Skip to main content
eScholarship
Open Access Publications from the University of California

Essays on Platform Markets

  • Author(s): Barach, Moshe Adiel
  • Advisor(s): Morgan, John
  • Tadelis, Steven
  • et al.
Abstract

This dissertation is comprised of three studies investigating how the provision of public information facilitates matching in online marketplaces.

The first study examines how employers on oDesk.com, the world’s largest online marketplace, use public information in hiring. I show that when employers are searching for someone low-skilled the provision of coarse information is sufficient. When employers are looking for someone high-skilled they will pay fixed screening costs to acquire information beyond what is provided by the platform. When information is not provided by the market- place all employers will pay to acquire more information. This leads to more matches and hiring quality workers at a lower price. However, the cost savings from hiring these low-cost, but high-quality workers does not outweigh the upfront cost of information acquisition.

The second study investigates how employers on oDesk.com alter their hiring behavior when some applicants, who were selected to be of higher quality by a machine learning algorithm, are covered by a money-back guarantee (MBG). The MBG increases the probability of transacting for employers looking for high-expertise talent. However, conditional on hir- ing, these employers do not substitute to observably different employees. Employers that are looking for lower expertise applicants substitute to higher quality applicants who are backed by the money-back guarantee. A follow-up experiment found that, given some reasonable assumptions, employers found the MBG to be more useful as a signal of applicant quality than as a risk-shifting mechanism.

The third study uses a series of empirical tests to isolate the mechanism through which the weather affects bidders’ willingness to pay (WTP) on eBay. I find that some variation in bidders’ WTP across multiple auctions can be explained by weather affecting mood. Changes in mood can alter WTP through two channels. Mood can directly alter the perceived value of the item or mood can influence cognitive processes, changing bidding strategies and indirectly changing the bid amount. I concluded that rain alters mood, which alters bidders’ perceived value of an item on eBay leading to bidders increasing their bids.

Main Content
Current View