The economics of the money-back guarantee

Back in July in The Conversation, Yalcin Akcay (Melbourne Business School) and Tamer Boyaci (ESMT Berlin) wrote an interesting piece on the economics of money-back guarantees:

...retailers are increasingly offering extra services such as warranty plans, free shipping and guarantees to reassure them. Selling with the “money-back guarantee” is a prime example of this.

This is because the economics of the money-back guarantee can work for retailers. These businesses allow customers to return products that do not meet their expectations — as a result of poor quality or a mismatch in taste — for a full or partial refund. Essentially offering their customers an insurance against the perceived risk of the product.

And research shows these retailers make a profit with this type of guarantee, given specific conditions.

For an article on "the economics of the money-back guarantee" though, I think it missed the most important economic aspects of this practice. Many goods are what economists call experience goods - goods where some of the characteristics (such as the quality of the good) are known to the seller, but are unknown to the buyer until after they have completed their purchase and received the good. The quality of the good is private information, and this creates an adverse selection problem which is especially problematic in the case of online sales.

In the absence of any other information (see below), since buyers don't know the quality of the goods they are purchasing online, they will assume that any goods on offer online are low quality. This is a pooling equilibrium - in the buyers' eyes, all goods are the same (low) quality. This lowers the amount that buyers are willing to pay for online purchases. Since the buyers aren't willing to pay much for goods of unknown quality, this drives sellers of high-quality goods out of the marketplace, because they can't receive a high enough price to justify selling their high-quality goods. The market for high-quality goods online collapses - the market fails.

Markets (including online markets) have developed ways to deal with this adverse selection problem, and generate a separating equilibrium (where high-quality and low-quality goods can be separated). One way is through signalling. With signalling, the informed party (the seller in this case) finds a way to credibly reveal the private information (the quality of the good that is being sold) to the uninformed party (the buyer). There are two important conditions for a signal to be effective: (1) it needs to be costly; and (2) it needs to be costly in a way that makes it unattractive for those selling low-quality goods to attempt (such as being more costly to those selling low-quality goods). These conditions are important, because if they are not fulfilled, then those with low quality could signal themselves as having high quality.

Money-back guarantees are a good example of an effective signal of high quality. They are costly - as Akcay and Boyaci note:

Customer returns cost retailers more than US$260 billion (equivalent to 8% of total retail sales) annually in the United States alone.

Those goods that are returned must then be sold at a discount, which is costly to the retailer. Money-back guarantees are also more costly if you are selling low-quality goods, since low-quality goods will be more likely to be returned. This makes it unattractive for sellers of low-quality goods to offer money-back guarantees. So buyers can be more confident that they are buying goods that are high-quality, when a money-back guarantee is offered. They can use this information to separate goods that are more likely to be high quality (those with money-back guarantees) from those that are more likely to be low quality (a separating equilibrium).

It is this signalling aspect that is the most important, when it comes to the economics of the money-back guarantee.