Adapting to the post-Covid market for public transport

FOLLOWING THE VERY SHARP loss of ridership due to Covid, public transport provision in Britain – and broadly equivalent countries - largely survived, but only due to very large amounts of government support. This has left a legacy of concerns over sustainable ongoing levels of public funding, and about what should be the new expected commercially viable levels of service provision in a materially changed marketplace for bus and rail use, and where is public support needed and justified.

Whilst overall ridership recovery (in terms of trips per capita) was in 2023/4 around 85% of the 2018-19[1] level in Britain, this masks significant variation, with stronger recovery evident in some urban areas and market sectors, notably long-distance rail. Some services which had been almost completely suspended during the pandemic have largely revived, notably the more substantial bus-based park and ride schemes and the express coach network. The latter, unlike local bus and rail, did not receive revenue support to offset ridership loss during the pandemic. Other modes have seen changes too, for example cycling increased during the lockdowns, and car use fell and has not yet returned fully to pre-Covid levels, NTS data for England showing a drop in car driver trip rates of about 4% between 2019 and 2023[2]. These wider patterns are more fully explored in John Siraut’s analysis of the latest National and London statistical travel data in TAPAS, 3 February 2025 (see article). But it was not just the levels of use that changed, but the character of demand, propelled by working from home (WFH), new ways of shopping (particularly a switch on-line) and other lifestyle changes for leisure and entertainment.

Substantial regional variations can meanwhile be seen within Britain, Wales showing the lowest recovery in bus demand at about 70%. Here there has been no stimulation from the £2 fare cap as in England, or the free travel for under 22s as in Scotland. London shows a slightly lower recovery level within England – both for bus and rail – probably associated with higher levels of working from home (WFH). Rail in London displayed a mixed pattern, with the largely-orbital Overground routes sustaining about 97% of pre-Covid total ridership but the Underground somewhat lower at 88%. The successful opening of the immediately popular Elizabeth Line complicates the picture[3]. Between 2018-19 and 2023-24 Underground train vehicle-km fell by 7%[4]. The latest quarterly data for national rail use (July -September 2024) meanwhile show an encouraging growth in trips of 9% over the same quarter in 2023[5]. So what are the most significant implications of the new structure of demand?

The national fare cap has boosted bus use particularly on longer distance routes

One post-Covid expectation raised by some commentators was that behaviour change might enable a ‘flattening out’ of peaks in demand (primarily for the journey to work) due to a shift to WFH. There was already evidence of WFH prior to Covid, as shown in a study sponsored by TfL and British Telecom of which I was a co-author[6]: Both employers and employees saw attractions stemming from greater flexibility, reduced demand for office space and reductions in commuting effort and costs. The more senior staff in professional activities were more likely to be suited to this, hence typically rail rather than bus commuters. A dramatic shift toward WFH was imposed by Covid lockdowns, including sectors previously averse to this (notably financial services). A partial reversion to previous patterns has occurred, but a complete one is unlikely despite recent evidence of some employers seeking to get staff back into their offices for more - or all - of the time.Rather than working wholly from home, an arrangement emerged amongst many employers and employees of a ’hybrid’ pattern, in which 2-3 days per week are worked at home, the others at the traditional workplace. This reduces commuting and office space needs, while enabling face-to-face contact at the workplace to be maintained, although it brings challenges to public transport operators in determining appropriate service levels. If evenly spread throughout the working week, a reduction in peak resources would be possible through thinning out frequencies, but in practice WFH tends to be focussed on Mondays and Fridays – with marked reductions in the morning peak in particular - leading to a mid-week period with substantial peak demand, and also busier levels of weekend use than before. However the Tuesday-Thursday demand level itself is substantially lower during holiday periods.

Indeed, there is a general pattern of greater passenger retention and new activity at weekends, both for bus and car (reflected in the DfT’s daily travel index since Covid) and reported by many rail operators, which could be seen as part of a new spreading of demand within the week as a whole.

For the local bus sector, the peak travel impact of WFH has been less than for rail, since bus users tend to be in lower-income occupations more anchored to workplaces (albeit with substantial local variations). A strong morning and afternoon peak demand for education travel (mainly schoolchildren) remains, adding to the morning peak work journeys. However, off-peak bus demand has been hit by the substantial drop in travel by older people on concessionary passes since the pandemic which has continued with little sign of full recovery. This seems to have been associated with fears of infection, and a shift to home delivery of shopping.

In 2018-19 concessionary travel represented 28.1% of all bus trips in England outside London, falling to 21.9% in 2023-23[7], a drop 6.2 of percentage points (and an even greater absolute reduction in volume, given the drop in total ridership). Some bus operators are reporting a complete recovery in fare-paying traffic, with the drop in concessionary travel thus representing all the net fall in total demand. More recently, some recovery in off-peak shopping travel by fare-paying passengers, and some new leisure travel has been evident, stimulated in England by the £2 maximum fare from January 2023, (increased to £3 in January 2025). This increased travel on rural and longer distance routes where the savings were most significant. The National Travel Survey for England indicates the changes in demand by trip purpose between 2019 and 2023 calendar years within each mode, which can be seen as a proxy for demand by time period. For local buses outside London, the changes are remarkably small. For example, commuting trips comprised 20% and 21% respectively, education 23% and 21%, and shopping 24% and 23%[8]. Given that overall bus demand is down, this implies a similar absolute drop in each category. Conversely for rail, commuting fell much more steeply from 47% to 37% of all trips, with a growth in the leisure share substantially up from 26% to 35%. Education travel grew from 8% to 13%, which, if occurring mainly at peak times, would offset some of the reduction in peak demand associated with the drop in commuting. Note that the NTS covers all days of the week, and runs throughout the year, whereas some demands are more concentrated. This is particularly true for school travel which occurs only on Mondays to Fridays and during a term time, which is shorter than the adult working year. Hence a proportion of the fleet may be required purely for this purpose. So, alongside this, to what extent could operators reduce costs in response to reduced demand and output?

Very sharp reductions in output were made in the first lockdown in 2020, but thereafter service levels (expressed as vehicle-km) were increased from the low base, well in advance of ridership recovery. Local bus services provision in England outside London is now around 83% of pre-Covid levels (but substantially higher in London itself at 95%, the cuts being largely in the central area) For national rail, the service level was back to 92% in 2023-24. The cut in bus service levels outside London is clearly worrying in terms of reduced travel opportunities for those who depend on this mode, and the attractiveness of it to others, a reason that TfL resisted such significant cuts. The bus sector has displayed greater flexibility than rail in reducing costs with output. In part, this is due to the cost structure, with a substantially higher share of energy cost within the total (about 13%, compared with 6% for rail operators) enabling pro rata reductions with output. Rolling stock leasing charges in contrast represent a very much larger share of total costs for rail than depreciation and leasing for bus fleets. Given an intention of retaining resources to enable service resumption after the pandemic, these costs could not be avoided. However, the bus industry has also shown greater flexibility in varying staffing levels with output. Apart from a dip in the peak Covid year (2020-2021) local bus-km per staff member have remained remarkably stable, being 4% below the 2018-19 level in 2022-23. In contrast the equivalent rail indicator for train-km per Train Operating Company staff member is 15% lower. Looking over a longer period, from 2010-11 to 2022/23, these differences are even more marked, the rail productivity indicator in 2022-23 being 26% lower, but bus only 3% lower[9]. This is attributable to substantial increases in total rail staffing levels as demand rose in the years before the pandemic, as well as lower train-km post-Covid. A similar finding is evident in a report from the Office of Rail and Road, assessing change from 2013-14[10]. Combined with revenue changes, the overall effect is that total public spending in the bus sector within England outside London was lower in real terms in 2022-23 than 2018-19, albeit higher on a passenger trip basis[11]. This primarily reflects substantial reductions in concessionary fare compensation and BSOG. Conversely, public spending on rail support has remained stubbornly high, around 230% of 2018-19 figure in 2023-24[12], rising from £5,418m to £12,434m, and specifically identified by the Chancellor in her first post-Election statement last July as the main cause of the Department for Transport’s significant overspend of £2.9bn. “But instead of developing a proper plan to adjust for this new reality, the government handed out cash to rail companies to make up for passenger shortfalls, but failed to budget for this adequately” as she put it.

In contrast, the government (and its predecessor) have seemed more relaxed about bus support, especially as it has been more closely targeted.

What then, are the future options?

In the short run, the improvements by local transport authorities funded through the government’s Bus Service Improvement Plan grants should offset some of the reductions in service levels outside London since Covid, hopefully reflecting opportunities for increased ridership, stimulated already by the fare cap, rather than simply replacing withdrawn facilities.

Increases in population and in particular large increases in planned housing would also be expected to stimulate additional demand, especially if steps are taken to avoid the location and design of new housing that locks in so called ‘car dependency’.

A simple return to the pre-Covid status quo is unlikely anywhere however, given the fundamental changes in travel behaviour, even with some further scope for passenger recovery. Whilst employers are urging (or insisting upon) a greater proportion of days worked at the office - or even fully five-day office working in some cases - WFH seems here to stay at some level. Benefits to employers in being able to recruit from a wider range of potential staff and save on office space still apply and there is resistance amongst employees to a return to traditional five days a week working. Concessionary travel by bus is meanwhile unlikely to fully return, also affected by rising car availability amongst the elderly within the successive cohorts to whom concessionary travel is becoming eligible. Reductions in concessionary fare compensation have meanwhile already affected total bus industry income.

However, there are opportunities for operators to reallocate resources from weekday to weekend output, notably through working agreements for rail staff to incorporate Sunday working in regular rostering.

There are also considerations regarding when engineering works and line closures are undertaken, now weekend travel levels are similar to weekdays in some places

The more substantial problem may lie in the uneven distribution of demand (especially rail) within the working week associated with hybrid working, so that rolling stock and infrastructure needs for rail, and fleet levels for bus, are determined by Tuesday-Thursday demand. It therefore becomes particularly important to ensure maximum fleet availability on Tuesdays to Thursdays, perhaps by re-timing vehicle maintenance work. The cost structure for bus may also become more capital-intensive, accelerated by the shift to electric buses which have about twice the capital costs of diesels but helpfully lower maintenance, and (dependent on electricity pricing) energy costs.

Ideally, there should be means of encouraging a better spread of demand through pricing, as use has shifted from period travelcards or seasons to pay-as-you-go with capping, enabling more variation by time period, although this does reduce revenue received in advance. Simply extending off-peak pricing to all-day on Fridays had very little effect when trialled by TfL last year, but perhaps greater differentials (with increases on Tuesdays-Thursdays) might have more impact. It also remains in employers’ interest to maximise use of office space and thus spread demand during the working week. Ironically new types of ‘flexible day’ season tickets may make variable pricing more difficult.

In the longer term, other changes affecting the passenger transport industry as a whole may have more impact. Driverless operation is already becoming more widespread on urban metros worldwide, and may become possible for buses – perhaps initially on BRT /bus priority corridors and then in designated urban /commercial areas where suitable traffic control is possible, and ultimately on roads in mixed traffic. For buses, this would sharply reduce operating costs, and enable higher frequencies with smaller vehicles, improving overall attractiveness of the service. For national rail a general improvement in labour productivity is needed in any case.

A bigger threat to rail could be the sheer size of the support bills faced by the government if the return to state ownership of rail operators does not work out well – and always a concern in the Treasury. Buses in urban areas, in contrast, seem to have considerable support from both central government and the transport authorities in the metropolitan areas and some larger towns and cities, where they figure prominently as a sustainable transport policy tool.

The underlying opportunities to increase ridership and revenue by raising service quality and marketing initiatives also of course remain for both bus and rail.

I’d be pleased to have further discussion on the issues raised here, both at the small round table being kindly hosted by Transport for London on March 20th (see below), or on the TAPAS platform where this article is also being posted.

References and Links:

1. Based on trips reported by operators in DfT tables Bus01a and LRT0101, and Office of Rail and Road (ORR) table 1220, showing 2023/24 compared with 2018/19. Converted to trip rate per head by reference to ONS mid-year population estimates.

2. ‘NTS Table NTS0303a ‘Average number of trips by main mode (trips per person per year): England, 2002 onwards’

3. TfL Travel in London Annual Overview, December 2024. Trends from a base year of 2019

4. From Dft Table LRT 0105

5. Office of Rail and Road ‘Passenger Rail Usage July to September 2024’ published 19 December 2024

6. White, P., Christodoulou, G., Mackett, R., Titheridge, H., Thoreau, R. and Polak, J. ‘The role of teleworking in Britain: its implications for the Transport System and Economic Evaluation’ Paper at European Transport Conference, Netherlands, October 2007

7. DfT Table Bus 01d.

8. Derived from NTS Table 0409a ‘Average number of trips by purpose and main mode: England, 2002 onwards’, with percentage shares by purpose calculated by the author, rounded to the nearest whole number.

9. Derived using bus-km data from DfT table Bus02a, and staff from table Bus07a. Rail data from Office of Rail and Road tables 1243 (passenger-train km) and 2233 (TOC staff)

10. Office of Rail and Road ‘Discussion paper on rail industry productivity’ April 2024

11. Derived from DfT Table Bus05bii

12. Derived from ORR Table 7270 ‘Government support to the rail industry (April 2023 to March 2024 prices) Great Britain’, based on the ‘total operational funding’ category, less freight support, i.e. excluding funding for enhancements such as HS2.