The Unintended Consequences of Headquarters’ Involvement in Decentralized Transfer Price Negotiations: Experimental Evidence

Abstract

This study investigates how headquarters’ involvement affects the efficiency of transfer price negotiations. Although prior research explores autonomous transfer price negotiations, evidence suggests that headquarters can become involved in these negotiations, particularly after they fail. Although the likely intention of headquarters’ involvement is to overcome inefficiencies arising from decentralized managers’ inability to agree on a transfer price, we suggest that such involvement can reduce agreement frequency and the efficiency of transfer pricing in coordinating transfers between divisions. Reduced agreement may occur because involvement can reduce managers’ perceived responsibility for the negotiation outcome and because they may expect headquarters’ decision to be more favorable for them than a negotiated price. Headquarters’ involvement can also reduce the coordination efficiency of transfer pricing because of information asymmetries and headquarters’ decision biases in interpreting negotiation failure and using available information. In an experiment, we manipulate whether headquarters’ involvement is absent or present. We also manipulate whether headquarters suggests a nonbinding price (weak involvement) or whether it imposes a price on divisions (strong involvement). Consistent with our predictions, we find that headquarters’ involvement reduces the frequency of negotiation agreement and the coordination efficiency of transfer pricing. Efficiency is reduced more when involvement is strong rather than weak. We contribute to research by studying managers’ negotiation behavior in the realistic setting of potential headquarters’ involvement and by providing evidence on headquarters’ biased perceptions of negotiation impasse and the unintended consequences of its involvement. This paper was accepted by Brian Bushee, accounting.