First off, Democratic Socialists of America is in the midst of a membership drive. If you’re not a member of a political organization, I encourage you to join one (and the Democratic Party doesn’t count). Voting just isn’t enough when things are this bad. I have been a member of DSA for the past four years, and while it’s by no means a perfect organization, it does have 75,000 dues-paying members and it is winning victories. Feel free to ask me about DSA or any other org–I think being organized is a much more important political decision than who you vote for.
I’ve harped on this before, but it bears repeating: when examining a social question, one of the first tools you should reach for is class analysis. If you are able to ask questions like Is there a conflict between classes at work here? What could I conclude if so?, you are already ahead of most of the mainstream discourse.
If not class analysis, then what other tool might we use? An alternative way to think about a question might be a from first principles approach. In that approach, you define some basic models for actors and decide some axioms that seem reasonable. Then you start searching for conclusions from these axioms and seeing how well they describe the real world. You can get pretty far with this approach, and it does prove useful. You could define a category of worker and a category of owner and see that a worker and an owner have conflicting interests. But you will eventually find the limitations of a first-principles approach and you will need to do additional investigation.
Let’s talk about AB5. This California law, passed in September of 2019, makes it illegal for a business to hire workers as independent contractors when they should be full-time employees unless the worker’s role satisfies a three-part test:
(A) The person is free from the control and direction of the hiring entity in connection with the performance of the work, both under the contract for the performance of the work and in fact.
(B) The person performs work that is outside the usual course of the hiring entity’s business.
(C) The person is customarily engaged in an independently established trade, occupation, or business of the same nature as that involved in the work performed.
Clearly this applies to a company like Uber, as hiring drivers to transport paying passengers is the entirety of their business. But a coalition of gig companies are refusing to comply with this law and have sued the state to prevent the enforcement of AB5.
Prop 22 is a ballot measure funded by that coalition of gig companies that would alter the application of AB5 to gig workers. This relevant section would classify drivers as independent contractors so long as they satisfy a four-part test:
(A) The network company does not unilaterally prescribe specific dates, times of day, or a minimum number of hours during which the app-based driver must be logged into the network company’s online-enabled application or platform
(B) The network company does not require the app-based driver to accept any specific rideshare service or delivery service request as a condition of maintaining access to the network company’s online-enabled application or platform
(C) The network company does not restrict the app-based driver from performing rideshare services or delivery services through other network companies except during engaged time
(D) The network company does not restrict the app-based driver from working in any other lawful occupation or business.
Now, let’s use the first-principles approach. Maybe we can define an employee as someone who sells their labor time to an employer in a rigid way–they have to clock in and clock out and be fully dedicated to their task while they’re on the clock. A contractor, on the other hand, sells the product of their labor. They can work whatever hours they want, so long as the product is produced. These definitions feel pretty good, and if you’re prone to self-assuredness you might think you could use definitions like these to decide how the labor market should naturally function. A gig company is simply paying a worker for their product, a ride given or a meal delivered. The worker clocks in when they want and is, technically, free to take or leave an individual order. Then if you go into the text of AB5, you might see a strange list of professions excluded from the three-part test: direct sales salespersons, real estate licensees, commercial fishermen, workers providing licensed barber or cosmetology services. You could conclude that those exceptions are arbitrary and non-sensical; this law makes no sense.
Now let’s try to contextualize AB5 within a framing of class conflict. Here’s a very brief narrative that I find helpful: As productivity and wages enter their fifth decade of divergence and it becomes clear that the recovery from 2008 only helped the top cut of society, we’re seeing any jobs that do come back return with worse wages and benefits and less stability. The shift to precarious labor is a net step back for gains workers have made over the past century. It’s unclear what to do about this, or if anything can be done under late capitalism, but it is clear that a small handful of people are getting rich off of exploiting this sectoral shift. The app platforms have brought in hundreds of billions of dollars of investment globally to build infrastructure around precarious labor, and it seems clear that if they succeed in winning the monopolies they seek, precarious app-based labor will become a permanent feature of work in America. That’s an unconscionable future—as markets soar most people will struggle at subsistence wages. Stopping this trend is critical. Any (legal) means are justified in stopping the platforms from cementing this future. It would be wise to come up with some definitions of employee and contractor that challenge Uber’s business model so we might slow down this devastating loss of workers’ rights.
I consider this a class-sensitive explanation, whereas the first-principles approach is class-blind. Seeing the loss of power the working class is facing, representatives of the working class advocated for AB5, a law that would slow the backsliding. The class analysis comes first, the action second. So is Prop 22 a conflict between people who use class analysis and people who reason from first-principles? Certainly not. There is too much money at stake for the gig companies to disregard so powerful a tool. Uber doesn’t have some staff philosopher on payroll that defines a first-principles ontology of the business landscape and then orders the CEO to implement it.
What happens at Uber is what happens at any firm under capitalism. The CEO is a manager appointed by the class of shareholders to represent their interests. The interests of Uber’s shareholders overlap strongly with the interests of all owners of capital. The interests of Uber’s drivers are sometimes overlapping and sometimes conflicting with the interests of Uber’s owners, and the drivers’ interests are mostly overlapping with the interests of most other workers. Knowing that his job is to maximize returns on capital, the Uber CEO orders their PR and legal teams to come up with a set of laws and arguments and propagandas that accomplish the goal of minimizing payouts to workers and thus maximizing profits. Uber proposes a set of definitions that interact to produce a set of conclusions that they can stay on the right side of by introducing a variety of abstractions that satisfy the letter of the law.
Uber may not prescribe […] a minimum number of hours during which the app-based driver must work, but its task-assignment algorithm is designed to use monetary incentives to make sure drivers are on the road when they want them to be. Uber may not require the app-based driver to accept any specific rideshare service, but if a driver habitually rejects rides they will be penalized. Drivers are free to drive during non-incentivized hours or to only take certain rides, but the design of Uber’s pay structure guarantees that a driver will not make a living wage by doing so. The incentives end up mattering a lot because average driver pay falls over time in mature markets, as once Uber has won a regional monopoly (or half of a duopoly) they stop subsidizing drivers’ wages in the region. Finally, Uber touts the importance of flexibility for its drivers, but the vast majority of rides sold through Uber are sold by full-time drivers (ed: previously I stated that most drivers were full-time; this is not the case).
By placing the Prop 22 debate in the historical context of the rise of precarious labor, we can see that Uber’s goal is to get essentially the same kind of labor that a full-time employee would provide but to pay them for their output rather than their time. This shifts risk from the class of Uber’s owners onto the class of Uber’s workers. If Uber were to pay an hourly rate but a driver doesn’t get any rides, Uber has lost money. But as it is if a driver sits at the airport for an hour and gets no rides, they have lost their time but Uber has paid nothing. In this class conflict, each side wants to push risk onto the other.
This makes the question of Prop 22 pretty obvious to me. It’s a conflict between workers and owners, and it will have spillover effects that will affect me. If another worker’s wages are lowered and more of their pay is shifted to owners, then the price I can seek for my own labor is lowered as I now have to compete with a worker who makes less. I have a lot more in common with an Uber driver than I do a major Uber shareholder, so it’s clear to me that I will vote no on 22.
I like to use this newsletter as a way to discuss what could be instead of just what is, so I’d like to consider what a “better” gig economy would look like. One defense of the gig economy is that gig companies have found new business models that add value to people’s lives—mobility through rideshares and the convenience of food delivery are quite nice. Today, these companies are unprofitable even with the poor labor practices they embrace, but it’s useful to imagine what would be necessary to create a comparable amount of value without the hyperexploitation.
One model would be a Cooperative Labor Contractor (CLC). It’s kind of like a staffing agency that’s run by its workers. A group of workers would form a CLC which would be owned solely by the workers and managed through democratic means. The CLC then enters into contracts selling various labor-based products to gig economy companies. If I were an entrepreneur with a novel idea for delivering, say, on-demand pharmacy prescription refills, I could contract the CLC to perform the labor-intensive aspects of my business that are core to it. In this model, the entrepreneur owns the brand and the business, but the labor is provided by the CLC. The CLC would pay workers as its own employees and it would be responsible for representing the workers’ interests with the delivery business. If the workers don’t like the incentive structure proposed by the delivery business, they would simply vote against taking the contract (or vote out managers who take unsatisfactory contracts).
This model is better in several ways. It makes innovation easier, actually; rather than managing a massive workforce, an entrepreneur manages the unique aspects of their business and only has to sign one contract with the CLC. With CLCs active, we should expect to see a flourishing of new ways to apply labor. Call centers would hire from CLCs, as would AI companies. The CLC is also less vulnerable to exploiting employees (though not immune to it) since it has democratic oversights built in, whereas today’s gig companies and corporations are totalitarian in nature. We would see actual innovation in gig companies for the first time, as well. Today’s gig companies innovate on new ways to arbitrage around labor regulations; by guaranteeing satisfactory labor practices, companies would have to compete on value rather than exploitation.
We need solidarity-entrepreneurs to start CLCs and we need new regulations that give CLCs legal structures analogous to the variety of legal structures available for legacy capitalist corporations. We also need to pull gig companies off of hyperexploitation and onto CLCs; this is a place where regulation would also be useful. After having seen the inability of gig companies to take care of their workers in times of distress like the pandemic, any politician should see that we need less precarious ways to organize gig labor. Over the next decade, I could see states and municipalities making the operation of non-cooperative gig companies more and more onerous until they are forced to hand over their staffing to worker-directed enterprises like CLCs. Stopping Prop 22 is just a tiny step along the way.