Trump(et)ing the terrors of nationalized and regulated 5G networks

The latest craze to strike the world of telecommunications policy is advocacy for the construction of centralized, nationalized 5G mobile networks. Although they faced significant opposition, US media reported recently that the idea had at least been investigated by the Trump administration. However, it is far from new.

Last year Chorus, the firm contracted to build most of New Zealand’s centralized, all-but-nationalized nationwide fiber broadband network, floated the idea that it should be given the exclusive right to build the country’s sole 5G network and sell connections to structurally separate retailers who would then sell services to consumers. The same concept also underpins South Africa’s latest proposed Telecommunications Act amendments.

An appealing appeal

For those with scant economics backgrounds (or perhaps a telco with a grand strategy to appeal to politicians for regulation to reduce its competition — and even better, to get a free ride on the financing if government funding or spectrum gifts are forthcoming), the idea seems quite attractive. As Tim Wu exhorted last week:

A national 5G network would be a kind of 21st-century Tennessee Valley Authority. The government would build or lease the towers . . . and set up a public utility that sold bandwidth at cost. . . . cheap bandwidth would be made available for resale by anyone who wanted to provide home broadband or wireless, thus creating a new business model for small local resellers.

Wu did place a couple of caveats on his plan — that it should not be done by seizing the assets of existing private providers, and because governments tend to be bad at managing telecommunications networks, it should be government-owned for its first decade or so and then sold off to the highest bidder.

A deeply flawed model

The problem is that the long history of economics research into this business model — used by most of Western Europe, Scandinavia, the UK, and British Commonwealth nations for the deployment and delivery of fixed line telecommunications services — is that it is deeply flawed in terms of the incentives it provides for dynamic efficiency.

There is no doubt that foreclosing competition in the construction of a new network does lead to lower costs in the short run. There is no duplication of the assets, so static efficiency is maximized. However, there is a risk that without competition, the network built tends to be a gold-plated playpen for engineers, allowing them to experiment with the latest technology and charge consumers for the privilege, even though consumers may not place the same value on the bells and whistles.

The real problem however comes in the long run, because the regulatory model used to keep the stack of dominoes upright tends to delay the deployment of the next generation of technology and makes it almost certain that government subsidies will be required to get it built. Consumers get cheap prices, but in exchange for forgoing the benefits that come from the deployment of an even better technology (i.e., 6G) in the future.

Counting the real costs

The key lies in determining the “cost” at which the services are resold. In economics, this is not the accounting artifact reflecting the initial investment made in the past but a price reflecting the current opportunity cost of the investment — that is, what the network is worth today. And what it is worth today is measured by what it would cost to replace it using today’s technology.

The problem is that having sunk a large amount of money into the 5G network, the network operator fears the asset becoming “stranded” if a newer, better, and possibly cheaper technology that customers value more comes along. Because regulators don’t like the inefficiencies arising from stranded assets, the regulated price at which retailers get to buy the resold services on the old network is periodically reset to reflect the opportunity cost of the old network, which is determined by the cost of building a new network using today’s technology. If this is less than the cost of building the old network (and in telecommunications this has been the pattern observed for a long time), then regulated prices go down.

Given the changes in costs, rivals could enter the market by building the new network, providing real product differentiation for consumers but stranding the incumbent’s assets. If, however, rivals can buy access to the old network at a long-run incremental price that is equivalent to what they would have paid if building their own network, then asset stranding is avoided. Without facing any of the commercial risk that comes from making their own investment, rivals have been given a “free option.” They can resell the incumbent’s services and thereby avoid the same risk that they too in the future might face competition and asset stranding from a rival building a newer, cheaper, and more capable network. Asset stranding is avoided, and consumers get to pay the cheaper prices for the same old services. But the dynamic efficiency gains that would have come from the newer and better services (had the new network been built) are foregone.

Striking an (im)balance

However, the problem gets worse. As the incumbent’s regulated price keeps decreasing to reflect the changing costs of new networks, it will never recover the costs of its initial capital investment. So when the network comes to the end of its natural life, the incumbent cannot invest in new infrastructure. Furthermore, even though the rivals could invest, they won’t because they have observed what price regulation did to the incumbent’s business case. The end result is an “investment strike.” Nobody is willing to invest in the new network under the regulatory rules. Hence the only options are deregulation (unlikely) or for the government to fund the next generation of technology. As deregulation is seldom in the personal interests of regulators or politicians, government investment prevails.

This is the business model playing out right now across fixed-line telecommunications markets in Western Europe, the UK, Scandinavia, Australia, New Zealand, and most of the British Commonwealth as fiber networks replace copper. Even though the governments initially built the copper telephone networks and for the most part sold them on to private investors post 1980, the problem is not in the ownership of the network but in the regulatory regime required to price access to services. So even with Wu’s caveat of selling down the network after a decade or so, the problem cannot be avoided.

It would be a tragedy if vibrant (oligopolistically) competitive mobile markets were reduced by access regulation to the moribund state of monopoly fixed-line markets, dependent for their survival on government life support subsidies. However, such a scenario is not necessarily a bleak one for regulators and politicians looking for ways to become indispensable in an otherwise turbulent telecommunications world. Caveat (policy) emptor!