Is Relational Exchange Possible in the Absence of Reputations and Repeated Contact?

Experimental methods are used to study two central components of an incomplete contract: the ex post process of negotiation, which fixes contract terms and surplus shares in response to stochastic shocks, and the ex ante process of transaction-specific investment made in anticipation of future surplus shares. The 2 × 2 treatment design features either high or low discount rates in an alternating-offer bargaining setting, and transaction-specific investments that are either commonly observable or private information and thus unobservable. The analysis finds results consistent with the notion that despite random and anonymous matching of contracting parties, and the noncontractability of investments, parties index their surplus sharing agreements to transaction-specific investments when these investments are observable. Parties who invest relatively more tend to receive larger shares of ex post surplus, which in turn promotes additional investment and so overcomes part of the underinvestment problem inherent to these sort of sharing contracts. The analysis suggests that relational exchange is possible in settings where individual reputations and frequently repeated contact are absent, and thus supports Macneil's claim that social norms of equity are central to relational exchange.