Gains from Product Variety and the Local Business Cycle

Net product entry is procyclical, which amplifies fluctuations in consumer welfare over the cycle if consumers have love for variety. Using barcode-level data covering grocery expenditures in 26 major cities, I establish that differences in city-level product entry are largely uncorrelated with local economic conditions. I provide evidence that city-level changes in product variety over the business cycle are driven instead by multi-city retailers who introduce new products simultaneously in all cities in which they operate. This suggests that product introduction by multi-city retailers can propagate business cycle shocks. To quantify the impact of this mechanism, I develop a quantitative model of retailer product choice that relates the welfare gains from product entry in each city to demand growth in every other city. The model implies that the contribution of other cities' business cycle shocks to each city’s price level is proportional to the share of other cities in retailer revenue. Since the share of outside cities in retailer revenue is 63 percent in the average city, the impact of other cities' shocks on product entry is substantial. The presence of multi-city retailers makes net product entry more correlated across cities than they would be if retailers operated in only one city. In a counterfactual in which retailers do operate in only one city, the variation in gains from new product entry across cities would be 47 percent higher than in the baseline.