I revisit Ngai and Tenreyro (2014)'s empirical analysis of seasonal match quality in American Housing Survey (AHS) data. Using 1999 data only, Ngai and Tenreyro show that homes purchased in the summer season are occupied longer and have fewer and less costly renovations soon after purchase, pointing to superior match quality for households who move house during the thicker summer market. However, applying the same methods to other years of the AHS substantially weakens these results. In addition, I document heaping in a key variable, the prior move month, and implement a multiple imputation correction. Ngai and Tenreyro's use of a coarsened measure of duration seems to largely overcome the biases that heaping introduces.
