The Italian Amazon Case and the Notion of Abuse

Abstract

In a very recent decision, the Italian Competition Authority (ICA) condemned Amazon for making Prime and other Amazon services accessible exclusively to commercial customers who delivered their goods using Amazon’s logistics service – called Fulfillment by Amazon (FBA) – in lieu of other independent logistics service providers.

However, it is unclear how the ICA framed such behavior. It explicitly qualified Amazon’s conduct as a case of self-preferencing, although it did not build this charge by meeting the liability conditions that the European General Court affirmed in Google Shopping. At the same time, the Amazon decision is replete with words and expressions evoking a tying case: First, the ICA frequently refers to the company’s different product combinations as well as to the idea that Amazon’s business clients did not spontaneously chose such pairings. Furthermore, the ICA used the adjective “essential” to describe Prime and the other Amazon services, dealing with them as if they were essential facilities. Indeed, the remedy of making Prime and those services accessible on FRAND (fair, reasonable, and non-discriminatory) terms to anyone who meets certain quality requirements is reminiscent of the duty to share that is usually adopted in essential facility cases. Finally, there is room to argue that because it focused on the effects of the company’s practice, the ICA overlooked the form that practice took – and the class of exclusionary practices under which it could belong.

Now, such a legal ambiguity – or, at least, the ICA’s reluctance to pigeonhole Amazon’s conduct into a single class of practices and write the decision accordingly – raises a general theoretical question: if one believes – as we do – that the effects-based approach would be the most appropriate to assess monopolistic practices, why does it matter whether Amazon’s conduct is a case of tying, a case of essential facility, or a case of self-preferencing? More explicitly, if a dominant firm is said to be abusing its power when its conduct is likely to exclude rivals in an anticompetitive way without producing any efficiency or innovation gain in return, why does the form of that conduct – or the class of practices to which such behavior is said to belong – matter? Or should one believe that qualifying a practice as self-preferencing sorts things out, because such a qualification traces the conduct to a family of practices – discriminatory practices – that differs from that of exclusionary practices?
The chapter examines these issues.

Keywords: Self-preferencing, amazon, abuse of dominance, foreclosure

JEL Classification: K21