Behavioral Mechansim Design

The general question I would like to explore is: if Homo economicus is not an appropriate way to model behaviors and should be replaced by the new behavioral paradigm, then what does that mean to mechanism design and public policy, especially in the second-best world with non-contractual social interactions and generalized incre asing returns?

To narrow down, we firstly focus on the liberal trilemma– the impossibility of Pareto efficiency, preference neutrality, and voluntary participation (Bowles 2016). Instead of replacing the efficiency standard with “incentive efficiency”, Bowles suggests another solution by putting the restriction on preferences – creating better people.

Along this direction, the first question we might ask is:what kind of restrictions on preference would ensure that efficient outcome can be implemented voluntarily? For example, Dutta et al. (2012) and Lombardi et al. (???)investigate the impact of “partially honest”– truth-telling is preferred when it won’t worsen the outcome – on the implementation of Nash equilibrium. Hwang and Bowles show that some degree of the altruism would be sufficient to overcome the trilemma (as cited in Bowles (2016)). How about trust and fairness? Assuming that others are good people under uncertainty would encourage interaction that brings mutual benefits, while the social norm on fairness providing a clear distributive rule of the division of the mutual benefits may reduce the risk of coordination failure.

Another question would be: how to create better people or how can public policy encourage the evolution of pro-social norms? First of all, the policy maker should choose the target, which is related to the previous question. Among all the possible “social norms”, some are desirable in the sense that they are compatible with Parato efficiency and voluntary participation. Next, in those desirable social norms, some are efficient in the sense that they cost less to be chosen in the evolution with subsidies (the total cost depends on the size of subsidy required and the speed of convergence).

So one of my questions is: can we develop some conceptual tools to help us identify those social norms that are both desirable and efficient? My hypothesis is that it may be related to some versions of Kantian morality. After a close look at the Roemer’s version of “ Kantian equilibrium” (Roemer 2015), however, I found that it may not be a good candidate, due to the misleading way to model morality as an optimization pro