How to Segment a Search Market: Information Design and Directed Search

This paper examines when the public provision of information in search markets improves welfare. I consider a two-sided frictional search market in which buyers match with vertically differentiated sellers. The market is segmented into submarkets based on seller types. Such segmentation serves as a public signal that buyers use to direct their search. Given a segmentation, I characterize both the socially efficient and the equilibrium allocation of buyers across submarkets, and identify a Hosios-type condition under which the equilibrium allocation is efficient. I then analyze the design of surplus-maximizing segmentations, showing that the nature of search externalities determines when the constrained-efficient segmentation fully separates seller types or pools them into at most a binary partition.