Legitimation, Endorsement, and Stability

Intuition and formal theories of legitimacy suggest that endorsement of a social order should have important effects on compliance to and the stability of that order. Evidence from experimental and nonexperimental research fails to offer consistent support for that argument. This article contends that investigators have generally failed to recognize acts as objects of legitimation and entirely neglect variation in the sources of legitimacy. As a result, models of legitimation processes are not properly specified and the effects of endorsement are generally confounded with the effects of other uncontrolled sources of variation. We identify three objects of legitimation-persons, positions, and actsand three sources of legitimacy-propriety, endorsement, and authorization-and argue that each is important to a proper analysis of legitimation processes. The laboratory investigation which we report minimizes uncontrolled variation from those additional independent variables and we demonstrate that endorsement of an experimental task structure delays or prevents approximately 38 percent of change which occurs in task structures which are not endorsed.