Arbitration problems boil down to information design problems. The arbitrator always has to solve the information design problem, once she has done that, designing the arbitration problem is mechanical. [pdf]
A bureaucrat decides whether to approve a project based on his information, only a part of which is verifiable in court. Is more precise verifiable information always better? [pdf]
If you want to settle a legal dispute without the need of (further) discovery, but you cannot guarantee full settlement, then the optimal ADR protocol has the following properties: (i) a neutral-third party manages the information flow, (ii) the information disputants obtain from participating in the negotiations is independent from the information they provide in the process, (iii) even ex-ante symmetric parties are treated asymmetrically during the process, (iv) settlement is not guaranteed for any potential constellation. [pdf], [program ADR (matlab)]
If a borrower is in financial distress, he has better information whether this due to a (shorter) dry spell or because his investment is fundamentally rotten. That information is valuable to banks who have to decide whether to roll-over credit or to foreclose. Bankruptcy laws affect banks decisions and thus prices in the credit market.
To solve a problem by a deadline an agent can apply a trial-and-error approach, or she can aim to tackle the problem through a more structured, systematic approach. When should the agent switch between approaches? [pdf]
Coordination on standards is hard if firms have private information the quality of their innovations. A firm can persuade a competitor to participate in the SSO by credibly threatening to release information about its own quality. Vaporware, Newsleakage, and Betaversions are potential business strategies to exercise the threat [pdf]
How would an agent decide on the order of proposing projects to a principal? An accepted proposal forces the agent to work on it, once rejected it is hard to convince the principal to take up the project again. Is it wise to propose the agent-preferred project early on? Should she wait to signal that this truly is a good project?