Firms that set prices on both sides change who wins and loses in supply chains
This paper studies how firms trade with each other inside production networks and how that affects prices and the division of gains. The authors build a theoretical model in which every firm can influence prices both when it buys inputs and when it sells outputs. In that setting, the final price paid by consumers and the shares of profit and consumer benefit are determined by the network of trades, the technologies firms use, and consumer preferences.
To capture firm-to-firm trade, the paper represents the production network as a collection of double auctions. In this setup, firms interact as buyers and sellers, and their ability to influence prices on both sides is called multilateral market power. The model makes the size and the split of the economic “surplus” — the total gains from production and trade — an outcome of the technology, where firms sit in the network, and how consumers value the final goods.
A common simplification in past work is to assume firms are price-takers when buying inputs. That is called unilateral market power: a firm can affect its selling price but not the prices it pays for inputs. The new model shows that this simplification has systematic effects. When analysts assume price-taking on input markets they tend to understate the final price paid by consumers and overstate how much surplus flows upstream to input suppliers.
These differences matter for policy questions. For example, predictions about how a merger changes welfare depend on who actually holds market power in the network. If the standard price-taking assumption is used, it can give a misleading picture of who gains or loses from a merger and by how much.
The paper is a theoretical contribution and its conclusions follow from the model’s assumptions about auctions, technology, network structure, and preferences. That means the quantitative implications for real markets will depend on how well those assumptions match actual industries. Empirical work would be needed to test how large the effects are in practice and how they vary across sectors.