We compare first-price auctions to an exchange process that we term "multilateral negotiations." In multilateral negotiations, a buyer solicits price offers for a homogeneous product from sellers with privately known costs, and then plays the sellers off one another to obtain additional price concessions. Using the experimental method, we find that with four sellers, transaction prices are statistically indistinguishable in the two institutions, but with two sellers, prices are higher in multilateral negotiations than in first-price auctions. The institutions are equally efficient with two sellers, but multilateral negotiations are slightly more efficient with four sellers.

No abstract

We use the experimental method to compare second-price auctions with ÔverifiableÕ multilateral negotiations in which the sole buyer can credibly reveal to sellers the best price offer he currently holds. Despite the two institutionsÕ seeming equivalence, we find that prices are lower in verifiable multilateral negotiations than in second-price auctions. The difference occurs because low-cost sellers in negotiations often submit initial offers below the second-lowest cost. We also compare the two institutions to previously studied first-price auctions and multilateral negotiations with non-verifiable offers. Second-price auctions yield the highest prices, followed in order by verifiable negotiations, non-verifiable negotiations and first-price auctions.One of the more interesting facets of voluntary exchange is how changes in the strategic environment lead to differences in transaction outcomes. For example, the pioneering experimental work in Chamberlin (1948) and Smith (1962) illustrates that changing the nature of the information available to participants can dramatically influence the price and efficiency of an exchange process. With this insight in mind, in this article we use the experimental method to examine and compare four exchange mechanisms in a procurement setting in which a buyer faces several sellers that have privately known production costs.The first two institutions are the first-price auction and the second-price auction. These well-known auction formats and their outcome-equivalent variants, the Dutch auction and the English auction, are used extensively to allocate products as varied as flowers, art, produce, fish, government securities, and mineral rights. Several governments have recently used auctions to allocate such valuable resources as radio spectra, electric power, and pollution rights. The second two institutions are variants of the multilateral negotiations introduced in Thomas and Wilson (2002). In this common exchange mechanism, a buyer solicits price offers from multiple sellers and then he elicits more favourable offers by playing the sellers off one another until he accepts one of the offers or breaks off the negotiations. Among other settings, multilateral negotiations are

We experimentally compare first‐price auctions and multilateral negotiations after introducing horizontal product differentiation into a standard procurement setting. Both institutions yield identical surplus for the buyer, a difference from prior findings with homogeneous products that results from differentiation's influence on sellers' pricing behaviour. The data are consistent with this finding being driven by concessions from low‐cost sellers in response to differentiation reducing their likelihood of being the buyer's surplus‐maximizing trading partner. Further analysis shows that introducing product differentiation increases the intensity of price competition among sellers, which contrasts with the conventional wisdom that product differentiation softens competition.

This paper evaluates both efficiency increasing and efficiency decreasing mergers in a procurement setting in which firms differ in the likelihood that they have high production costs. Profitable efficiency increasing mergers often decrease the expected price, but profitable efficiency decreasing mergers always increase it. For a particular pair of firms, there may be no profitable mergers. If there are, then the most profitable merger may decrease efficiency. Consequently, merging firms that are able to choose their postmerger level of efficiency may profitably elect to decrease it.