Simply being labeled "high-quality" doesn't ensure a product will live up to that claim. How can one be sure such claims are reflected in the product?
Suppose you’re interested in buying high-quality kitchen knives. You start the search online and Google Shopping returns an abundance of options, sortable by review and price. Many knives are labeled high-quality, but how can you be certain that this claim is reflected in the product?
Opacity around product quality, with examples ranging from “lemons” in used car markets to toxic financial assets, is known by the term adverse selection, and it creates a number of market inefficiencies. To work around this problem, many markets have developed methods of “signaling” that are used to differentiate high-quality from low-quality. In the case of used cars, for instance, a salesman might signal his trustworthiness and the quality of his cars through professional certification or warranties on every purchase.
Marketers have long theorized about the effects of signaling. But a recent paper by Yale’s Kosuke Uetake, with Kei Kawai from New York University and Ken Onishi from Singapore Management University, offers the first-known empirical analysis of this phenomenon. Their results quantify the extent to which signaling affects market outcomes and welfare relative to the same market in the absence of signaling.
Using the peer-to-peer loan website Prosper.com, Uetake and his coauthors studied how prospective lenders were able to interpret and apply signals provided by borrowers. Because the integrity of a borrower is not easy to determine, the trade of unsecured loans is considered a classic example of a market that suffers from problems of adverse selection. A key feature of Prosper.com, however, is that each borrower can post the maximum interest rate that he or she is willing to accept. It turns out that that those who are willing to pay high rates are more likely to default than those who post low rates. For lenders, this implies that the posted interest rate loosely signals the creditworthiness of the borrower.
With this information, Uetake and his colleagues devised a model that tested three different scenarios based on the Prosper.com environment: a market in which interest rates are interpreted as signals, a market in which they are not, and a market in which lenders were fully aware of the integrity of borrowers. This design allowed the investigators to quantify the extent to which credit markets suffer from adverse selection, and the extent to which signaling can mitigate this effect.
In most cases, Uetake and his coauthors found that signaling increased the overall welfare in markets riddled with adverse selection. That is, when a group of borrowers included a number of people who were likely to default on their loans, then signaling through interest rates proved an efficient mechanism for improving market outcomes. Interestingly, this was not true among groups of borrowers with good credit and a high likelihood of repayment—people with an AA credit rating score. In this case, lenders remained wary enough of low interest rates that they passed up opportunities to loan to creditworthy borrowers; signaling from those who were most creditworthy actually reduced overall market welfare.
As the first paper to empirically examine and provide analysis of signaling in industrial organizations, Uetake and his coauthors have taken a critical step to developing a far richer understanding of the costs of adverse selection along with the potential benefits of signaling.