This discussion paper led to a publication in 'The RAND Journal of Economics', Volume 44, Issue 3, pages 391–424, Fall 2013.
It studies the incentives to merge in a Bertrand competition model where firms sell differentiatedproducts and consumers search for satisfactory deals. In the pre-merger symmetricequilibrium, the probability that a firm is the next one to be visited by a consumer is equal acrossfirms not yet visited. However, in the short-run after a merger, because insiders raise their pricesmore than what the outsiders do, consumers start searching for good deals at the non-mergingstores. Only when they do not find any product satisfactory enough, they continue searching atthe merging stores. When search costs are sufficiently large, consumer traffic from the non-merging firms to the merged ones is so small that mergers become unprofitable. This new merger paradox,which is more likely the higher the number of non-merging firms, can be overcome in the mediumtolong-run if the merging firms choose to stock their shelves with all the products of the constituent firms, which generates sizable search economies. Such demand-side economies can conferthe merging firms a prominent position in the marketplace, in which case their price may even belower than the price of the outsiders. In that case, consumers visit first the merged entity andthe firms outside the merger lose out. Search cost economies may render a merger beneficial forconsumers and so overall welfare may increase.