Estimating the overcharge of the liquefied petroleum gas cartel in Brazil's Federal District: evidence from before-and-after and difference-in-differences approaches

Abstract

Context: the liquefied petroleum gas (LPG) distribution and retail market in Brazil’s Federal District engaged in collusive practices that raised prices and triggered an investigation by the Administrative Council for Economic Defense (Cade), culminating in the cartel’s conviction in 2023. Objective: estimate the overcharge attributable to the LPG cartel and demonstrate the relevance of Cade’s intervention in restoring competition and consumer welfare. Method: combined application of the “Before and After” and Difference-in-Differences (DID) approaches, comparing prices before, during, and after the collusive conduct to isolate the cartel’s causal effect on prices. Conclusions: The estimates obtained from the “Before and After” methodology indicate overcharges ranging from 2.82% to 5.94%, while the Difference-in-Differences model shows values between 3.73% and 4.56%. Considering the most comprehensive specification, which includes time fixed effects and standard errors clustered by neighborhood, the estimated effect loses statistical significance, suggesting that part of the price variation may be explained by common spatial and temporal factors. Overall, the results indicate that, although there is no robust evidence of a price increase directly attributable to the cartel, CADE’s intervention was relevant in dismantling the coordinated conduct and preserving market competition, thereby promoting a more competitive and stable environment for price formation.