Abstract: This paper explores how innovations in advertising technology reshape consumers’ brand preferences – propensity to purchase certain brands over others despite similar prices – and contribute to changes in the market structures of grocery products. I first document that the average market concentration and markup for consumer-packaged goods decreased by 10% and 5% between 2010 and 2016. Next, I show that the aggregate cost function of the advertising industry shifted significantly during the same period. To structurally estimate model parameters, I construct a multi-product, multi-sector heterogeneous firm model with endogenous advertising and entry/exit decisions. I find that advertisements have greater influence on consumer brand preferences in categories where brands are less substitutable. On the macroeconomic level, my model suggests that changes in advertising cost structures are the main driving forces behind redistribution of firm market shares. The counterfactual analysis shows that if the advertising cost structure had not changed, the average markup and market concentration would have both increased between 2010 and 2016.
Abstract: Based on item level micro-data collected from around 35,000 U.S. supermarkets from 2006 to 2016, I compare different approaches to construct location-specific price indexes from scanner data sets and propose a method that generates unbiased, free of “chain drift” indexes while staying computationally feasible. I show that the resulting indexes are consistent with official statistics released by the U.S. Bureau of Labor Statistics and discuss ways to treat missing data in the process. My results show significant regional variations in inflation rates across U.S. locations from 2006 to 2016: cities have lower inflation rates than rural areas. This finding potentially raises new questions on the effect of central banks’ monetary policies on regional inflation rates.