I’d like to talk about something that I kind of see as an inevitability in the tech world. You might agree or disagree with me, but hopefully I can expose some readers to a new concept.
Background
In the last decade or so (and especially since the pandemic) large tech companies have across the board engaged in what Cory Doctorow calls enshittification. To describe this process simply, large tech companies have hit a wall and found it harder to produce value, as all of the low-hanging fruit has already been picked. In order to maintain growth, they must capture value from “inefficiencies” that already exist.
This “inefficiency” might be the price elasticity of a commodity. For example, say Netflix charges $10 a month for a subscription, but they could raise it to $12 before seeing their subscriptions drop. If you have ever played Rollercoaster Tycoon, you are already familiar with this idea. When you check how much you can charge for the bathroom people start to complain about it being too expensive, you are acting as an optimizer, maximizing the profit of your park. You are eliminating “inefficiency”, but did you actually solve a problem? No, you just maximized profit at the expense of everyone else.
Why would users tolerate the Enshittification of every single big tech offering? Shouldn’t the market be creating competition from smaller firms? The answer is because there is no realistic way to compete anymore. The “digital enclosures” that were created during the 70s, 80s, 90s, 00s, and 10s are almost impossible to break into.
average startup idea in 2024
Go look at the list of YCombinator startups. How many post-2013 do you recognize?
(Note: I’d like to put some actual numbers behind my point here, but I don’t have the time or energy to value all of these YC startups. So let’s just use brand recognition as a stand-in for success.)
https://www.ycombinator.com/companieshttps://www.ycombinator.com/companies
This is even with the help of YCombinator and all of Silicon Valley’s venture capital firms. Try as they might, they cannot produce value on the scale that they were once able to. So much so in fact, that earlier in 2024, YCombinator put out a post asking for startups in robotics, space exploration, defense, etc. When was the last time an investor asked someone to make a company for them to invest in?
https://www.ycombinator.com/rfs
To over-extend a Marxian idea, we could view the era of the Apple’s, Microsoft’s, Facebook’s, and AirBnB’s as an era of primitive accumulation for the high-tech sector. The amount of work necessary to produce a home computer in the 80s was several orders of magnitude less than in 2024. The same goes for operating systems, websites, apps, etc. The initial investment cost has become so high, that only state or corporate-level players can compete.
The “digital enclosure” that is created by the high startup cost, as well as entrenchment of existing user bases, means that there is nowhere for users to go, and no way for competitors to enter the market.
This is anecdotal, but there was only one company I interviewed at during my 6 months of unemployment in 2023 that had actually generated profit. It was a company that got people refunds on their property taxes. In other words, the business model was completely built around an inefficiency in the IRS’ appraisal calculator. I repeat, value creation is dead.
In conclusion, every large tech company has expanded its borders to the point where they can no longer grow. They have even grown into each other’s borders as they settle into competitive equilibrium. No new players can enter the field because of startup costs, and the only way to keep profits growing is to turn inward towards their workers and consumers.
Thankfully, this stagnation of capitalism is actually helpful for the development of cooperatives!
What is a Cooperative? What is a Federated Cooperative?
A worker cooperative is a “firm” (company) where every employee is an owner. Decisions are made democratically. Decisions like how much to pay new hires, how to redistribute or reinvest profits, and the election of management. These are usually called “bylaws”.
A federated cooperative is a network of cooperatives who are loosely organized, and can optionally have some centralized body that handles activities that can not be done within a truly distributed system. In the example of the United States, the Federal Government has final say over interstate trade, since this is an intractable problem for states to solve, since they don’t have jurisdiction over each other.
Federated Cooperatives and the potential for disruption
(Note: I am not sure if I explain this part very well, please let me know if this doesn’t make sense and I can edit it)
A company like Uber provides services across the United States. However, all customers and drivers use the same app. Because of the massive scale and relative computational complexity of route planning, the cloud costs for a company like Uber are massive. Nevertheless, Uber finally managed to turn a profit in 2023— only after several years of having a near-monopoly on ride sharing.
Uber is a great model to think about the Federated Coop model for several reasons that we will get into. Hopefully it will please the reader to know that someone has already started a ridesharing cooperative with plans for federation (https://drivers.coop/), and that it has been very successful.
Uber’s services are geographically distributed across the United States. We could think of each city as its own “Uber Island”. Typically a rider will not need to go from one island to another. Because of this, we could break Uber into many geographically isolated Ubers. NYC Uber could be a totally different company than Asheville, NC’s Uber.
Each geographically-isolated Uber could be a cooperative owned by the workers. It would be very inconvenient for users of Uber in different locations to need different apps though. This is where federation comes in. Once one cooperative has built their software, they can share it with another cooperative.
There is no concern about competition, since both cooperatives are geographically isolated. The cooperatives could co-own the code, or create a third cooperative, a developer coop, that is funded by the driver cooperatives. The developer cooperative would maintain the application and the backend, or each driver’s coop could maintain its own backend, with the central organization just pointing to their individual URLs based on your geographic location.
Okay, but this just sounds like dumb communism. Everyone knows that’s not financially viable
I’m not here to argue that worker cooperatives are better for workers. That part is obvious. I am here to argue that worker cooperatives with federation are vastly superior to the current corporate model.
(Disclaimer: I am not a business dude, this is not financial advice. Please correct me if my understanding of finances is incorrect)
Continuing with Uber as our business that we want to grind into dust, let’s look at the breakdown of their revenue on page 54 of their 10-K filing (it’s a little scary but I promise we can get through it together!)
https://www.sec.gov/ix?doc=/Archives/edgar/data/1543151/000154315124000012/uber-20231231.htm
See? It’s really not that bad
This is just a breakdown of how much of their revenue goes to what. 60% goes to “cost of revenue” which is just the cost of doing business. Let’s just say this is fixed for the sake of simplicity.
12% goes to sales and marketing. This makes sense for a publicly-traded company. At some point you hit market saturation, and the only way to avoid shrinking is to constantly be advertising and trying to maintain your presence in the public’s consciousness. Fun fact: this is why every insurance company has like 10 mascots
Remember though that our cooperatives do not have competitive pressure with each other. They have no obligation to deliver growth to shareholders either. They only need to worry about their own salaries and potentially reinvesting into their capital accounts. So a cooperative alternative to Uber already theoretically has 12% higher profit margins.
General and Administrative covers a lot of things that don’t fall into other categories, including management salaries. For the sake of fairness we’ll say it remains the same for a cooperative, although you can make compelling arguments for why a democratic cooperative would have a lower number here.
Research and Development is 8%. During the growth stage of a company, these costs are actually valid. Research and development usually means new features that improve the customer experience. In the Enshittification stage though, research and development means maximizing profit extraction. Developing better algorithms to handle surge pricing calculations. But each of those metrics needs a new database table. Each calculation probably lives in a different service. You need to hire data scientists and big data experts to set up your data lake, which of course has a proprietary license.
I am being hyperbolic but research and development costs at a certain point are to make features that are hostile to users and actively make the service worse. So let’s throw those costs away too.
So 20%. A cooperative in the same business as Uber would have 20% less overhead. This profit could go back in to the cooperative or could be used to pay drivers more. The most important thing here is that the 20% gives the cooperative model a competitive edge in the market.
Why Nobody Uses Mastodon
Ok, people use Mastodon don’t yell at me, but at the very least Mastodon didn’t kill Twitter.
Mastodon didn’t kill Twitter for a few reasons. Being kind of hard to use for the layperson, being too federated, network effects, and most importantly not being better than Twitter-- being so much better than people are willing to make that jump.
For an Uber alternative to succeed it wouldn’t need to be that much better feature-wise, it would need to outcompete Uber in the free market. To get real simple Uber would need to be cheaper for riders, and more profitable for drivers.
This seems like a crazy hurdle right? Like how could a cooperative be simultaneously cheaper than Uber, while paying drivers more? Remember our 20% lower operating costs? A cooperative could spend 5% to subsidize cheaper rides, and another 5% to pay drivers more, while still bringing in 10% profit to the firm.
If you ever talk to your Uber drivers, you’ve probably heard them say that Uber will subsidize rides during the off-season. Uber does have cash reserves and its own stock that it uses to subsidize rides for the benefit of keeping drivers available. This could be used to squash individual cooperatives by subsidizing rides to out-compete the cooperatives.
However, this method would be very capital intensive and could lead to a death spiral if it failed. Example: Uber spends lots of money to crush driver cooperatives, decreasing its profitability. Investors, seeing the margins worsening, flee. Uber’s stock holdings are now worth less, and if Uber wants cash to continue subsidizing its’ rides, it must sell even more stock than before.
Alternatively, if some investor in Uber has already sunk enough money into the company, they may extend Uber a line of credit to help them squash cooperatives. So there’s still risks.
Even with the risks though, I would still argue that worker cooperatives, under certain conditions, can easily out-compete corporations in the market-- and Enshittification makes the conditions for cooperatives even better.
In Conclusion
I could keep adding on to this post but it’s already too long. I think that federated cooperatives could cannibalize quite a few existing tech companies. For sectors that are difficult to disrupt with cooperatives (Social Media, Cloud Computing, Amazon’s distribution networks). I would hope that federated worker cooperatives could pool resources to eat away at bits of market share, and also launch political campaigns to try to nationalize certain things (payment processing, cloud compute, online marketplaces), or at least lobby the absolute shit out of the government to break up monopolies and make them easier to unionize.
Basically I see federated cooperatives as a way to siphon capital to workers, and by extension give actual political power to people whose incentives are aligned with the working class.
This also applies to the games industry. I am too lazy and unqualified to do a financials breakdown of Microsoft’s games division, but it’s very clear from Microsoft’s actions this week that there are MASSIVE inefficiencies in the video game industry as well that are due to enshittification and just capitalism in general. Fun fact: did you know Bobby Kotick (the former CEO of Activision) got a $60 Million dollar bonus before he left amidst sexual harassment allegations? To put that into perspective, that’s the total cost of developing and marketing HALO FUCKING 3.
Federated cooperatives of independent game studios would make it possible for independent devs to make AAA games. Many dev teams are already distributed globally at this point anyways. We can always discuss potential models below in this thread.
I’d also like to say that with all the open source technologies available, the current cheap cost of computing, crazy-ass stuff like Postgres, and the maturity of these open source technologies, I think that cooperatives self-hosting their own back ends is extremely doable. Just keep the number of features reasonable and build something stable not infinitely scalable.
Related Stuff
Genius Investor Warren Buffet figures out how much Californians will pay for chocolate
(Good example of capitalist value “extraction”)
Where does profit come from?
(People still argue about surplus value today don’t come @ me economists)
Democratic Confederalism
(This is about governments, but this paper got me thinking about federated coops to begin with)