Post-Cartel Behavior: ssessing the effects of antitrustpolicy on Brazilian fuel market

Papers assessing the antitrust effect on cartel cases usually take the form of a quantifying approach, measuring the impact on prices with methods like before-and-after dummyregressions, difference-in-difference, or synthetic controls designs. However, these approaches have some downsides (notably, the requirement of establishing an exogenous date or breakthrough event, based on assumptions that may not be accurate). To over-come this weakness, we applied Structural Break Analysis (Bai and Perron Test) and Markov Switching Regressions to four cases in the Brazilian fuel market (Brasilia, Belo Horizonte, São Luís and Londrina) to analyze the effectiveness of competition policies. As a comparative test between MSR and Bai Perron procedures, our paper shows that the former was more sensible to transitions between regimes, without missing breaks,and exhibited precise results. From the point of view of the antitrust policy evaluation, our findings indicate a low capacity of the antitrust authorities to extinguish price-fixing practices in targeted markets.