That was the topic of a panel I was involved in at the Chicago Booth Annual Antitrust Conference last week. One thing you get from a Chicago panel is a diverse range of opinions. The panel was motivated by an article in the Yale Law Journal by Lina Khan written when she was a student and, having heard her speak now, she has ‘future politician’ written all over her. Khan throws the book at Amazon far more than any economist would which, of course, is why, for the most part, I don’t agree with her stance.
But before getting to that, let me point you to this concise explanation of why digital platforms tend to become concentrated provided by Ben Thompson (of Stratechery)
Ben was on the Amazon panel as well. The point is that digital platforms can be viewed as solving the ‘new’ hard problem that arises when digital technology shatters existing ways of doing things. In effect, Amazon is aggregating a variety of services from search to pricing to logistics and distribution of online content, to solve the problem of making online shopping easier, cheaper and safer. That is why most of us tend to be fans of Amazon’s existence. But just being a fan doesn’t mean we don’t want more competition.
My view on Amazon is similar to my view on other antitrust matters in this space. A static perspective rarely gives rise to an antitrust concern while a dynamic perspective could. The problem is that there is a great deal of uncertainty in that regard which makes pre-emptive enforcement a challenge. Here is my articulation of those views.
Back to Khan. She identifies many areas where Amazon has, in her opinion, violated antitrust law. The tough part in her argument is that Amazon rarely has what would be regarded as a monopoly. Online sales are still 9% of the US retail and Amazon has half of these overall but in individual verticals it varies. So you need to be willing to define markets narrowly to see Amazon in a bad light. She does, however, document some instances which do not look great at a first pass. In the end, Khan and many others in the conference believed that the problem lay at the foot of the consumer-centric tests for antitrust that are followed world-wide. They want a broader notion of harm that includes harm to labour, small business disruption and even threats to democracy.
On this I disagreed vigorously. At this point in the video, I went into a tirade (listen through my discussion of Amazon’s ‘war on shopping’). My point was that other countries — other than the US — had done just fine using traditional, economically grounded antitrust practices to deal with these sorts of issues and the US has pretty much chosen to be lax on this. I pointed out that Amazon had been quite open — with AWS and third party sellers — to potential competitors which makes it really easy to detect bad practices when they arise. Their current open behaviour means that it is easy to see when they go astray. You just have to want to do that when it happens.
Amazon is tricky. They are big. Competitive pressure on them comes from all over the place. And they have technological advantages. But using antitrust to discipline them is not an unfathonable problem. It can be done. The only thing is that you have to wait until they actually get the market dominance that people believe they currently have.
Digitopoly 
Thanks for this, I agree. Where does that leave the other policy concerns, the ‘broader notion of harm’ though? Antitrust law should be restricted to antitrust problems, but part of the reason we care about rights and externalities in digital markets is because of concentration – if small suppliers have no credible alternatives (in itself probably not a good argument in Amazon’s case, but as an example), and we care about the impact on their ability to conduct business and earn a decent wage, how do we intervene? And how do we consider the potential anticompetitive effects of such intervention (‘supplier protection’ measures might be much more affordable for Amazon than a smaller competitor)? There is no immutable hierarchy of policy objectives, it seems to me, and competition needs to be part of the mix even if we don’t want competition authorities worrying about the overall recipe.
There are two groups of Amazon customers. There are the buying customers who go to the site, place orders and so on. There are also the selling customers who sell their goods on Amazon and sometimes use Amazon logistics services and sometimes don’t. I think the bigger anti-trust problem involves the second group since they are both customers and competitors.
I’ve spent all my 30 years in the tech sector. I started selling and installing computers when I got out of college in 1988. I’ve spent the majority of my time on the consulting side (selling); about a third inside IT departments of traditional types of firms (consuming).
What I have learned – and what economists seem to haven’t the slightest inkling of how to comprehend or process – is that the rules of capitalism (never mind that many of its attributes are no longer enforced) do not apply when the vast majority of customers of computer technology make decisions. They cannot be informed because most cannot understand the technology they are shopping for. Due to this, they have no idea what a fair price is for any product or service. Most tech companies have made this the key tenet to their marketing – do not provide any information (I’ve shredded every vendor for purposely standing up dramatic web sites that are filled with flowery white papers, testimonials, and name-drops – but no facts on which to actually make any decisions. Huge conferences and elaborate road shows are orchestrated to convince the great unwashed that the crap they were sold last year is junk and the stuff they’re selling THIS year is FANTASTIC – and ALL of your competitors are using it, leaving you in the dust. Product development is at a point where we have VHS vs. Betamax showdowns every other year – not every 10 years.
The sales strategy is pure peer affiliation – convince the customer that everyone else who’s smart uses their stuff. One might say, “well, there are smart people in IT departments who will surely guide their employers to the best purchases,” but that is sadly not the case. First, many IT people are in it to fluff their resumes and get better jobs somewhere else, and are motivated to guide the company to buy popular things so that they can be more attractive to other companies. Second, many IT people are not as smart as you think/hope. Third, many IT people are completely susceptible to the affiliation fraud perpetrated by these many technology providers. Last, and certainly not least, the many solid IT people who advise “deciders” on the right technology choice are overruled because the “deciders” are the absolute target of the affiliate fraud marketing strategy (e.g. want to be able to brag at the country club about what they bought, etc.).
So when I read economists opining on the question of whether or not technology companies are abusing customers, leveraging market power, accumulating market power, etc. I can’t help but laugh because the conclusions are based on principles that do not apply. Amazon is highly complex business with highly complex products and has demonstrated a mastery of the art of tech customer manipulation (both retailers using their platform for sales and folks hosting web apps, among others). To appreciate their power, you only need to ask one simple question, “If you’re going to shop for and purchase a dog collar online, do you search for it in Google [wait, do they abuse their power?] or do you go to Amazon?” If you were a retailer of dog collars and want to sell online, what does the answer to this question mean to you? And if you’re Bezos and setting commissions for retailers, what does the answer mean to how you price?
Clearly they have massive market power. Attempts to minimize it by portraying various relative market sizes as de minimis is a diversion. The effort used to create the diversion would be better invested in understanding how warped the tech industry is. Your conclusions will have a lot more credibility if you had a better understanding of the way decisions on technology are made.