Selective Consumption Tax Policy and Income-Expenditure Elasticities of Obesity-Linked Consumption Goods

Objective: To identify how consumption of 12 goods – alcohol, cigarettes, fast food, items sold at vending machines, purchases of food away from home, cookies, cakes, chips, candy, donuts, bacon, and carbonated soft drinks – varies across the income distribution by calculating their income-expenditure elasticites. Design and Methods: Data on 20,040 households from 2009-2012 from the Bureau of Labor Statistics’ Consumer Expenditure Survey were used. The data were analyzed using ordinary least squares regressions and Cragg’s double hurdle model which integrates (i) a binary model to determine the decision to consume and (ii) a truncated normal model to estimate the effects for conditional (y > 0) consumption. Results: Income had the greatest effect on expenditures for alcohol (0.314), food away from home (0.295), and fast food (0.284). A one percentage-point increase in income (approximately $428 at the mean) translated into a 0.314 percentage-point increase in spending on alcoholic beverages (approximately $1 annually at the mean). Income had the smallest influence on tobacco expenditures (0.007) and donut expenditures (-0.009). Conclusion: Percentage of a household’s discretionary budget spent on the studied goods falls substantially as income gets larger. Policies targeting the consumption of such goods will disproportionately impact lower-income households