Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This study addresses the large bias in chained price indices that persists even at lower frequencies. The bias arises from intertemporal substitution caused by consumer hoarding, and is problematic for purchase‐based data. In order to resolve this issue, we propose a method for calculating changes in inventories and consumption using retailer scanner data. We construct a partial equilibrium model to estimate inventories and consumption and show that the model accurately predicts the sign and size of the bias. We also demonstrate that the bias is smaller for consumption‐based data and propose a particular type of price index that eliminates intertemporal substitution bias.