Abstract
We provide an ex-post analysis of the 2001 merger between the two largest brewers on the Swedish beer market. Difference-in-difference estimates suggest low price effects of the merger. This is well matched by a merger simulation, using a random coefficient logit model, which predicts price increases of only 0.4 percent. Knowledge of the retailers markup rules allow us to discard retailer behavior as an explanation for the pricing patterns. We further establish that without the divestitures required by the competition authorities, the price increase would have more than doubled to 1 percent (even though still low in absolute terms).