Britain’s Buses Became More “Productive” While Running Less Service

England outside London carried about 1.85 billion local-bus passenger journeys in 2024/25, compared with roughly 1.98 billion before COVID-19. Passenger journeys had recovered to 93.3% of their 2019/20 level, while bus service mileage had recovered to 86.0%. Put those two figures together and passenger journeys per service mile were about 8.5% higher than before the pandemic.

That is a legitimate productivity ratio. Operators have spent years adjusting schedules, controlling costs and concentrating resources where passengers are travelling. The difficulty comes when the ratio is treated as evidence that the passenger network itself has improved, because the underlying statistics cannot answer that question.

I examined the issue in How Can a Bus Network Become More “Productive” by Running Less Service?, the second article drawn from my broader The Bus Is Not the System report on what Britain and Ireland’s long experiment with bus governance can teach cities preparing to electrify their fleets.

The Department for Transport counts passenger journeys as boardings. Someone changing buses on the way to work will generally appear twice. Service mileage measures the distance travelled by buses providing local service, so dividing the two tells us how many boardings occur for each mile operated. It does not show how far passengers travel, whether important destinations remain reachable or whether the connections needed for complete journeys still exist.

Consider an evening service connecting a residential area with a hospital through a major interchange. It may carry relatively few people after the peak and be expensive for an operator to retain. Removing it cuts driver hours, energy use and vehicle wear, while the busy trunk route feeding the interchange can continue to perform well. The worker whose late shift can no longer be completed by bus stops appearing in the boarding data altogether.

There are good reasons to remove poorly used mileage. Britain’s deregulation experience produced genuine pressure on operating costs, and London achieved major efficiencies through competitive tendering while retaining public control over network specification. Congestion, staffing, maintenance and vehicle utilization all affect bus economics, so asking operators to run every historical route regardless of demand would be a poor use of money.

The harder decisions arise when a service contributes to journeys beyond the passengers sitting on that particular bus. A feeder can make a high-frequency trunk route useful. Evening coverage can determine whether someone can take a job with irregular hours, and a modest cross-town service can eliminate the need for a second household car. Those benefits have limits, and sometimes a timetable change, better transfer or smaller amount of targeted service can preserve much of the value at lower cost.

This is where metropolitan network planning becomes important. Individual operators are well placed to understand their vehicles, labour costs, schedules and fare revenue. A transport authority has to look across operator boundaries and judge what happens to employment access, healthcare trips, education and other complete journeys when one piece of the network changes.

The post-pandemic comparison with London and Dublin illustrates how different choices can be made. London kept bus mileage broadly stable even while passenger journeys remained below 2019/20 levels. Dublin expanded scheduled service substantially through BusConnects and subsequently recorded passenger volumes above its 2019 level. The statistics use different definitions and baselines, so they do not support a simple ranking, but they do show that service supply is partly a policy decision rather than merely a response to current fare revenue.

English legislation has started to acknowledge the same issue. The Bus Services Act 2025 requires authorities using enhanced partnerships to identify socially necessary services where withdrawal could materially harm access to employment, essential goods or social activities. Authorities still have to work within their budgets and legal powers, but the framework recognizes that the revenue collected by one service is not a complete measure of its value.

A fuller performance picture would keep the traditional operating measures and add accessibility. Cost per vehicle-kilometre, reliability, service delivery and patronage all remain useful. Alongside them, authorities need to understand which neighbourhoods can reach major workplaces, hospitals and colleges within reasonable travel times, whether the journey still works in the evening or at weekends, and whether a practical return trip exists.

Fleet electrification makes those questions more consequential. Battery-electric buses are now the reference technology for urban transit, and deploying them brings long-lived decisions about depots, charging equipment, grid connections and contracts. Building that infrastructure around a fragmented or poorly specified passenger network can preserve today’s institutional boundaries for decades.

That is one of the central findings of The Bus Is Not the System. Competition can improve the efficiency of bus operation while the passenger network still requires an institution able to plan across routes, operators, depots and infrastructure. Electrification does not remove that responsibility; it makes the consequences of getting it wrong harder to reverse.

The 8.5% improvement in journeys per service mile tells us something useful about the services that remained. To know whether passengers are better served, authorities also need to understand what disappeared with the mileage and whether those losses made the rest of the network less useful.

The detailed British statistics are explored in How Can a Bus Network Become More “Productive” by Running Less Service?, while the governance, contracting and electrification implications are developed in the full The Bus Is Not the System strategy report.