This study had several limitations. Smoking and alcohol consumption were measured retrospectively via self-report, potentially limiting the accuracy of our findings. In addition, while participants may have moved out of states between Waves I and II of the NESARC, data on participant residence at Wave II were not available. However, moving from one state to another is not expected to be systematically related to cigarette taxation, and this limitation would likely create additional random error and result in an underestimation of the true impact of cigarette taxation on reductions in drinking behaviors. Cohort attrition could bias our findings if participants lost to follow-up were different in their smoking and drinking behaviors. Further, increases in excise cigarette taxes take effect in states at different points throughout the calendar year (and can occur multiple times), and states varied in the length of time between cigarette tax increases and Wave II interviews. This likely biased the estimates of the effects of increases in excise cigarette effects downward, making it more difficult to detect a result (e.g., cigarette tax increases occurring just prior to
