What bad signal costs Britain

Since Buffering Britain launched on Sunday, we’ve already written in the Spectator, spoken to GB News, and had our launch video viewed more than 300,000 times.

We’ve also been inundated with supportive messages from people who, like us, experience the impact of poor phone signal every day. The audience we’ve gained so quickly demonstrates how much of a problem this is.

But how much is it costing?

Together with the Centre for British Progress, we’ve had a serious go at the number (full report here). Their conservative estimate is that Britain’s mobile network congestion is destroying £490-785 million of consumer welfare every year (in plain English, what people would pay to have a signal that actually worked). London alone - with just 13 per cent of the country’s mobile users - accounts for £150-260 million of it. That works out at around £27 a year for every Londoner, against roughly £10 a year for every mobile user in the rest of the country. A small but recurring tax on every smartphone in Britain, paid back to no one.

Here’s how we worked it out.

The hard question is what bad signal is worth, i.e. what people would pay to have it fixed. You cannot get that from market data, because no operator in Britain sells you “the version of your contract where the signal actually works at 6pm.” So the model borrows willingness-to-pay estimates from a large European study in which more than five thousand respondents made trade-offs between internet packages at different speeds, latencies, and prices. Those choices reveal how much a unit of speed is worth to a user (what economists call a discrete choice experiment), and what the rest of us call asking people what they would actually pay.

Those valuations are then applied to UK congestion data from Opensignal, which measures real-world network speeds throughout the day. The numbers are stark. In London, mobile speeds at 5pm are less than half what they are at 3am. That gap - between what the infrastructure could do and what it actually does at the moment you most need it - is what the model prices.

Then, we made two adjustments. The first was income. The original survey was Polish, so the model scales willingness-to-pay up to UK incomes (a smaller difference than it used to be), with London getting an additional uplift. The second is honesty: people in surveys tend to overstate what they would actually pay in real life, so the model applies a 30-50 per cent discount, consistent with the academic literature on this bias. That’s where the range comes from.

The London piece is the most rigorous part of it. Opensignal publishes hour-by-hour speed data for the capital, so the peak-hour congestion being priced is measured, not assumed. The rest-of-UK figure rests on a careful extrapolation, because no equivalent hour-by-hour series exists for any other city. We assume congestion outside London is milder than inside it.

These figures are a floor

The numbers capture only the welfare cost of slower speeds. Worse latency during peak hours (e.g. the lag that makes a video call unusable) is modelled separately and flagged as illustrative, because the UK data needed to calibrate it properly does not yet exist. The figure also excludes business users, the million-odd daily inbound commuters into London, and reliability - the dropped calls and timeouts that never show up in a speed average. Each of these is its own meaningful number that’s very difficult to calculate with the data available. So treat £785 million as an underestimate.

With a different, broader methodology, a separate report by WPI Strategy for VodafoneThree arrives at a much larger number: it estimates that fixing Britain’s signal would add £6.6 billion a year to the economy after a decade, by unlocking the 49,000 businesses that, on their modelling, never come into existence because would-be founders cannot rely on the network. The Centre for British Progress prices the welfare we are losing now. VodafoneThree prices the growth we are foregoing.

For a government supposedly focused on delivering economic growth, refusing to build mobile infrastructure is costing up to £785 million a year in direct losses, and billions more in growth that never arrives.

Britain has chosen this. We’re here to make it choose otherwise.