RE-thinking market failure in the light of the imperfect state

We propose a formal re-definition of the concept of market failure based on the idea of the imperfect state. In the neoclassical taxonomy, a decentralized regime of exchange is a market failure if its laissez-faire equilibrium solution is welfare-dominated by a technically feasible alternative. If the state is perfect (that is, it is benevolent) and competent (that is, its transactions cost of intervention is zero), then every market failure can be remedied with a welfare gain. On the other hand, if the state is imperfect (that is, either non-benevolent or with non-zero transactions cost of intervention), the state intervention to correct the market failure can be welfare-reducing. Extending the logic behind Williamson’s remediableness criterion and Stiglitz’ constrained Paretoness, we introduce a new taxonomy of failures: the concept “proto-failure” now denotes any failure which laissez-faire interaction cannot remedy with a welfare gain. The label “market failure” now denotes a proto-failure, which the relevant state can correct with a welfare gain. A proto-failure that the relevant state cannot correct with a welfare gain we call “Remediableness Condition efficient”. We use the net welfare metric, which explicitly accounts for transactions cost of intervention as efficiency criterion. The new taxonomy is equivalent to the old if the state is perfect, that is, all proto-failures are market failures. When the state is imperfect, the set of market failures is smaller than the set of proto-failures. A proto-failure is a necessary—but not a sufficient— condition for a welfare-improving government intervention. This paper follows the Williamson counsel to “push the logic of positive transactions cost to completion”.