Bundling Time and Goods: Implications for Hours Dispersion

We document the large dispersion in hours worked in the cross-section. We account for this fact using a model in which households combine market inputs and time to produce a set of nonmarket activities. To estimate the model, we create a novel data set that pairs market expenditures and time use at the activity level using data from the Consumer Expenditure Survey and the American Time Use Survey, respectively. The estimated model can account for a large fraction of the dispersion of hours worked in the data. The substitutability between market inputs and time within an activity and across a sizable number of activities is key to our results. We show that models that lack these features can only generate one third of the observed hours dispersion.