Forty years after Ronald Reagan, the antitrust pendulum has swung back. So far most of the discussion has focused on whether Facebook and other tech giants deserve “scrutiny.”

Social media brands But what then? Simply replacing Facebook with a new monopolist would do nothing to make consumers better off. The real question should instead be whether we can design relief that increases competition going forward.

The usual Sherman Act choices are limited. Money damages are often uncertain, unprovable, and may never be asserted at all. This undercuts deterrence. Antitrust law has long recognized this defect and tries to compensate by awarding three times actual damage. However this trebling is itself a guess, simultaneously guaranteeing that some monopolists will be undeterred while driving others out of the market entirely. Injunctions are similarly limited.Because criminal sanctions must be narrow and precise, rapidly evolving digital markets quickly outgrow them.

No wonder then, if antitrust lawyers are intrigued by “structural relief” proposals like breaking up existing giants or blocking mergers that would create new ones. But these remedies also face challenges in the digital world. Social media, after all, derive their value from letting users interact. This means that consumers often choose platforms simply because they are popular. Suppose, then, that the government broke Facebook into two competing services. That would surely increase competition, and this might translate into lower ad prices or more free software. But the new platforms would also have smaller audiences. Would consumers really prefer this trade?

There is a better way. More than a century ago the Supreme Court famously ordered the breakup of Standard Oil (1911). The surprise came the following year in the Terminal Railway case, where a monopolist bought up the only three bridges over the Mississippi and used them to limit railroad competition. Following Standard Oil , the Court could have split the company in three. But that would have added just two competitors, and still left all three bridges badly underused. In fact, the Court did something much cleverer, telling the offender that it could keep its facilities – but only if it opened them to any competitor who agreed to share operating expenses. Compared to divestiture, the new system delivered more competition and less downtime for the bridges. What’s not to like?

The question today is whether the same thing can be done for social media platforms. In some ways the digital economy has simplified the problem. Physical assets like bridges, after all, get congested and need maintenance. This requires complicated agreements which members can covertly manipulate to raise prices. By comparison there is no limit to the number of consumers who can share software.

Still, there is a problem. Monopoly, it turns out, is not entirely bad. Where the Sherman Act sees a threat, intellectual property law sees incentives to innovate. At least in theory, Terminal Railway relief could deliver so much competition that firms were forced to shrink their R&D investments. At the same time, we know from history that many firms – for example IBM in personal computers and Nintendo for games – have deliberately chosen open standards to attract customers. This shows that innovation and sharing are often compatible. The good news here is that Facebook is so insanely profitable. Regulators should take comfort in the fact that early 20th Century interventions into similarly lucrative industries like automobiles and airplanes had little or no impact on R&D.

What would sharing look like? Probably the most traditional Terminal Railway solution would be to pool key data (e.g. User posts, “Followers” and “Likes”) in a central depository open to any company that agrees to make reciprocal disclosures. Consumers would then set up home pages with whatever company offered the most added value in the form of software tools, newsfeeds, and (ideally) enlightened privacy and censorship policies. Not surprisingly, the hardest questions would involve innovation, most obviously when some depository member decided to change an existing data category or create new ones. Here, the usual 20th Century impulse of creating a nominally independent central management body to approve changes seems hopelessly restrictive. Better to let every member add whatever categories it likes to the depository, and then let competitors decide which ones their own products should build on.

Americans are understandably angry at Big Tech. All the same, politicians should deliver more than pandering and violent rhetoric. Journalists and voters should press firebrands to explain just how their proposals will improve life for consumers. If nothing else, this will help citizens separate statesmen from demagogues going into the 2020 elections.