalities.

Keywords: fiscal revenues; inequality; price elasticity; tax progressivity; tobacco excises.

Figure 1

Smoking prevalence (a) and conditional (demand) intensity (b) trends by income group. Source: Authors’ calculation based on HBS data. Notes: Smoking prevalence is defined as the share of the households with positive tobacco consumption, while smoking intensity represents the number of cigarette packs a household with positive expenditures on cigarettes smoked per month. Cigarette prices are defined as municipality/year average cigarettes’ unit values (ratio between total expenditure and quantity) and are expressed in real terms (2006 = 100). Income groups are constructed based on total household consumption (a proxy for income) per capita.