A Mechanism Design Approach to Identification and Estimation (with Brad Larsen, January 2017)
This paper provides a new, nonparametric identification and estimation approach for a variety of incomplete information games, both static and dynamic. The approach relies on the Revelation Principle, exploiting the incentive compatibility of the direct revelation mechanism corresponding to the underlying and unspecified game, rather than attempting to solve for or specify the extensive form or equilibrium strategies of the game directly. We illustrate the approach using simulated and actual data from bargaining settings.
Private ownership creates monopoly power, harming the dynamic efficiency of asset allocation. Common ownership improves allocative efficiency, but eliminates incentives to invest in the common value of assets. We propose a simple partial private ownership system, Harberger licensing, for public assets. Lessors self-assess a price at which they commit to sell the asset to any interested buyer and pay a tax on this price. In a calibrated dynamic trade model, setting Harberger tax rates using a simple rule-of-thumb – half the observed turnover rate – increases steady state value by 4.6% of asset prices under full private ownership.
Implementability, Walrasian Equilibria, and Efficient Matchings, with Piotr Dworczak (2017), Economics Letters 153 pp. 57–60.
In general screening problems, implementable allocation rules correspond exactly to Walrasian equilibria of an economy in which types are consumers with quasilinear utility and unit demand. Due to the welfare theorems, an allocation rule is implementable if and only if it induces an efficient matching between types and goods.