CTIC Research Fellow Giovanna Massarotto argues that computer science can help regulators monitor markets and identify algorithmic collusion.
At ProMarket, Giovanna Massarotto, Research Fellow at the Center for Technology, Innovation & Technology (CTIC) explores how regulators can better enforce antitrust laws in computer-run digital markets by incorporating lessons from computer science to combat algorithmic collusion.
Massarotto’s scholarship focuses on how technology affects society and the intersection of law, economics, and computer science.
She is an active scholar and author of Antitrust Settlements: How a Sample Agreement Can Drive the Economy, published by Wolters Kluwer, and of papers in many prestigious journals, including the Journal of Competition Law & Economics and John Marshall Review of Intellectual Property Law.
Enforcing competition principles in computer-run digital markets became critical in 2023. This past year saw the widespread use of algorithms, including artificial intelligence, in many facets of our lives. In business, algorithms have become a necessary tool to remain competitive. With their deployment comes concerns about potentially anticompetitive conduct by firms, such as algorithmic collusion to fix prices or allocate markets.
Many legal scholars and legislators are proposing to use ex ante regulation to deal with digital markets (regulation that imposes precise prohibitions and obligations by default, such as the Digital Markets Act) instead of ex post antitrust intervention (which implies an investigation and a response to an infraction, such as a fine or a consent decree). However, ex ante regulation has been unsuccessful in the past, even in less dynamic physical markets, such as electricity and transportation. For instance, George J. Stigler and Claire Friedland’s 1962 study revealed “[t]he ineffectiveness of regulation” in electricity markets, mainly because the “regulatory body [was] incapable of forcing the utility to operate at a specified combination of output, price, and cost.” The capability of the government to regulate ex ante has become even more challenging in fast-moving technological markets run by increasingly sophisticated algorithms. Thus, I proposesomething different, which is to incorporate lessons from computer science into our current set of law-and-economics tools for the analysis of antitrust violations ex post. Otherwise, the interpretation of how computers can engage in anticompetitive behavior remains incomplete, if not impossible.
To validate my proposal, I connect the collusion problems we deal with in law to problems of agreement on a shared value in computer networks. In my working paper, “Using Computer Science to Detect Cheat Tolerant Cartels,” I use the Byzantine Generals Problem (BGP) employed in computer science as an analogy for the cartel-cheating problem relevant in antitrust law… .