Working Paper Series no. 746: Everyday Regular Prices

The ability of monetary policy to affect the dynamics of the economy in response to business cycle shocks depends on the adjustment process of prices. Microeconomic evidence documents higher price flexibility than what is suggested by aggregate inflation: Prices change often, but a large share of price adjustments are actually temporary sales, that is, price transitions which are later reversed. Regular prices are instead much more rigid. It is an open debate in the literature whether sales are relevant from a macroeconomic perspective. If they are time-dependent `comeback prices' that follow sticky plans that are basically unresponsive to macroeconomic shocks, they contribute little to price flexibility relevant for the transmission of monetary policy. Inference on the firms' pricing behaviour, and their responsiveness to monetary policy and to changes in demand, is therefore hard to extract from micro price data in general, since it is unclear whether sales event respond to demand conditions and to aggregate macroeconomic variables - and whether they do so in the same degree as regular prices.

Using a novel dataset from a large supermarket retailer in a European country that never engages in temporary sales, we establish that prices are actually as sticky as regular prices. Circumventing the debate on whether sales have to be included or excluded from price adjustments, we find evidence consistent with state-dependence price setting in a multiproduct firm. In particular, our data exhibit responsiveness of prices to changes to aggregate demand shifts, a more than trivial share of very small price changes, and synchronization of price changes across items especially within the same product category. Price rigidity and the extent of state-dependence are heterogeneous across items. In particular, we find that pricing of top sales items (and even more of private label ones) is more flexible and state-dependent. Indeed, the extent to which prices react to an exogenous shift in demand is more than twice for 1% top sales items with respect to that for the overall sample and about four times for private label top 1% sales items that are private labels (arguably characterized by particularly high profit margin). This is consistent with price setting in a multiproduct firm that may rationally choose to be inattentive to information that is costly to acquire, absorb, or process for some items more than for others ones. Indeed, a multiproduct price setter may more often revise and change prices of items that are more important for the firm, minimizing in this way the loss incurred when failing to adjust.