Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We consider the efficient allocation of a single good with interdependent values in a quasi-linear environment. We present an approach to modeling interdependent preferences distinguishing between "payoff types" and "belief types" and report a characterization of when the efficient allocation can be partially Bayesian implemented on a finite type space. The characterization can be used to unify a number of sufficient conditions for efficient partial implementation in this classical auction setting. We report how a canonical language for discussing interdependent types - developed in a more general setting by Bergemann, Morris and Takahashi (2011) - applies in this setting and note by example that this canonical language will not allow us to distinguish some types in the payoff type - belief type language.