Evidence on search costs under hyperinflation in Brazil: the effect of Plano Real

Plano Real put an end to hyperinflation in 1994 and significantly altered price-setting behavior in Brazil. This paper investigates the impact of Plano Real on search frictions. I estimate a nonsequential search model for homogeneous goods to structurally retrieve consumers' search costs. The dataset comprises 11,673 store-level price quotes collected from 1993 to 1995 by FIPE to calculate the Consumer Price Index (CPI) in the city of São Paulo. I choose 15 brands to analyze: 7 food items, 4 industrial goods, and 4 services. To quantify the extension of search costs, I focus only on geographically isolated markets, defined as all stores that sell a certain brand within a radius of 6 km. The empirical strategy consists of using Plano Real as a structural breakpoint in the data. I estimate the model splitting the data into before (January 1993 to June 1994) and after (August 1994 to December 1995) the plan, and I find evidence on first-order stochastic dominance of the search-cost ibution of the former into the latter; that is, search costs are higher during hyperinflation. The majority of consumers search only once or twice before buying an item, but this share is marginally higher during hyperinflation (84% vs 79%). In addition, after Plano Real, a larger share of consumers is willing to quote prices in all stores before committing to a purchase. I also document evidence of the effect of the plan on shrinking price-cost margins. When searching is less costly, stores lose market power.