In this paper, we question a very deep-rooted bias in the economic literature with regard to conflict and revolution. Conflict in general and revolution in particular are not necessarily 'dark side of self-interest' or bad things. They may be sources of political and economic efficiency depending on their incidence on institutional change. Revolution escapes from Hirschman's dichotomous corrective mechanisms of 'voice' versus 'exit'. We consider revolution as scream exiting from existing rules and voicing new rules. Our theoretical framework suggests that revolutions are a source of institutional innovation characterized by a process of deinstitutionalization and re-institutionalization. The Veblenian 'selective adaptation' on the basis of differentiation, inheritance and selection through a survival struggle provides an endogenous explanation of its different stages. We explore the impact of revolutions on institutional change by focusing on two major types of revolutions, namely the elite revolution and the mass revolution. The former is led by a fraction of the old elite group, while the latter is driven by social groups and classes that were previously non-dominant. The 1688 British Glorious Revolution and the 1789 French Revolution are emblematic figures of these two types of revolutions. The great demarcation between property and sovereignty was accomplished by both revolutions traversing through different patterns of institutional innovation. Our references to the British revolution are exclusively limited to old and new institutionalist authors to show the consistency between their theoretical framework and conflict as a means of bargaining. By contrast, our analysis of French revolution tries to test the consistency of our theoretical framework in light of empirical evidence.