This paper – which can be found here – develops a rather orthodox argument about market leveraging and foreclosure. In short, the argument is that: “Under certain conditions, digital platforms can harness the power of reputation to steer users to favored suppliers. This steering forecloses non-favored suppliers in a related, though distinct, relevant market. Where favored suppliers are able to split the resulting rents with the platform, the strategy s rational. The resulting foreclosure reduces efficiency and consumer welfare.”
In short, “Platform markets exhibiting substantial algorithm-driven reputation competition can facilitate a unique type of competitive harm. This article briefly summarizes the marketplace evolutions that have fiercely intensified [algorithm driven] reputation competition. It then describes this novel second-order, out-of-market competitive harm that can arise as a result of certain conduct in such markets.”
Given this, it is unsurprising that the paper begins with a discussion of why reputation is relevant to competition. Antitrust doctrine generally treats “consumer choice” as being good. Yet the antitrust enterprise has not fully grappled with the “paradox of choice”: a larger assortment of choices can cause consumers to make less-than-optimal decisions. The downsides of information abundance—which can be experienced instead as information overload—prompted the rise of services that compile and refine information into a more useful “finished” product. In this context, reputation has emerged as one of the most vital facets of competition in many modern markets. A number of platforms – Uber, Yelp, Google Maps, Amazon—either incorporate or (in the case of Yelp) essentially comprise reputational mechanisms that are vital to their functioning because of the importance of filtering.
The paper then moves on to develop a theory of harm based on the relevance of reputation for competition. “In the presence of information overload, users often look to platforms to serve as filters. Here, reputation (via algorithmically ordered ranking) often plays a central role. The danger is that a platform may have the power to tilt the real-world playing field in favor of its own favored counterparties. Where this is so, a platform could be incentivized to alter its reputational system in order to deliver consumers to its favored partners. The ultimate effect is foreclosure of non-favored sellers from the real-world market. The resulting rents from the real-world market would allow the favored partners to compensate the platform for its exclusion of the non-favored sellers.” A case study of this – which is the author is at pains to describe as merely illustrative – is a recent merger between real estate portals which was approved without conditions by the FTC.
This is a short paper that develops a theory of harm in line with standard antitrust principles applicable to digital platforms. For its commitment to sticking to the fundamentals of antitrust alone, it deserves to be commended. While presently there are still only few theories of harm that are applicable to the new economy, I think this is a worthwhile endeavour – even if the theory itself would probably benefit from further refinements.