It's called competition
Amazon should be in the business of killing its competitors, even third-party sellers on its platform
Today, I’m going to talk about how great Amazon is and how the antitrust claims made against it for competing with its third-party sellers are mostly bunk.
But before that, I wanted to take a moment to give a quick intro to the blog itself. Will and I have a lot of passion for market mechanisms to be able to solve problems, especially informational problems. That’s why this is called ‘Prediction Markets’ and it’s why we full-time push prediction markets as useful social tools for people to be able to hedge risk and find new information they didn’t have before. Here, we have a space where we can apply that sort of knowledge in a way that we also wanted to when we were economic students.
Recent developments at the FTC to bring new antitrust claims to the major technology firms have heavily stalled out. A judge dismissed its suit against Facebook by noting that the FTC did not make a case for what exactly it is Facebook monopolizes. Internally, it’s apparently had a dramatic change in morale and satisfaction with leadership.
Amazon, Apple, and the other tech giants are very different from the monopolies of old, who (at least are alleged to have) restricted supply and raised prices. These platforms have ‘monopolies’ on consumers, not any products or inputs. As a result, a lot of the old ways we’ve dealt with monopolies don’t really apply—but new ways don’t really either. Chair Khan famously got her start by bashing Amazon’s ability to abuse third-party sellers on its platform, but I think that’s actually a great example of how Amazon is improving consumer welfare and competing with who they need to be.
Amazon and third-party sellers
Amazon receives considerable bad press because it both operates a marketplace and competes in that marketplace. The argument goes something like this:
Amazon is able to abuse its monopoly position by
Using private data only available to itself to
Push their own products ahead of competitors
Only, Amazon doesn’t have a monopoly on anything. Even if they have a majority of e-commerce sales, e-commerce isn’t actually a distinct market from normal retail. Amazon is competing with the hundreds of thousands of brick and mortar retailers in the United States. In fact, competing with in-person commerce was the premise of the entire business, not winning the (at the time) minuscule e-commerce market. Today, Amazon’s not even the largest retailer in the United States, which is still Walmart. Neither has even 10% of the market. Amazon doesn’t act like a monopoly, either; it has low, market-rate prices and invests an enormous amount.
If Amazon isn’t a monopoly, then this whole antitrust claim is backwards. If you support competition, you should support Amazon engaging in heavy competition with its third-party sellers. A worse, and plausible scenario, would be that Amazon takes its cut from the sellers, sits back, and chooses not to care about whatever happens in that market. Maybe even support a single winner and reap the benefits. Rather, customer obsession has led them to decide that they can win in new sectors and markets, to the benefit of all of us, who can now get our socks easily and cheaply with Amazon Basics.
The last few years have shown how unnecessary Amazon actually is to be in the retail business. Hundreds of thousands of businesses have set up on Shopify, selling products through their own shops. Shopify actually appears to be more successful at doing it. The loudest anti-Amazon ‘victims’, like the CEO of Diapers.com, don’t really grapple with how the long-term impact on their market from being crushed in competition by Amazon is positive for the consumer. Diapers are, unsurprisingly, still readily accessible for very low prices on Amazon, but also in Walmart, Target, and anywhere else you can think of.
The types of products hardest to not be on Amazon are those which are very commodity like and where consumers are price sensitive–something like diapers is actually a good example of this. You won’t find Allbirds, Birkenstocks, KKW or Kylie Cosmetics on Amazon, because those are closer to luxury, differentiated products that people are willing to seek out to buy uniquely. But if consumers are just looking for the lowest price, who cares that much even if Amazon competes with them, or even prioritizes its own products? In every retail store you’ve been to, the layout of products has been deliberately chosen in order to get you to pick some goods instead of others. Why should online markets be any different?
Network effects are real, and Amazon and Walmart have both become pretty large; however, in the age of the internet, I’m convinced every day that this is less important than the conventional wisdom suggests. In 2007, the Guardian ran this headline:
It may be trite, but competition genuinely is only a click away. This is much less of a barrier to compete than it is to even drive to another store pre-internet!
The ‘use of private data’ point is heavily overrated, if it’s true at all. Since this is all private Amazon internal behavior, there isn’t actually much concrete evidence on what Amazon does with whatever it has. I don’t doubt that it probably has some unique information–like what its users are doing and considering while using their website–its difficult for me not to believe that this information is not mostly available through some of the many market research firms that are out there. And clearly firms are still using the platform in enormous numbers, suggesting that the actual threat of this is still much less than the benefit of using Amazon’s platform. Amazon certainly doesn’t have information that actually is critical to one’s business, like cost structure.
Because it’s not a monopoly, a priori, it’s not clear to me that Amazon isn’t allowed to even arbitrarily prioritize its own products. After all, these third-party sellers are on the Amazon platform at its discretion. Banning them from the platform is definitely worse–something a real House bill would do right now–, but it’s allowed to do that.
I think it’d be useful to think of there being three different types of products sold on Amazon:
Products bought by Amazon wholesale and re-sold to you for cheaper. This was the only type of product offered from Amazon’s launch in 1994 until 2000. This is about 40% of Amazon sales.
Third-party products sold by a third-party seller using Amazon’s platform. Now, this is about 60% of Amazon sales.
Amazon’s own private label goods, like Amazon basics. This is a tiny fraction (~1%) of the Amazon sales.
That timing point is critical. By 2000, third-party sellers were clamoring to pay Amazon in order to access their marketplace, which had built up a brand and would let them avoid the difficulty of setting up their own online delivery businesses. It seems perverse to punish Amazon by limiting or preventing them from doing (1) or (2), or even (3).
Concern about Amazon’s private label goods is well over blown—not only is it a tiny fraction of its own market, its tiny compared to almost any retail. Putting your own products next to competitors is a key part of almost every store—have you ever been to a Trader Joe’s?
Now that you’ve read this far, have you considered that oil prices actually fell considerably under the Rockefeller monopoly?