Commute duration and health: Empirical evidence from Brazil

There have been many empirical studies associating commuting time and health outcomes in the last few decades. Their general conclusion is that commuting and health are negatively related. The validity of their findings, however, is questionable, given their lack of good identification strategies to correctly account for omitted variables. In this paper, we analyze this relationship using a large and unique nationally representative sample of Brazilian individuals, coupled with the use of propensity score matching techniques, and the application of an exhaustive set of standard falsification tests and sensitivity analysis that may prevent one from claiming a causal link between the two variables. Our results indicate that individuals with more than one hour of commuting appear to have statistically higher probability, ranging from 1.9 to 4.6 percentage points, of reporting bad health status when compared to a person whose commuting time is less than one hour.