If it is revealed that a group of 8 soft drink manufacturers ( from an industry of 10 soft drink manufacturers) had engaged in a cartel between 2005-2010, how do I estimate the impact that their collusion had on prices ?.
I was thinking along the following lines: I try to study the trend in pricing of the non cartel group (the two other manufacturers) from say, 2000-2010. I observe the trend in pricing of the cartel group during the same time period. I then try to see if there are differences in pricing between the two groups during the cartel period (2005-2010)which cannot otherwise be described by the difference in prices during regular times(2000-2005). If there is a difference in prices, then I can show the effect that the cartel had on prices.
Now if I follow this reasoning, it’s based on the assumption that the prices of the non-colluders were not affected during the cartel period. Is this a reasonable assumption to make ?. If it is the case that the non colluders had a large share of the industry would they price independently of the cartel ?.