I study Cournot competition under incomplete information about demand while assuming that market price must be non-negative for all demand realizations. Although this assumption is very natural, it has only rarely been made in the earlier literature. Yet it has important economic consequences:...

Search engines face an interesting tradeoff in choosing the way to display their results. While providing high quality unpaid, or “left side” results attracts users, doing so can also cannibalize the revenue that comes from paid ads on the “right side”. This paper examines this tradeoff,...

I study Cournot competition under incomplete information about demand while assuming that market price must be non-negative for all demand realizations. Although this assumption is very natural, it has only rarely been made in the earlier literature. Yet it has important economic consequences:...

I study Cournot competition under incomplete information about demand while assuming that market price must be non-negative for all demand realizations. Although this assumption is very natural, it has only rarely been made in the earlier literature. Yet it has important economic consequences:...

I study Cournot competition under incomplete information about demand while assuming that market price must be non-negative for all demand realizations. Although this assumption is very natural, it has only rarely been made in the earlier literature. Yet it has important economic consequences:...

The challenge of effective merger enforcement is tremendous. U.S. antitrust agencies must, by statute, quickly forecast the competitive effects of mergers that occur in virtually every sector of the economy to determine if mergers can proceed. Surprisingly, given the complexity of the regulators...

We study the screening or nonlinear pricing for discrete consumer types when the single-crossing condition need not hold. A method is developed to reveal social eff±ciency or distortion of equilibria from market observations. The two known indicators of effciency, namely absence of almost-envy...

We study discrete-type screening without Spence-Mirrlees condition. Under non-separable and non-concave cost, all packages can be distorted in equilibrium, even when only the participation constraints are active. This and other paradoxical effects, shown by examples, are caused by some kind of...

It is shown that supplementing any usage-based pricing with a flatt-fee scheme is profit improving when the savings from transactions costs and from deadweight loss by using flat fee exceed the additional production costs. We use a most general model, without many traditionally used assumptions...

We study a multidimensional screening problem with minimal restrictions on valuations. Our ?-relaxation of the constraints excludes bunching and cycles in the graph of active incentive-compatibility constraints. Therefore, the Lagrange multipliers do exist and enable us in characterizing...