Measuring commuting in the American Time Use Survey

The journey between work and home plays an important role in daily time use, acting as both a fixed time cost of labor force participation and as a constraint on time for other activities. Data from the American Time Use Survey (ATUS) offer the opportunity to examine commuting behavior and its relationship to demographics, labor market characteristics, and the amount of time spent on other activities. Previous analyses have been complicated by the difficulties of obtaining commuting time measures from the ATUS. Travel information can be difficult to interpret in the ATUS, and many commuting trips are likely misclassified using stock measures of work-related travel. To address this shortcoming, I review the strategies of previous researchers to reclassify travel. After surveying possible methodologies, I focus on applying to the ATUS a methodology applied to the National Household Transportation Survey (NHTS). Detailed time information in the NHTS allows me to compare both aggregate commuting measures and the timing of commuting in the two surveys. I further extend the analysis to compare to journey-to-work information in another commonly used dataset, the American Community Survey. These comparisons and the methodology provided serve to enable and validate further analysis of commuting behavior using the ATUS, leveraging the advantages of this dataset.