Effects of Tobacco Abuse Liability-dependent Taxes in the ETM

In a within-subject design, cigarette smokers, stratified by age, will purchase tobacco products under three tax rate conditions based on abuse liability, that is taxes based on: (1) product category score (i.e. score ranging from 0-1 based on abuse liability effect sizes); (2) product category rank (i.e. position of a product relative to all others; e.g. 1st, 5th); and (3) product tiers (i.e. low-, medium-, high abuse liability), as well as a control. In the ETM, the effects of these interventions will be examined on cigarette demand and other products substitution.

The investigators will use a repeated-measures (within subject) analysis of variance (ANOVA) to test if there are differences between conditions (i.e., tax proposals) independently for each product. The models will include study design parameters for order effects to account for counterbalancing the tax conditions in the ETM. Each tax condition will be compared to the control. All distributional assumptions will be checked prior to analyses and the appropriate transformations will be employed, if needed. For each significant result from the omnibus test, the investigators will perform planned contrasts to test the proposed hypotheses previously described. Note that the omnibus test can result in a significant finding while the contrasts of interest may be non-significant. In this case, the investigators will only report the results from the planned contrasts.

Additional analyses might be conducted.