With the potential for passage of the America Innovation and Choice Online Act (AICOA) and discussion of self-preferencing among Big-Tech platforms ought to come discussion of remedies for that self-preference. Apart from damages and injunctive relief, AICOA is small on discussion about what to do with repeat offenders. Indeed, while other aspects of the AICOA are important and promising, there is not much guidance for courts in the remedies provision, aside from generalist language about the court being able to enter relief as necessary to prevent and restrain the activity.
With or without AICOA, antitrust enforcers have means to eliminate bottleneck monopolies that enable self-preferencing. Bottleneck monopolies lie somewhere along the chain of production and distribution of goods or services. Usually both before and after the bottleneck, the markets are workably competitive. Although there may be many producers of basic inputs, processed in another workably competitive market, the products must pass through the essential facility to reach the final consumer. The monopoly then intervenes in the otherwise workably competitive sequence of markets and confers on its possessor control over access from upstream sources to downstream distribution or consumers. Such a monopoly can exact a substantial amount of wealth by imposing a monopoly charge on those seeking to use its essential facility.
We address the traditional means of curtailing monopoly power of bottleneck owners and their flaws. We propose two potential alternatives that could be better suited to modern day bottlenecks.
The Traditional Problem of Bottlenecks: Self-Preferencing
The problem of vertical integration has been present in virtually every regulated industry. Indeed, the principal reason for regulation has been to assure that a vertically integrated monopoly does not use its monopoly power to create prices that are out of the realm of competition or not in the public interest.
A monopolist in control over the essential facility will generally have an incentive to raise price and restrict output. However, a monopolist that controls the essential facility but competes in the market has tremendous incentive to capture shares of the downstream market.
Regulators have employed two traditional means of “eliminating” self-preferencing: (1) the “Cleaver” and (2) the “Watcher.”
Structural attempts to remove self-preferencing have led to remedies that sought to sharply cleave the ability and the incentive of holders of essential facilities. Bifurcation of ownership provides, in theory, a neatly carved and observable remedy.
But the problems with such a remedy are threefold. First, as AT&T demonstrates, what was broken up can reunite. Second, cleavers have a way of multiplying the problem rather than remedying it. For example, in DOJ’s initial proposed Microsoft remedy, two monopolies would have been created in the proposed remedy by the DOJ, one over the operating system and one over applications. Third, some bottlenecks are not so easily cleavable.
Conduct requirements that regulate control but not ownership have the potential for remedying the issue, in theory. As an example, when electricity dispatch is done by a vertically integrated firm, its incentives are to exclude independent generators despite regulatory rules that deter such conduct.
But regulatory solutions require the combination of oversight and penalty sufficient to deter future conduct. In many instances in the past, the penalty has merely been to disgorge ill-gotten gains. In prior instances where this has been the remedy, “crime pays.”
Also, regulatory solutions in antitrust are frequently not permanent. To the extent that there is insufficient continued progress in the market toward competition, elimination of the decree can be perilous.
Some Modest Alternatives
If the ownership and control of the bottleneck is redesigned to curtail the incentives to exclude or discriminate, this changes the goals of the operator or owner of capacity. In this revised world, the owner of unused capacity has every incentive to sell it to the highest bidder because it can gain little or nothing by withholding a small piece of the capacity. Similarly, if many owners control a bottleneck that is expandable, they will have the incentive to expand its capacity because they gain little or nothing by restricting capacity. Their gains come from lower cost, easier access to their downstream or upstream markets.
Thus, redefining the ownership and control of the bottleneck is central to harnessing the ordinary incentives of the market process to induce conduct that approximates what would have happened in a functionally competitive market. Yet some bottlenecks involve situations where the users would be best off to have participation in the decisions regarding the structure and operation of the bottleneck itself. These are situations in which shared governance of some kind is important. Such situations call for a collective ownership of the facility involved. We have labeled this the “cooperative” model.
In contrast, in other contexts the operation of the bottleneck is relatively distinct from the activity of its users. In these situations, the central incentive rests in the ownership or control of the capacity of the bottleneck itself. If no one owns (controls) much of that capacity and the rights of use are transferable, then those with initial ownership will have the incentive to sell or assign the use right whenever the price (value) exceeds their own value for the right. This will not eliminate “monopoly” prices at times of scarcity, but the use rights will be distributed based on the value to the buyer-user and not for longer run strategic reasons. We refer to this remedy as the “condominium” solution to the bottleneck problem.
Alternative 1: The Cooperative Model
This model envisions shared user ownership of the bottleneck. Such transformation of ownership should change the incentives governing the operation and potential expansion of the bottleneck.
The individual owner-user stands to gain little by trying to impose a monopoly price on users including itself or by restricting access to the bottleneck by new entrants. So long as there are many owners, the primary objective should be to manage the entity so that it operates efficiently and with as much capacity as possible. The owner’s gains come from their upstream or downstream commercial activity. The bottleneck is a transaction cost and potential constraint.
An example includes grain elevators in the early 1900s. When it was very costly and time consuming for farmers to move grain very far from the farm to a grain elevator on a railroad line, competition among rural elevators was limited. Any region was likely to produce only enough grain on average to support one or two elevators. In only a few areas did farmers have access to as many as three elevators. This conferred on the elevators, especially when only one or two served the area, considerable power to exploit farmers.
Farmers, with help from some grain trading firms, resolved the issue by establishing cooperative elevators. These elevators had on average greater storage capacity than either the vertically integrated elevators or the independent elevators. The results were that farmer revenue went up because the monopsonistic buying practices could be circumvented. The cooperative elevator had sufficient grain that it could fill rail cars and ship to several alternative markets. Thus, the monopsony bottleneck was eliminated.
This and other examples involve enterprises that require substantial continued engagement of the participants in the governance of the enterprise. With such shared governance, the enterprise will be developed and operated with the objective of serving the interest of all participants.
Two characteristics are significant to determining whether a cooperative model for addressing bottleneck power is appropriate. First, when the structure and operation of the bottleneck is complex and interdependent with the users’ activities, the users are likely to find it very important to be directly engaged in the decisions about those issues. Second, where the size of the bottleneck can vary substantially from highly restrictive to very open, again the users’ interest is to be engaged with the decision about scale and that in turn is best implemented by participating in the governance of the enterprise.
The more the bottleneck interacts directly with other aspects of the users’ or suppliers’ activity, the more those parties will benefit from active involvement in the decisions about the nature and scope of the activity. Examples could include an operating system, an electric transmission system, or social media platform. In each, there are a myriad of choices to be made about design and/or location. Different stakeholders will have different needs and desires.
For example,Twitter is currently discussed in terms of singular incentives that fail both users, advertisers, employees and other stakeholders. A cooperative model would require input from users as to changes that devolve the platform (paid checkmarks that allow for impersonation, for example.)
This method requires that no party or group dominates the decision process and all parties recognize their mutual need to make the bottleneck as effective as possible for all users. Enhancing use is a shared goal, and the competing experiences and needs could be negotiated without unilateral action that could devalue the platform.
Alternative 2: The Condominium Model
The condominium model is one in which a distinct entity administers the “facility” in which the occupants “own” their specific units. Examples of such structures include the current rights to capacity on natural gas pipelines, rights to space on container ships, and administration for standard essential patents. These examples all involve situations in which the user has a right to use some capacity or right but the administration of the system rests with a distinct party rather than the kind of cooperative arrangement discussed earlier. In a full condominium analogy, the owners of the units would have the right to terminate the manager and replace it. Thus, if there are several potential managers, the prices of the managerial service would ideally be set by the market.
The central difference between a condominium model and the cooperative one is that the management of the bottleneck is readily separable from the uses being made. Thus, a pipeline manager can operate the pipe while individual rights holders deliver gas to be transported. Such separability between operation and use often involves a facility with fixed capacity such as the pipeline or rail line. The manager can coordinate the use and the owners of access rights are able to use the facility as needed. But this is also a feature of some other more open-ended bottlenecks especially those involving patents or copyrights. There the user is interested in having the right to use the IP, but the management of the licensing process particularly when it involves many potential users is quite distinct. Thus, BMI and ASCAP provide administrative services that link licensees to rights owners beyond anything that any pair of users and owners might accomplish.
Another feature of this model is that it implies that the rights of use/access when constrained will be tradable. Much as a condo owner may elect to rent the condo to someone who values it more. Scarcity creates the potential for discriminatory exploitation when a single monopolist holds those rights. By distributing access rights to many owners, the opportunity for discriminatory or exclusionary conduct is removed and the owner has only the opportunity to earn rents (high prices) from the sale or lease of its capacity entitlement. This dispersion of interests results in a clear change in the incentives of a rights holder. Thus, the finite capacity of the bottleneck should be allocated to the highest and best users as measured by their willingness to pay.
Potential applications of this model in addition to the examples already given include such disparate bottlenecks as trash dumps, airport gates, or an online platform for the sale of goods. Trash dumps provide an interesting parallel to some modern issues. The barriers to entry into trash collecting are low. But there are very few locations for disposing of trash. Private ownership of those facilities confers substantial power on those trash haulers that are vertically integrated. But it would not be hard to require that dumping rights be separated from ownership and operation of the dump. In that case, each rights holder would have the entitlement to deliver a set quantity of trash each year. Provided that such entitlements are sufficiently dispersed, no owner would have an incentive to try to exclude competition but would be willing to sell or lease its rights if other potential users were willing to pay a sufficient premium.
Human genome research could also work under a condo model. The operator would collect and publish data. Such an endeavor would potentially enhance research. Patent pools provide another example. For example, In Re Ciba-Geigy, the FTC required compulsory licensing of certain IP to competitors. An easier remedy would be to have an independent controller with a goal to maximize research.
In the context of Amazon, a controller could optimize product placement, with premium placement occurring for those who value the placement the most, rather than based on who owns the platform. Thus the controller would have as their goal to maximize the value of the platform.
We do not present these solutions as panacea. We anticipate some initial difficulties in implementation. And they may not work for every situation. However, they offer a decidedly better chance of achieving good overall results compared to traditional remedies.
Stakeholders will have incentives to monitor and call out transgressions that go against their interests. This should reduce monitoring costs. Because of the threat of calling out violations, remedying any transgression ought to be more efficient. Abuse would potentially be remedied without any intervention that could draw resources. Moreover, such a remedy would potentially be durable beyond a decree period. However, we propose that such a decree be presumptively renewable with review every five years. And ownership that goes wayward could be “rebooted.” Distortions in the balance of ownership could be reset in the decree review process.
In sum, there is potential for our proposed remedies to prove workable. Highly contentious markets with multiple diverse stakeholders have great potential for success.