Detecting collusion: An application of the synthetic control method

Abstract

The Spanish antitrust authorities charged car makers with information sharing and collusion in the Spanish market in the period 2006 - 2013. Using data on car prices for models sold in a set of European countries, in the period 1993 - 2011, we measure how explicit collusion affected car prices in Spain in that period using the synthetic control method. By following a demand-driven approach, that is, based on the fact that the elasticity of demand is heterogeneous across segments, we focus the analysis on the most representative car segments within the Spanish and European market, selecting a repre- sentative car model for each of them. Our results indicate that sharing information caused higher prices in A and B segments, highlighting the great power of explicit collusion. In particular, we show how in oligopolies with a large number of initial members and in cartels that increase its size over time, such an agreement can be maintained, even in periods of economic crisis. Focusing on the segment with the largest market share, the C-segment, there is no evidence that sharing information had an effect of increasing final prices, suggesting that higher competition within the segment may succeed in nullifying the harmful effects of explicit collusion on social welfare.