The Problem of Inefficiency: Market Failure or Modeling Failure?

Abstract: Most economists believe that we can model inefficiency or “market failure” with rational choice models, but we cannot. This paper identifies an implicit modeling error found in all rational choice models that claim to model inefficient behavior. Put simply, if we program an agent with a binary choice set containing just A and B, we can’t complain that they don’t choose C. If choice C is a worse option than A and B then the model is a valid simplification. But if there are any C choices that are better than A and/or B, as alleged in market failure claims, we have a “model failure” because the agent’s best possible option, C, has been incorrectly (and inexplicably) excluded from the modeled choice set. This modeling error is fixed by building the ex post market failure “solutions” into the ex ante choice sets of the agents. But with this correction we discover that the so-called “government”, “central planner” or ”bumbling bureaucrat” that we typically invoke ex post to mandate a tax or regulation is really just a “governance entrepreneur” offering an efficiently governed rule to the parties as an efficient choice. Correctly modeling governance as a choice sheds light on a vast market for privately ordered governance. It also generates a surprising but Coasian implication: just as the ability to easily transact over private rule choices can make the public law irrelevant, the ability to easily transact over private governance choices can make the public governance irrelevant.