High-resolution behavioral economic analysis of cigarette demand to inform tax policy


Correspondence to: James MacKillop, Department of Psychology, University of Georgia, 100 Hooper St., Athens, GA 30602, USA. E-mail: jmackill@uga.edu



Novel methods in behavioral economics permit the systematic assessment of the relationship between cigarette consumption and price. Towards informing tax policy, the goals of this study were to conduct a high-resolution analysis of cigarette demand in a large sample of adult smokers and to use the data to estimate the effects of tax increases in 10 US States.


In-person descriptive survey assessment.


Academic departments at three universities.


Adult daily smokers (i.e. more than five cigarettes/day; 18+ years old; ≥8th grade education); n = 1056.


Estimated cigarette demand, demographics, expired carbon monoxide.


The cigarette demand curve exhibited highly variable levels of price sensitivity, especially in the form of ‘left-digit effects’ (i.e. very high price sensitivity as pack prices transitioned from one whole number to the next; e.g. $5.80–6/pack). A $1 tax increase in the 10 states was projected to reduce the economic burden of smoking by an average of $530.6 million (range: $93.6–976.5 million) and increase gross tax revenue by an average of 162% (range: 114–247%).


Tobacco price sensitivity is non-linear across the demand curve and in particular for pack-level left-digit price transitions. Tax increases in US states with similar price and tax rates to the sample are projected to result in substantial decreases in smoking-related costs and substantial increases in tax revenues.