Parametricity

The political economy of platforms

Posted at — May 3, 2020
A platform for hamburgers. Think GrubHub.

In this note, we’ll try to understand what is a platform, and whether they can be improved with a “coop” structure.

In particular, we’ll try to understand:

  1. What are the tensions inherent in platforms as an organizational form, and how can they be resolved?
  2. What are the prospects for substantive impact of platform cooperatives in the presence of competing capitalist-run platforms?

What is a platform?

Before really understanding platform coops, it is necessary to define “platform” and examine the various types of people interacting with it.

Most people are familiar with platforms. Uber, Doordash, TaskRabbit and the like come most immediately to mind. Others include Amazon Web Services, the Amazon marketplace, and the Apple App Store.

Let’s provisionally adopt the following definition that generalizes the foregoing examples:

Provisional definition (platform)

A platform is a social/technological formation which facilitates more direct forms of production and sustains itself by taking a portion of the proceeds of that production.

This definition implies the existence of three kinds of parties:

  1. Operators. The party responsible for software development, marketing, participant acquisition, etc. In other words, the party whose activities produce the platform itself. This could be, for example, a capitalist firm along with its owners.
  2. Producers. This party gives some of the proceeds of its production to the platform, either in the direct form of exploitation of labor, or in the indirect form of a fee. This can appear in a direct form, as with Uber drivers or TaskRabbit workers, or in an indirect form as with AWS.
  3. Consumers. This party consumes the goods and services produced by the platform producers.

Internally these parties may have their own power structures and dynamics. We will mostly postpone analysis of that for now until we examine some concrete examples.1

With this view, it is clear platforms are set up for certain conflicts, because one can see these three parties’ interests are not at all aligned. A few examples:

The power of the intermediary position of operators

“By default”, operators are in a very advantageous position in this three-way conflict:

What are the material conditions that have led to the existence of platforms?

This varies on a case-by-case basis. For a VC-funded platform-operator firm, the most immediate reason is that enough VCs thought that they had the potential to be very profitable.

Uber is an instructive example. I doubt so many VCs would have gotten on the Uber train if it were structured as a traditional firm in which drivers were employees. In the platform organizational-form, Uber has a much stronger position from which to exploit the labor it oversees. A few reasons:

Do platforms facilitate production which would not have happened otherwise?

This also varies on a case-by-case basis. I think one can confidently say the App Store (in the context of capitalism of course) made the development of certain apps which would not have otherwise been possible, since otherwise there would have been no business model for them.

On the other hand, did the fact that Uber is a platform cause rides to happen that would not have otherwise? In some sense it did, because it helped make it possible to exploit drivers more than had been possible before which (along with VC subsidization) drove prices down enough that people used cabs more. But this is quite a different story to the App Store example.

What is a platform coop?

The most common organizational structure of a platform operator in the present-day is that of a for-profit corporation. In broad strokes, control over production and profit distribution is split between shareholders and high-level managers and workers at the operator corporation. The firms are operated for the extraction of profits for shareholders.

A platform coop is a platform in which measures are put in place giving producers and/or consumers some degree of control over the operator. This may be in a minimal way, e.g., say producers just get a say in where profits are distributed. Or it can be in a more significant way, e.g., say producers and/or take an active role in defining the priorities of platform development and have more granular control over operator decisions.

In principle, one can set it up with any distribution of powers between the three groups (or subsets of the groups.)

It’s important to note that the operator organization itself may have investors and each producer or consumer may even itself be a capitalist firm (e.g., in the case of AWS) and certain conclusions will depend on the internal structure of the operators and producers (and even of the consumers).

Challenges for coops in capitalism

Any coop which is beholden to capitalists — either indirectly for funding, or more directly to investors who have a controlling stake in the operator corporation — is at risk of betraying the interests of its producers, consumers.

The reason is simply that those capitalists will have interests that are not aligned with the producers and consumers, and will use its leverage with the operator to steer the operation of a platform in the direction of those interests.

An interesting case study will be Savvy, an organization whose business function is to provide medical patients with gigs with pharmaceutical and medical services companies that need information from patients. For example, patients can take surveys.

Here the producers are the medical patients and the consumers are the pharma and medical services companies2.

Just a few weeks ago, Savvy raised funds from Bryce Robert’s hiply-branded Indie VC. Savvy says it will remain majority producer-controlled. This is a good starting point since the consumers are capitalist firms that do not have the patients’ interests in mind. On the other hand, unless the producers are highly organized independently of the operators, they are may find themselves exploited.

Why? In the worst case, one must assume the operators themselves are completely aligned with the interests of their capitalist investors. And so the operator’s primary interest is in profitability3.In the last analysis, the investors hold all the cards as they can pull the plug on the enterprise.

It is interesting to note that in the case of Indie VC, the information in this interview seems to imply that they may have a 3X cap on returns, at which point the operator can buy them out. Given the pressure on capitalists to maximize profits, my guess is this will forever be a niche feature of the investment landscape and not be a widely deployed option. Moreover, such financing will likely not be an option for high-risk, high-initial-capital-requiring ventures.

Let’s return to analyzing how this power structure could play out. Suppose there is a feature which Savvy could add to their platform which would result in increased profitability at the expense of the well-being of the producers. For example, some psychological manipulation in their UI which would lead them to take more gigs than would be best for their happiness, as with Uber.

The operators, owing to their interest in profitability, may be fine with such a decision and the producers may have no transparency or oversight over such a decision, being as it is a low-level operating detail of the platform. The “consumers” (i.e., the medical companies) would be very much in favor of such a change.

So there need to be mechanisms in place for the producers to regulate updates to the platform itself. This may be the case with Savvy, I couldn’t find any information on what the coop structure actually is.

Generally speaking, the operators themselves will require some initial capital to start the platform. And ultimately, under capitalism, capital can always restrict access to capital for platform-coops relative to their non-coop competition to starve the coops out.

Technological tools for enforcing control over the operator

As discussed, the main problems with platforms come from misalignments in the interests of operators with other platform participants. This can be mitigated by giving producers+consumers some degree of control over the operator. That control may be difficult to enforce on a day-to-day basis with legal or social means (strikes, boycots). In that case, there are some technological tools which may be of use.

We categorize these by which power asymmetry the tool mitigates.


  1. One additional important note. We can talk about the “legal” instantiation of platforms, where all three entities are legally distinct. But we can also think of platforms as a pure organizational form. That is, much of the analysis here could still apply, say, to capitalist firms which internally have sub-organizations of “operators”, “producers”, and “consumers”. ↩︎

  2. This situation is independently interesting as a coop where the producers are relatively disempowered individuals and the consumers are powerful capitalist firms. ↩︎

  3. There are a few hitches here: the investors are likely interested in pursuing strategies leading to maximal expected profit, while the operator-managers more in a lower variance but lower profitability strategies. ↩︎