【正文】 235812。s problem. American Economic Review, 63(2): 134139. Sappington, D. (1991). Incentives in principal agent relationships. Journal of Economic Perspectives 3(2): 4566. Tan, Guofu and Yilankaya, Okan, (2004). Ratifiability of Efficient Collusive Mechanisms in Second Price Auctions with Participation Costs. *.pdf Tommasi et al, (2003). Centralization vs. Decentralization: a PrincipalAgent analysis. *.pdf Vafai, Kouroche. (2005). Collusion and Organization Design. Economica, 72: 17–37Williamson, Oliver E. (1998). Transaction Cost Economics: How it works。 What definition of collusion am I going to use?The methodological approaches observed in the literature I have encountered so far have primarily been adopting positivist, quantitative methodologies in their approach to researching. They tend to use what they consider to be statistically relevant information to build representative equations that they then, justifiably, appropriate into algebraic equations with are then further manipulated to reason their propositions. Although significantly appropriate, I do not feel that these techniques would be transferable for my propositioned research topic – only the conclusions derived. However, it is understood that when determining a research methodology it is suggestible to incorporate both quantitative and qualitative methodologies into the research. As such I will find the grounding, for the framework I will construct, in the positivistic results acquired from quantitative methodologies used by the authors referenced in the literature. From there I intend to engage in case studies so I can receive and interpret the reality of collusive situations. In saying that, it bees apparent that the positivistic methodological paradigms, that attempt to explain objective reality through deductive methods, are not appropriate for my methodological research intentions but are suitable as referent information describing positivistic/statistical ‘truths’. And it’s for this reason I reject the positivist methodological paradigm for the developments of a framework, instead I am looking for an inductive paradigm from the following: Critical Theory, Constructivism or Realism. I can further rule out Constructivism as it is Inmensurable and relates to the subjective dimension. That leaves both Critical Theory and Realism. Unfortunately, at this stage, I find it incredibly difficult to differentiate the two – let alone try and chose one over the other. Both seem to offer themselves as being exceptionally valid。 essentially a kind of information ransom from the perspective of the principal. However, by colluding, agents incur costs to one another (Quesada, 2002). For this reason not all collusive situations are beneficial to all of the agents involved because the costs may outweigh benefits (Bolton, 1991). As pensatory measure, the relative benefit of collusion is often distributed proportionately equal to the market share that each of the collusive parties had when they first entered the collusive arrangement (Roos, 2003). At first this sounds like a fool proof way for the colluding agents to gain stability and control over the market, but such static market places are the prime location to find new entrants with cleverly diversifying strategies. A new entrant to the market place may delay their inclusion to the collusive arrangement (at high cost to themselves) with the knowledge that the more market share they have before they agree the greater proportion or the benefit they will receive. This reintroduces a petitive nature into dynamic collaborative arrangements as it reminds us that each independent organisation has its own strategical goals and objectives. “Heterogeneous firms make collusion, entry, exit, and investment decisions within an evolving environment.” (Roos, 2003. pp1) As such, the need to evaluate the ideal petitive and advantageous course of action regarding collusive action for the organisation is a major strategic concern. The principal can benefit, if aware of the arrangements, by utilizing this informational advantage by favouring the excluded agent and extracting maximum rent – boycotting the colluding agents if possible. The agent can benefit by forgoing immediate collusional behaviour in favour of a more timely application. Obviously these options are not always available as they are constrained by timeliness and information access (Roos, 2003).Hinging heavily on game theory, colluding agents “…are assumed to devise plex conditional punishment and reward regimes to create the incentives necessary to sustain an optimal degree of collusion.” (Roos, 2003. p4). The agents need to make the incentives strong enough, or the punishment severe enough so that a multiple party