Development economics seems to rediscover the importance of domesticating violence as a condition of democratic empowerment. According to the theoretical framework developed by North, Wallis and Weingast (NWW), developing countries are defined as limited access orders that reduce the violence potential of powerful elite organizations through the manipulation of economic rents and the negotiation of more or less stable elite coalitions. NWW's framework must be credited for reintroducing social orders in economics' research agenda; allowing for different paths to development, and highlighting violence and conflict as a central force in societies. In this paper, we first go back to the discussion of two main challenges in NWW's theory and methodology. The first one is their pluralist understanding of open access societies. The second is the way they link micro-level behavior to macro-political changes. We show here that, though alternative strategies could be followed, the methodological problem underlined by critics is somehow inevitable. We then apply the alternative frameworks we designed to Thailand, which has been through what observers may call a " failed democratization process " in the last decades, in order to bring out the key challenges to build up analytical frameworks of failed processes. 2