The Labour Party has set out plans “to deliver fast and free full-fibre broadband for all by bringing parts of BT into public ownership and creating a new British Broadband public service”. This is bad economics.

There is no doubt that the wider availability of broadband would be a ‘good thing’. Indeed, the case was well made in a recent CEBR report, commissioned by BT’s Openreach itself. Improved digital infrastructure would bring substantial economic benefits in terms of increased consumer welfare and business productivity, as well as being important for social inclusion and connectivity.

Nonetheless, the case for state ownership and control is flimsy. For a start, full-fibre broadband is not a ‘public good’ in the economic sense of something whose benefits would be hard for a private sector provider to monetise. In particular, unlike classic public goods such as flood defences, individuals who are not willing to pay for broadband access can easily be excluded.

What’s more, while there might be some positive spillovers (‘externalities’) that could justify a proportionate level of public subsidy, most of the benefits of broadband access are enjoyed by the users themselves. Private providers can therefore charge for these benefits and are already incentivised to deliver an optimal service.

Of course, this may not work perfectly in practice. But the UK telecoms market is now one of the most competitive and dynamic in the world. This has already delivered huge improvements in quality and value for money, particularly since privatisation, precisely because of increased competition and the profit motive.

Looking forward, Openreach has argued that “up to 90% of the UK could be attractive for full fibre investment by the private sector”, implying that taxpayer support might only be needed for the remaining 10%. But Labour’s plans go far beyond this, with up to 100% of the work expected to be done by the state.

Indeed, Labour seems to take it for granted that the government can do a better job than the market. Does anyone fondly remember how quick and easy it was to have a landline installed in the 1970s when the state-owned Post Office had a national monopoly? No, me neither.

Labour is holding up South Korea as an example of how the state can help deliver top quality broadband. But even there, the service is provided by competing private companies, and customers are expected to pay.

A better comparison might be with Australia, where the government set up a national network operator which acquired others to form a state monopoly. Guess what? Australia’s ‘National Broadband Network’ (reviewed in this NERA report) has resulted in long delays, huge cost overruns, and failed to deliver the expected results.

The UK government’s own track record in picking winning technologies is pretty dreadful, including in telecoms. Full-fibre broadband is the gold standard – at the moment. But it would be typical if Labour directed tens of billions of taxpayers’ money into a particular model only for something better to come along.

Here many people hold on to the romantic view, popularised by Mariana Mazzucato, that the state was responsible for the global technological revolution in the first place, and conclude that the state should take over the delivery of broadband too.

However, this narrative confuses the invention of things such as the World Wide Web or GPS, and the work done to turn these inventions into something practical for widespread use. It’s almost always the free market, driven by the dreaded profit motive, that has created the commercial applications.

In the case of broadband, there are doubtless ways in which the market could be made to work better, including a strengthening of pro-competitive regulation and a reduction in barriers to the roll out of broadband, such as easier access to land and properties for all providers.

But Labour’s plan would actually strengthen the dominance of a single player, and a state-owned one at that. This is only likely to drive other providers out of the market. (If we’re still tied to the EU’s rules on state ownership, procurement and competition, the European Commission might have something to say about this as well.)

Labour is also wrong to claim that state broadband would be ‘free’, or, in the words of John McDonnell, “literally eliminate bills for millions of people across the UK”. This pitch is odd in itself: if this many people are already willing to pay £30 per month for broadband, what’s the point in making it free for everyone, regardless of need?

In any event, someone would still have to pay – whether as current or future taxpayers, as customers receiving a worse service, or as shareholders whose assets are being expropriated, or some combination of all of these.

The financial costs are also significant. Labour has estimated the upfront capital cost at about £20 billion and said that this would come from its ‘Green Transformation Fund’, i.e. more borrowing. (Labour’s estimate was based on the ‘national monopoly’ model in this Fathom report, which also pointed out the flaws in this model, including less incentive to innovate and a lengthy delay of 3 to 5 years before it could be implemented. Few reading this report would have concluded that this was the optimal solution.)

Similarly, the cost of bringing parts of BT into public ownership would be paid for by issuing new government bonds and swapping them for shares, i.e. more borrowing. All this is on top of the substantial costs of Labour’s existing plans to renationalise other utilities, the water industry, and Royal Mail.

In principle, the increase in debt would be offset by the acquisition of an asset. But borrowing is still borrowing, whatever the purpose, and someone would still have to be willing to hold the new bonds. Potential investors are already worried by Labour’s broader economic agenda. They may not be reassured by Mr McDonnell’s statement that the broadband plan is the limit of nationalisation, especially as he appeared to rule out any move on BT only a few months ago. Any attempt to pay less than full value would only increase that nervousness, as well as hit all shareholders – big or small.

Valuing assets is also difficult, and any numbers are intrinsically less certain than the costs of the debt that finances them. There’s no guarantee that any new public sector assets would keep their value under public management, especially given conflicting priorities and demands.

As for the ongoing running costs, Labour says these will be “more than covered” by taxing “UK-based multinationals on the share of their global profits that reflects their UK share of their global sales, employment and assets”. This is presumably a nod towards the bureaucratic ‘formulary apportionment’ system being proposed by the OECD, which would be difficult to introduce unilaterally.

However, Labour’s original press release actually named Amazon, Facebook and Google, which suggests that the real target would be US-based tech giants. Like the current government’s planned (and seriously flawed) ‘Digital Services Tax’, this would appear to be a de facto international tariff and therefore potentially in breach of WTO rules.

Either way, while companies may be a popular target for new taxes, they are only legal entities and cannot bear the economic burden themselves. The reality is that all taxes are ultimately paid by people – including employees and customers, as well as shareholders. What’s more, turnover taxes have particularly large deadweight costs, are more likely to be passed on to consumers, and are a major deterrent to investment.

Finally, Labour’s costings are already being debunked. Various sources, quoted in the Financial Times and the Guardian, have put the upfront capital costs at between £30 billion and £40 billion, compared to Labour’s estimate of around £20 billion, and the annual running costs at between £1 billion and £2.6 billion, compared to Labour’s £230 million.

In addition, there’s the bill for renationalising Openreach itself, for which estimates start at around £12 billion, and taking on its share of BT’s pension liabilities, which could be another £4 billion to £5 billion.

These costs would multiply if, as seems likely, the government also has to take over other companies who are unable to compete with the new state-owned giant. Indeed, these companies are already shelving investment plans, and the government may well have to end up nationalising the entire industry. It’s not difficult to come up with a final price tag as large as £100 billion.

In short, surely there are more effective, efficient and fairer ways of improving digital infrastructure, with much less public money? There is perhaps a case for an increase in targeted public subsidies to help deliver broadband to poorer communities, alongside other measures to strengthen competition and reduce barriers to investment. But creating what is soon likely to become a state broadband monopoly would be a huge and costly step in the wrong direction.

This is an extended version of a blog first published by the Institute of Economic Affairs

• Applied economics
• Digital economy
• Nationalisation