Unveiling Demand Function Dynamics: A Scalable, Cross-Market Estimation Using Point-of-Sale Data

This study examines the dynamics of demand functions for a wide range of consumer products in Japan over the past three decades. Using highly granular point-of-sale (POS) data and machine learning techniques, we estimate key demand function parameters, namely price elasticity and demand curvature, from which we also derive markups for each market and period. We find that while the aggregate-level median price elasticity and markups remained relatively stable over the long term, with substantial cross-product heterogeneity underlying this stability, recent years have seen a modest decrease in the absolute value of median price elasticity and a corresponding increase in markups. Furthermore, our analysis shows that demand curvature increased until the mid-2010s and subsequently declined. Importantly, our panel analysis reveals a significant relationship between the estimated demand function parameters and key factors, including labor force participation and market concentration. In particular, we identify the rise in female labor force participation as a key driver of the recent decline in both absolute price elasticity and demand curvature. These findings suggest that these shifts in socio-economic factors, such as increased labor participation, might have altered consumer behavior, leading to diminished price sensitivity.