Disagreeing About What the Buyer Might Learn

I determine buyer-optimal information when the buyer anticipates worst-case disagreement over the induced signal distribution. The solution is a “twisted” version of the no-disagreement benchmark.