Testing the conflict-performance assumption in business-to-business relationships

There is little empirical research published testing the interdependency between conflict in business-to-business relationships and commercial performance. The "conflict-performance assumption"-all other factors being equal, relationships where conflict is low will outperform relationships where conflict levels are higher-remains central in the marketing channels' literature despite insufficient and contradictory empirical evidence. There are several explanations for the lack of a clear relationship between conflict and performance. Rosenbloom [J. Mark. 37 (1973) 26] theorises that the relationship between conflict and channel performance follows an inverted U-shaped curve, where conflict is most productive at moderate levels and least productive at very low or high levels. Others have argued for a simpler, linear relationship between conflict and performance, usually negative in nature. Various theories about the conflict-performance relationship are empirically tested in a large marketing channel, using a number of dyadic and monadic measures of conflict (latent, perceived, and affective) and two objective measures of performance (effectiveness and efficiency). A linear model (performance declining as conflict increases) is adequate to explain the relationship between dyadic measures of both perceived and affective conflict and channel effectiveness. A threshold model is found to be superior to a linear model in explaining the relationship between dyadic measures of perceived and affective conflict and efficiency. Conflict increases slowly as efficiency falls until a threshold is reached when conflict escalates. Practical implications include that companies need to consider whether performance criteria affecting efficiency are as important to their business partners as those affecting effectiveness. If they are not, then business partners should be rewarded for meeting any such criteria that are more important to the one side of a relationship dyad than to the other. © 2002 Elsevier Science Inc.