BY MARK JAMISON
Some politicians, academics, and regulatory practitioners are increasingly confident that there is something wrong with the structure of Big Tech. They are equally confident that smart regulators can remold Big Tech and make things better. They are half right.
The European Commission’s (EC) Digital Services Act and Digital Markets Act — which, according to the EC’s website, are about “shaping Europe’s digital future” — are examples of this confidence. Why do these politicians want to remold Big Tech? That can be a long story, but at the core of the answer are two beliefs: first, that US companies (whose services Europeans opt to use) are too popular; and second, that there is something fundamentally wrong with private businesses (which the EC calls “private rule-makers”) making product decisions. Perhaps it is lost on Brussels that its decisions — not Big Tech’s decisions — are keeping Europeans from having an amazing digital future.
The beliefs that Big Tech’s structure is wrong and that regulation should remold markets aren’t confined to Europe. Numerous people advising the incoming Joe Biden administration are advocating that the US attack Big Tech by remaking companies and designing markets. (See here, here, and here.) And the violence at the Capitol last week only intensified the cries for government control and breakup.
But again, these confident advocates for remaking Big Tech are only half right. They are surely right that something is wrong. How do we know? Digital markets are always changing. As a 2019 report commissioned by the UK government titled “Online advertising in the UK” concluded, “The online advertising market is characterized by a fast pace of change in terms of market structure, practices, and competitive dynamics and is likely to evolve significantly over the next 6 to 12 months.” If a structure were “right,” it wouldn’t always change — and change fast.
But these confident advocates are surely wrong that their interventions would make things better. How do we know? In addition to falsely believing that there is a predictably “right” structure, they hold several faulty premises. One is that market structure determines market performance. This line of argument was developed in the 1950s, but since then there has appeared a long line of research demonstrating that market structure emerges from industry economics and from regulation, with more regulation stifling rather than promoting competition.
Another faulty premise is that smart regulators know how businesses and markets should be designed. We have known for decades that this assumption is false. Nobel Prize winners in economics have explained the problems that arise when a government fails to recognize its limited knowledge and the perverse incentives of government control (e.g., Friedrich Hayek and Milton Friedman), the inherent imperfections in regulatory design (e.g., Jean Tirole), and the intrinsic limits of our brains to make complex and well-informed decisions (e.g., Daniel Kahneman, Richard Thaler, and Vernon Smith).
What happens when regulators try to mold markets? Government controls limit the amount of information brought to bear in decision-making and misalign incentives, information, and accountability. Reduced information lowers efficiency and innovation. Misaligned incentives encourage decisions that benefit regulators and powerful stakeholders more than consumers. And regulated markets would adapt slowly to economic and technology changes: As the EC acknowledges, its controls are always trying to play catchup with industry and consumers. So if the regulator engages in ex-ante molding of markets and companies, the regulator will always be a drag on progress.
What would work better for consumers than regulators choosing market structures? Pretty much anything, but an important step would be Big Tech companies being clear, candid, and consistent with customers about their terms of service. Facebook users were caught flatfooted by Cambridge Analytica’s use of Facebook data even though Barack Obama’s campaigns had used even more for years. (See here, here, and here.) Some social media platforms modified their rules on election postings as the 2020 election drew near, making it hard for citizens to develop alternative channels. And Twitter labeled tweets on one side of the election disagreement as “disputed” but not tweets from the other side.
Finally, Big Tech should practice better business management and strategy. YouTube’s demonetization practices seem staff-driven and arbitrary. (See here and here.) Amazon has apparently failed to control employees who, contrary to company policy, used rivals’ data to compete against them. (See here.) And Facebook’s issues often emerge from its business model, not from bad conduct. (See here.)