This post is part of a symposium on root and branch reconstruction in antitrust. Read the rest of the symposium here.
Sanjukta Paul (here and here), Marina Lao, Hiba Hafiz, and Marshall Steinbaum all have written about the antitrust issues raised by contract workers forming associations that would negotiate collectively with the firms that purchase their services.
This post builds on their work by evaluating whether antitrust joint ventures doctrine would permit contract workers to engage in coordination. Properly designed, an association could pass muster under current Sherman Act antitrust jurisprudence of BMI, NCAA and Dagher. To do so, a worker-owned firm or cooperative association would need to structure itself in a way that creates efficient integration, rather than naked price fixing. If it is able to provide a more efficient, higher quality product, then it could achieve higher worker incomes without engaging in anticompetitive conduct. However, this may prove cold comfort, because an association will likely face serious start-up problems — as well as litigation costs defending the association — that could prevent its success.
An Efficient Association
To illustrate the idea in a concrete way, consider the example of the drivers who provide services to ride sharing firms like Uber and Lyft, and delivery services like Door Dash and Amazon. These drivers are treated as independent contractors rather than employees and are unable to unionize. To satisfy BMI and FTC cases, the association needs to do more than negotiate payment terms. It also needs to create efficiency-enhancing integration. It also must escape NCAA, which permitted the NCAA to set the rules of football but condemned their joint negotiation of broadcast fees. This integration would also allow the association to satisfy the Agencies’ Competitor Collaboration Guidelines.
How might an association create efficiency-enhancing integration? The association could screen drivers for criminal and driving records and administer personality tests and other metrics to find drivers that would provide good service, periodically inspect drivers’ vehicles for safety and quality, and have “secret shoppers” check service. These services reduce transactions costs for the firms who are the customers of the association and are not services individual drivers could efficiently provide. The association could provide liability insurance to the customer firms for claims made against the drivers who are members of the association. Because of the inspections and screening, the association would be an efficient insurer.
The association could argue that it creates a different and superior product than any of the individual driver-members can, as required by BMI. It also can do better than the client firms themselves. While Uber and Lyft do some screening, the association would do more. Because the drivers are owners, they also would have incentives to monitor each other, which reduces monitoring costs and increases efficiency. In addition, if the association grows and serves multiple companies, it can keep drivers busier by better accounting for the demand peaks and valleys of its diverse client base. By assigning drivers in a coordinated way, it can better match supply to demand, thereby providing client firms greater reliability of a sufficient number of drivers.
BMI members were permitted to engage in direct dealing with licensees outside of the association, a key fact for the Court. However, non-exclusivity creates a free rider problem for the driver association. Once the association screens and certifies drivers, client and non-client firms might free ride by hiring them independently. This is a standard efficiency justification for exclusivity. Members would be permitted to leave, of course, though the association might create reasonable switching costs to deter free riding, for example, by withholding a small fraction of fees owed.
The association would want to set the payment terms, rather than each driver setting its own fees. The NCAA Court said that the Colleges (or the individual conferences) should negotiate with the broadcast networks individually rather than NCAA setting prices. Law firm partners, doctors in joint medical practices, and economists in some consulting firms do not set prices independently, so it is not clear why the association should be treated differently. Nor do many labor-outsourcing firms. The association can explain that it is like these labor-outsourcing firms, but simply with workers having ownership stakes. Indeed, it is not clear why those firms should be permitted to price jointly, if the association is not. If this argument fails, the association might replace worker ownership with bonuses that have the same economic effects and incentives.
Suppose that the association grows and driver fees do rise. The association can show that while clients are paying higher nominal fees (e.g., per hour or per ride), the higher quality service would lead to a lower quality-adjusted fees. Low-cost insurance and other benefits also raise drivers’ real incomes without necessarily impacting downstream prices. By giving the drivers a greater stake in the enterprise, even aside from higher income, worker satisfaction could rise.
Joint fee setting is easier to justify if the association lacks market power. A new entrant association would not have market power. In addition, the association would be competing with clients themselves in attracting drivers independently. If the association grows large, it would argue that its success was the result of a better, more valuable product — that is, that it gained market power obtained “on the merits.” However, as it adds driver-owners, the Agencies could raise merger or exclusive dealing issues.
Even if an association escapes antitrust attacks, significant practical impediments could hinder its likelihood of success. The start-up would involve substantial sunk costs to develop the testing and monitoring protocols and services and promotion of the association, as well as litigation costs if its formation is investigated and attacked. The driver-owners are unlikely to have the savings to finance the start-up or bear the risk of failure. The association might borrow or bring in a venture capitalist as a partial owner. This latter approach creates some irony, of course. And it also leads to potential conflicts of interest if the capitalist has substantial control.
An existing labor union might sponsor or finance the association. Unions have experience and expertise, as well as the financial reserves. For example, the Teamsters have taken a supportive role for drivers in Seattle, albeit currently under a contract with the city. Why haven’t unions been more aggressive in creating such associations? One possibility is that the unions are pushing to have drivers classified as employees, where they can obtain labor law benefits and unionize.
Another pitfall is potential conflicts of interest between the association managers and the drivers themselves. The managers may have incentives to take actions that would drive association growth – which could justify higher salaries — even if that growth would not benefit the workers. It would be ironic for a worker-owned association to oppress the workers. But, in fact, the Dairy Farmers of America agricultural coop has been accused of abusive conduct towards its members.
Why Not Compete?
Rather than simply supplying driver services, the association might compete by creating its own retail service. This would create a worker-owned firm instead of placing workers in an arguably inferior position in the hierarchy, as discussed by Sanjukta Paul. But this vertical integration would be more complex, more expensive, and would meet even more resistance from the established firms.
If the association takes this approach, it also raises the question of whether the association could set the retail prices, or whether each driver would set its own prices. Again, the analogy to law firm and medical practice partners should apply. However, if independent fee setting were required under NCAA, the association might need to run an Amazon-like auction in which the drivers bid for rides. The association would need to tack on fees, which it distributes to driver-owners as dividends, in order to avoid lower worker incomes.
All things considered, such an association of independent contractors could succeed if it had a dedicated and charismatic leader and sufficient legal and financial backing. But becoming classified as employees with the right to unionize would be better since that status already comes with a set of rights and the power to negotiate with employers. This route has political support. The Biden administration withdrew the Trump administration’s Independent Contractor Rule in 2021, though a Texas judge reinstated it in March 2022. In December 2021, the NLRB invited comments whether to require such workers to be classified as employees.
Of course, these rules will not cover all such contract service worker. But for some workers, the integration provided by these associations might allow them to improve their working conditions, provide a better product to consumers, and collaborate on negotiating payment terms while escaping the antitrust per se rule against price fixing.