Between online and offline markets: A structural estimation of consumer demand

Abstract While the literature on comparing the online and offline markets from the supply-side perspective is extensive, fewer researches have focused on the demand-side. Using a large-scale point-of-sales (POS) dataset on consumer packaged goods, we estimate a structural demand model and conducted counterfactuals of the price change or product stockouts. Our key findings are: 1) the own-price elasticity of demand is higher offline (i.e., –1.74 on average) than online (i.e., –0.87 on average). 2) The cross-price elasticity is also higher offline (i.e., 0.17 on average) than online (i.e., 0.04 on average). 3) When a product is out-of-stock, a larger portion of the loss of the market share is recovered more in the common products in the offline markets than that in the online markets. Specifically, 6.55% of the loss of the market share goes to common products offline on average, while 6.05% of the loss of the market share goes to the common product online on average. Our further analyses show that, when considering substitute products, the impact of price increases is surprising. The retailer may not be worse off by increasing prices because the gains in the revenue from the substitute products can make up more than the losses. The manufacturers may lose revenue by increasing prices or product stockouts. However, consumers are hurt more in the offline markets following the price increases of the common products because the offline market is more concentrated on the common products.