While the use of algorithms can undoubtedly generate significant procompetitive benefits, this is not always the case, including when mergers and acquisitions are involved. Yet to date, the effects of algorithms on merger policy have been undertreated, both in the academic literature and in merger policy. Furthermore, while some scholars have acknowledged the role that algorithms play in merger policy, most studies relate only to algorithmic coordination. Given recent learnings on the myriad ways that algorithms affect unilateral and coordinated anticompetitive conduct, a more informed approach to mergers that involves algorithms should be adopted.
This article seeks to systematically analyze the effects of algorithms on merger control, focusing mainly on their use by market players. To do so, we first explore the potential effects of algorithms—especially those powered by AI—on both unilateral and multilateral market power. The analysis includes both the standalone and cumulative effects of the different types of welfare-reducing conduct, which has become easier to engage in due to the use of algorithms. We then translate these effects into recommendations for merger policy.
To do so, we first identify six main functions of algorithms that may affect market dynamics and merger control: (1) collecting and ordering data; (2) improving the ability to use existing data; (3) reducing the need for data, for instance by generating synthetic data; (4) monitoring; (5) predicting outcomes to determine how different types of conduct are likely to affect market conditions; and (6) making decisions, for example by determining trade terms or product/service variables.
We then set the stage for merger analysis by pointing to the potential effects of algorithms on consumer welfare. We show that, while algorithms offer many benefits that may increase both firms’ and consumers’ welfare, they may also exacerbate anticompetitive conduct. Toward this end, we explore the effects of algorithms in eight scenarios: collusion, conscious parallelism, high unilateral prices, price discrimination, predation, selective pricing, tying, and reduced interoperability. In each scenario, we analyze how the market conditions necessary for such conduct are affected by algorithms. We also offer examples, both real and theoretical, to show that algorithms may not only increase a firm’s market power but also amplify its ability to create and enjoy the benefits of market power.
These findings are then translated into merger policy, which must balance the benefits of algorithms with their potentially harmful effects. The article explores the effects of algorithms on procedural, substantive, and institutional features of merger control. Given that the effects of algorithms on market dynamics are still being studied, it is too early to make across-the-board recommendations. Nonetheless, the time is ripe for antitrust authorities to utilize merger control tools that are sensitive to the effects of algorithms. Accordingly, we suggest changes relating to merger control.
The article proceeds as follows: Part I briefly describes different roles algorithms play in the marketplace. Parts II and III analyze the competitive concerns raised by the use of algorithms on unilateral and coordinated conduct, respectively. We also suggest how these concerns should affect merger policy. Part IV analyzes the use of AI by merger regulators.
Continue reading the full text of this article in PDF format.
