New rules for regulating big tech

Breaking up big tech has been a topic of conversation in the media and in politics for years. But as is often the case with big, complex, global issues, despite all the conversations and potential pathways, there isn’t always a clear answer. Unintended consequences abound and the potential political ramifications should one get it wrong has resulted in little progress. Economist Jean Tirole won the Nobel Prize in Economic Sciences in 2014 for his work on the analysis of market power and regulation. He has studied monopolies, oligopolies, and their impact on innovation and growth. Tirole looks to a, perhaps, unlikely industry that could provide guidance when it comes to writing the rules of how to regulate big tech.

A monopoly is when no competitors exist, a duopoly is competition between two firms, and an oligopoly is a market structure with a small number of firms competing. True monopolies, while rare, do exist and Tirole says that looking at utilities and transportation providers are perfect examples.

“There are lots of firms which are monopolies or tight oligopolies so they can basically raise a price or lower quality without losing customers,” says Tirole. “The reason for that is that they don't have many competitors. They have high fixed costs. For example, if you are a train company it's very hard to duplicate tracks and stations. If you are an electricity generator you cannot build your own high voltage transmission grid.”

What Tirole finds most fascinating with these types of companies and markets is that they often become quite complaisant.

“It's very interesting how monopolies live a very quiet life and in the end they don't innovate very much,” he says. “If you have competitors, you’re kept on your toes and you really have to innovate, otherwise some new firms will come in with a much better technology and replace you. It's very hard to incentivize a monopolist to actually change. That's why often you need competition, you need new entrants who are going to create breakthrough products so as to replace a monopolist.”

In the 1980s, Tirole introduced game theory, typically used in conflict situations, into organizational structures to better understand non-competitive markets. Markets are imperfect for many reasons, ranging from market power to asymmetric information, and externalities. For regulators of any industry, they often don’t have the same level of access to information as the firms that they are trying to regulate do.

“It's very important to reduce asymmetry of information for one thing, but it's also very important to actually realize areas of asymmetric information and act upon it to change your regulation and that's a little bit what we did, with Jean-Jacques Laffont in the 80s,” he says. “We were trying to anticipate on the future reforms that were going to take place in telecoms and railroads and electricity and tried to make sure that those reforms will be done in the right way. You can benefit the public interest by keeping prices low or keeping subsidies low, what's called incentive regulation, or preference-based regulation.”

Tirole and Laffont did this by trying to incentivize firms by making them more accountable for their performance. If they reduced their costs, they would keep some of that cost reduction for themselves and then share it with the customer or client. They also examined the limits of this, but ultimately found that you could in fact incentivize firms to do better and be more efficient and that the ultimate advantage to the consumer would be lower prices. They also looked at ways to introduce competition.

“Take electricity for example,” says Tirole. “You cannot have two or three high voltage transmission grids, that would be too difficult to build, too expensive, and not very beautiful for the environment either. But you can have multiple production facilities that compete amongst themselves.”

Tirole’s early work, while it focused more on utilities and natural monopolies, laid the groundwork for how regulation of big tech may be approached. But how do you regulate something while it’s still evolving? Tirole says he isn’t against the idea of breaking up big tech but admits that it’s a much more complex landscape, though he does see several solutions.

“So first, we already have antitrust laws,” he says. “There are actually two kinds of antitrust laws. The first is about monopolization, or basically some kind of collusion between firms in the market to try to raise prices. The other is acquisitions. One of the issues we see now is firms buying their competitors or potential future competitors. That’s what’s called the abuse of a dominant position.”

Tirole says that while the first option would be an attempt to go back to the old style of regulation, he doesn’t see that as a feasible option because of the global nature of tech companies.

“It's harder for a country to regulate a global firm, it's almost impossible,” he says. “So my view is that antitrust is the main game in town but it has to be altered in many ways. The first is that it has to be more forward looking. A number of reports have proposed that information be collected, so as to be able to intervene faster and possibly stop firms from certain practices. I think that’s a good idea.”

This would entail requiring companies to notify a regulatory body or antitrust agency of any acquisitions they intend to make. The antitrust agency would then have the authority to deny any potentially problematic acquisitions. This would require a burden of proof, making it a different regulatory regime than what we’ve seen historically.

When it comes to big tech companies, Tirole is not surprised that there’s an inherent tendency towards monopolization. When everyone uses the same platform, they can communicate, interact, and engage with each other. People want to be where other people are, he says. The more network effect that’s created can also improve the product overall. More data equals more efficiency. Ensuring that data is used properly, ethically, and responsibly, however is key.

“We have what's called data for service arrangements, so we don't pay for all those wonderful services,” he says. “We give our data and those data are used for two things. The first is targeted advertising. And the second is that those data are going to be used for new services.”

“You might say where is the competitive harm?” he continues. “After all the consumer pays zero and they get these great services, so do we care about monopoly? And the answer is yes, we do care about monopoly because the advertisers pay a lot, which raise the cost of doing business and raise our prices, so a consumer was going to pay in the end indirectly because the advertisers are going to pay more.”

In general, Tirole says that the main worry is that data will be used the right way, and one company controlling too much of the data makes it hard to oversee. Sometimes data collection is useful, and other times it can risk becoming exploitative.

“The danger is not only the emergence of very powerful platforms which can not only charge high prices but also infringe on our privacy, engage in politics, and things like that,” says Tirole. “There is the issue of the future of labor, there is the issue of inequality, both domestic inequality and international inequality. It all depends on regulation. If we do things right, we’ll be much healthier.”