Substack

Monday, September 7, 2020

Examining the dynamics driving digital platforms

I blogged here about the battle between Apple and Epic on the former's predatory pricing of downloads of third party applications purchased through the AppStore. This post is an attempt at thinking through the first principles of digital markets and identifying the problems with the dominant narratives on digital markets.

Before that, bringing together the links to recent posts on the topic in this blog. This is about regulatory arbitrage enjoyed by internet companies, this and this about their questionable valuations, this about how internet companies compare with those from other sectors on innovation, this and this about how single-minded pursuit of efficiency (especially through digital technologies) is reducing resilience and engendering risks, this about how tech companies have become the wild west of capitalism, this is about the Amazon playbook of business practices which other tech behemoths too follow, and this and this about the Mathew Effect and unearned increments of corporate giants. Much of the evidence and logic for what follows draws from that provided in these links.

The first part of the post draws attention to the differentiators of the digital markets.

Any economic activity has two parts - production and its delivery to consumers. In the brick-and-mortar economy, production is about cultivation (agriculture) or manufacturing (industry) or performance (service). Delivery happens through a market consisting of wholesalers and retailers, with their respective physical locations. Like with goods and services in the brick-and-mortar economy, digital economy too has a production and delivery sides. The difference being that the delivery channel is predominantly digital and therefore virtual. 

Imagine a physical mall, with an owner. The owner lets shops out to retailers, who sell their respective goods and services to consumers visiting the mall. The mall owner collects a rent from the retailers and not a share of their revenues or profits. Neither does the mall owner exercise any control over the visitors (like being able to access private information and hold it, much less monetise them). Apart from the physical space, the mall owner's value proposition to both retailers is that a mall aggregates several retailers into one platform (like an e-marketplace). The mall owner monetises this  value proposition with a reasonable premium.

The same description applies to a physical shop selling consumer brands. Brands find the store a convenient platform to sell their goods. The shop owner is the aggregator in this context. All marketplaces are, therefore, a form of aggregation of the different sides of the market.

What keeps the mall owner from seeking to fleece the retailer or extract (like a mall entry fee) from the consumers is the competition from both other malls and also from stand-alone shops. 

With this analogy in mind, I'll point to two fundamental issues with the nature of digital delivery channel. 

1. The platform nature of technology applications means that marketplaces in the digital economy are private. It is fine to have private marketplaces (after all malls are privately owned), as long as there is competition. In the absence of competition, the platform owner becomes a extreme monopolist - leaving both the sellers and buyers with no option but to choose the platform. 

2. The real value of any digital platform is the network of transacting agents on both sides of the platform and the digital trails from their transactions. In other words, the value addition is an (almost) inevitable emergent property of the system. It is like the unearned increment of a landlord whose property happens to be adjacent to the new suburban metro station!

True, there is technology and innovation in the platform itself. But compared to its aggregation role and transactions data driven value addition, the technical (or functional) value proposition of the platform is arguably marginal. Put simply, the real value of a platform is the richness of its transacting networks. 

And if the platform is monetising their data, why should the clients who use the platform not charge it a fee for its use of their data?

If we take these two together, the digital platforms of today are equivalent to private malls, where the mall owners extract a share of the revenues of the retailers and also collect and use information about all visitors to the mall. 

Worse still, unlike the physical mall owners whose clients (retailers and buyers) and their transactions are fixed (or limited), the digital platform faces potentially no such restrictions. Besides, in unit transaction terms, the cost of investment required in case of the former is much higher than the latter.

These two factors need to be kept in mind while examining the regulatory case on market structure in the digital economy.

The next part of this post examines the argument that the exorbitant margins extracted by tech companies are just desserts for innovation and risk. 

There is no doubting that digital economy behemoths are innovative. The point is whether it (or anything) merits the extent of private appropriation that is currently the norm.

It can be argued (the links above outline the case) that the driving force for much of the innovation that follows a digital platform is endogenous and an intrinsic feature of their growth, an emergent property of a growing platform, for atleast two reasons.

One, what begins as a start-up offering a limited service, leads, in due course of time, to the creation of a platform through which opportunities to deliver other services too open up. The emerging versions of the platform is rarely a pre-planned or pre-conceived product. Instead it is an emergent feature of the original application as it grows. Growth itself creates unforeseen opportunities. And the entrepreneur (all things being equal) only responds to the emergent opportunities. 

Two, bigger the marketplace, greater the emergent opportunities. The commercial diversification opportunities that presents to a big platform like Amazon or Facebook or Google is qualitatively different and far superior to one that would present to a smaller marketplace. And this is less to do with the technical merits of the original platform itself, than the size of its networks and associated transactions data. In fact, the accumulated data by itself contributes significantly to the quality of the underlying services.

Incidentally, this also means that the risks assumed by these large platforms in pursuit of these emergent opportunities, while significant, need also to be seen in their true perspective. It also needs to be borne in mind that these companies have also come to acquire strong lobbying and influencing powers across all tiers of the government in the US. The point then is that they can experiment with the confidence that they have placed enough entry barriers to forestall competition.    

In other words, the major value proposition of a large digital platform company comes from the simple fact that it was the first mover. In typical markets, the first mover advantage disappears after a short period. But in digital markets, where marketplaces are private, the first mover advantage increases with time and places insurmountable entry barriers. 

Worse still, not only are these entry barriers restricted to the market for the original service, but to most new types of services and businesses that emerge from the platform of the original service. 

In light of all the above, the just desserts argument that supporters of digital platforms highlight is questionable. Where's a just dessert when the first-mover's profiteering is largely on the back of the exclusive privilege of a private marketplace platform in a mostly monopoly market whose network effects erects entry barriers not only on similar services but also all other services that can potentially be offered on the same platform, and whose innovativeness is itself a function of its size? 

Update 1 (03.11.2020)

Rishabh Kripalani and Thomas Philippon model two-sided digital markets and find that they need greater regulation that physical markets,
We study an economy in which consumers and merchants (sellers) interact on a two-sided platform. Consumers can share data about their tastes for different varieties of a single good with the platform which in turn sells this data to merchants. Data sharing increases gains from trade by improving match quality but gives more market power to the platform relative to the merchants which can reduce entry and consequently consumer welfare. This leads to an externality not internalized by consumers thus leading to more data sharing than is efficient. We highlight two reasons why more precise information increases the market power of the platform. The first is a copycat (private label) externality that increases the outside option for the platform of selling the good directly to consumers. The second is a consumer access externality that reduces the outside option of the merchants when information gets more precise, as more buyers are able to find their desired variety on the platform. Our model explains the qualitative differences between traditional retail platforms (physical stores) and digital online platforms and why the latter are more likely to require regulatory interventions that the former.

No comments: