FOR more than three decades, telecoms policy, at least in rich countries, has been a one-way street: more deregulation and more privatisation in order to foster more competition. This direction was set by America in 1984, when it broke up AT&T, its telephone monopoly. So there was much surprise at a recent memo, written for the White House by an official at the National Security Council, which argued that the next generation of mobile network, “5G” for short, should be built and run by the American government.

The 30-page paper was widely criticised, and quickly dismissed by experts and regulators. Protecting the network from Chinese hacking, the main reason for the proposal, does not require the state to run the entire network. Huawei, a Chinese maker of telecoms gear, is already all but barred from selling its wares to American operators. Government-run broadband would instead stifle competition and increase the risk of overreach by America’s own security agencies.

Yet the memo contains another idea that merits more discussion, and not just in America but elsewhere too. This is the proposal that 5G be rolled out as a national wholesale network that can be used by several service providers, just as some rail networks and electricity grids are.

High five

In the fixed part of the telecoms network—the cables that run underground, say—wholesale networks are already widespread. Under this model, the owners and operators do not also provide the services; these are supplied by separate firms, which share the network and compete with each other. Singapore and New Zealand have this sort of arrangement; so do cities in Sweden. Mobile networks have conventionally been integrated affairs, with operators both managing the network and also providing services (although they do sometimes sublet capacity to others). But sharing does happen. Rwanda has had a wholesale mobile network for some time. Mexico’s Red Compartida is expected to start up soon; it has been built by a private consortium, with the government providing radio spectrum and fibre-optic links to connect the base stations.

The obvious risk of a single wholesale network is that, without the cut and thrust of competition, it ends up acting like the sluggish and underfunded telecoms monopolies of old. Critics point out that the average speed of internet connections on Australia’s government-owned National Broadband Network lags behind that of most rich countries. South Korea, by contrast, has a system of competing broadband networks and some of the zippiest speeds on the planet.

But many people in South Korea live in clusters of residential high-rise buildings, which are easily wired up. For farther-flung networks, particularly in rural areas, the costs are higher. And 5G networks will anyway be more expensive to build than their forebears. They will eventually use higher-frequency radio waves, which cannot penetrate buildings and other obstacles; that means they will need more base stations and antennae. If every provider has to build its own 5G network, costs will be unnecessarily high—sometimes prohibitively so.

A single shared network would be cheaper. It could also increase competition for the services on top of it. Next-generation networks are supposed to become the connective tissue for all sorts of devices, from sensors to medical equipment (see article). If firms can lease capacity to create such networks without having to build them, in much the same way as firms use smartphones and app stores to reach consumers, the “internet of things” will be more vibrant. This kind of environment would also ease worries about the end of “net neutrality” in America. If one company discriminates against certain online content, consumers can switch to another.

Governments do not have to go as far as mandating the creation of a wholesale network, as Mexico has done, to get some of the benefits from sharing. Many states in America restrict or even ban municipalities from building networks. Eliminating such laws would be an obvious start. Regulators can also encourage other forms of sharing, for instance of spectrum, something America has started to experiment with. The White House’s 5G memo is an unlikely milestone on the path towards more sharing. But in questioning the need for a lot of competing networks, it is sending the right signal.