Using a conception of reciprocal rather than distributive justice, we
investigate the fairness of power strategies in exchange networks. We
propose that norms of fairness affect the risk of using different stra
tegies, by influencing the probability that they will provoke retaliat
ion or resistance. Using computer-simulated actors to manipulate strat
egies, we investigate how subjects perceive and respond to their partn
ers' strategies that vary in reciprocity and power base. As predicted,
subjects in power-balanced networks judged both reward- and punishmen
t-based power strategies as more unfair that reciprocal strategies, an
d punishment strategies as more unfair that equally nonreciprocal rewa
rd strategies. Punishment strategies provoked more resistance that rew
ard strategies but, unexpectedly, no greater retaliation. Over time, t
he instrumental effects of the strategies tended to overcome reactions
to injustice.