The impact of tobacco control policies on disparities in children's secondhand smoke exposure: a comparison of methods.

To examine the impact of cigarette excise taxes and smoke-free legislation on tobacco use among households with school-age children and adolescents as well as disparities in children's secondhand smoke exposure. We compare the results from models using causal inference techniques to those from cross-sectional models. We linked families of 6-17-year-olds from the 2003 (N = 67,607) and 2007 (N = 62,768) contacts of the National Survey of Children's Health with state-level cigarette excise taxes and smoke-free legislation total score (0 [none]-32 [very strong]) in 2001 and 2005. Parents reported whether anyone in the household used tobacco products. In adjusted causal inference models every $1.00 increase in cigarette excise tax between 2001 and 2005 was associated with a 4 percentage point decrease in household tobacco use between 2003 and 2007 (p = 0.008); however, there was no effect of smoke-free legislation on household tobacco use. Significant interactions revealed that cigarette tax increases were only associated with reductions in household tobacco use for parents of white children and, separately, lower income households. In contrast, in adjusted cross-sectional models, a higher smoke-free legislation total score was associated with a lower prevalence of household tobacco use. Stronger cigarette excise taxes decrease tobacco use among households with school-age children and adolescents, but smoke-free legislation at the state level does not change parental smoking. Since cross-sectional models cannot assess the direction of causality, evaluations should employ causal inference methods to help inform policy decisions to reduce disparities in adult smoking and, ultimately, protect children from secondhand smoke.