This study should be interpreted within the context of several limitations. The NESARC policy data can only be analyzed at the state level. While municipalities may enact local bans, we are not able to assess this degree of specificity as the NESARC data only provides state-level geographic data for respondents. Many states were covered partially by local ordinances prior to the implementation of statewide smoke-free bar and restaurant legislation, which likely biased the estimates of the effects of statewide legislation downward making it more difficult to detect a result. We did not have information on participant residence at Wave II and it is possible that some participants moved in or out of states that implemented smoke-free policies. The effect of this limitation is likely to be nonsystematic, adding random error to our results and thus leading to an underestimation of the true effect of smoke-free policies on AUD transitions. In addition to comprehensive statewide smoke-free bar and restaurant legislation, work place bans enacted as part of separate legislation may have also impacted AUD transitions. However, we found that workplace bans
