SOME LIMITATIONS OF THE MODEL
Every economic model makes simplifying assumptions, and the simplifying assumptions set limits on the applicability of the model. That the model in this chapter only considers the last, cooperative stage of the coalition formation and decisions is a limitation that will be reconsidered in later chapters.
The model presented here is strictly a rational-action, full-information model. The only informational limitation assumed is that cooperative decisions can only be made among individuals who are linked in a previously- formed network, and this is not relevant to this chapter, which takes the coalition as formed. However, if we recognize that human rationality is bounded and is commonly expressed in “optimally imperfect” rules of thumb (Baumol and Quandt, 1964; Simon 1978; McCain 2014a, Chapters 6-7) then the maximization model presented here can only be approximated.
The normalizations discussed in section 11.2.1 lead to the independence of allocative decisions and distributive decisions. Conventional as they are, they are far-reaching. If we apply the normalization of equation (11.9) to all products simultaneously, we are assuming that utility is additively separable in all products without exception. But this is almost certainly empirically false. Empirical realism would be served if instead we adopted the non-transferable utility approach along the lines of McCain (2013, Chapter 6, section 5), but for that model the numerical expression of the firm’s value created is measured in units unique to that firm. Conversely, when we measure the output of a firm as its value added, and total output as GDP, we are implicitly assuming additive separability, whether we acknowledge it or not. Perhaps additive separability is best considered as a local approximation to a more complex global reality.