What are suitable policy responses?

THE POST-PANDEMIC RAIL TRAVEL CHANGES raise some interesting issues for both transport planners, train operators – and government, regarding the economic and financial framework for the industry as it moves back under public control.

Although, as John Siraut shows, urban rail demand within the Tuesday – Thursday period has now largely recovered from the Pandemic hit, and shows some welcome signs of flattening out the peak, we do still have a problem of concentration in demand on these days, with lower demand on Mondays and Fridays, so that rolling stock and other assets are now generally only fully used on three days.

Other changes may be more subtle and vary between areas. The DfT report also includes operator comments on factors likely to have affected performance during the period examined. South Western comment that ‘commuter demand is roughly two-thirds of pre-pandemic levels’ with a strong concentration on Tuesday, Wednesday and Thursday, but Greater Anglia reports Mondays now ‘more comparable to mid-week’ although Friday peak remains quieter.

When comparing the urban areas shown, the very high totals for passenger arrivals and departures at Reading are striking, the second highest after Birmingham outside London, and exceeding those for much larger cities such as Leeds, Liverpool and Manchester (figure 5). This may point to a relatively large central area employment vis a vis urban area size, as well as impacts of the much improved Elizabeth line service.

In contrast the trends for Liverpool show a marked contrast with other cities covered. After a small recovery in 2023, the 2024 total shows a reversion to the drop of about 25% during Covid. This prompts the question of what might be happening to activity in central Liverpool and also how might this affect justification for expensive new stations at places like Headbolt Lane and Liverpool Baltic?

Note that overall aggregate rail demand (arrivals) is now higher than 2019 when London is included in total – 1.87m versus 1.84m arrivals per day (Figure 2 in John’s paper above), due to growth of off-peak travel and also the general effect of the Elizabeth Line in London. Relatively high evening passenger departures from London after 1900 (Figure 4 above) could also be affected by good levels of service on connecting public transport services (underground/bus) into the late evening compared with other cities, both for central area access and the home end of the trip.

Crowding has been reduced, passengers in excess of capacity in the morning peak falling from 3.9% in 2019 to 1.4% in 2024 (figure 4).

It is noteworthy that the 2024 calendar year NTS data released on 27 August show that per capita ‘surface rail’ usage was 21 trips per person per year over England as a whole in both 2019 and 2024, having risen from 18 in 2023 but within this commuting fell from 10 trips per year to 8, leisure rising from 6 to 7 (quoting trip rates to the nearest whole number – see NTS Table 0409 ‘Average number of trips and distance travelled by purpose and main mode’). A further interesting implication of the NTS data by income group (Table 0705a) is the recovery in rail usage in higher income groups, so that the highest income quintile made 41 surface rail trips in 2024 (from 31 in 2023) compared with the overall average of 21 (albeit not as high as 44 in 2019).

Broader implications of the issues raised from the rail passenger data were examined in a TAPAS network meeting hosted by TfL in March. This followed an article I contributed in LTT issue 908 on 3 February (Adapting to the post-Covid market for public transport – now available on TAPAS tapas.network/76/white.php) identifying some of the questions to be addressed. Presentations were made on this theme by myself, Geoff Hobbs and Charles Buckingham of TfL, and Richard Anderson, Managing Director of the Transport Strategy Centre at Imperial College London, who leads the The Community of Metros Benchmarking Group (COMET), a multinational consortium of metro systems Attendees at the event included those from regulatory and research organisations, and the bus industry, which is affected, but to a lesser extent, by similar user behavioural changes.

A general post- Pandemic recovery can be seen internationally. Data from the COMET observatory at Imperial College for a large sample of metro systems shows absolute ridership at September 2024 close to pre-Covid levels (or in some regions, higher) but noticeably lower in the case of North and South America. In terms of per capita ridership, recovery may be somewhat lower due to population growth.

In general, the event heard, public transport was hit more severely by the Covid pandemic than the car, but has now recovered to a to a level approaching the pre-Covid volumes. TfL’s ’Travel in London’ survey shows a drop in trips per day by all modes between 2019 and 2023 (but most marked for public transport). Trip rate per person per day by all modes fell from 2.21 in 2019/20 to 1.98 in 2023/24, but this also reflects a longer-term drop from 2.59 in 2005/06.

However, looking at specific time periods, more variety can be seen. Growth of Working From Home (WFH) and the associated shift to a hybrid working has reduced overall commuting (especially for rail, whose users more are likely to be in occupations suited to this) with bigger impacts on Mondays and Fridays. Conversely, at weekends demand has held up better and in some cases is now higher than pre-Covid. In the London case, TfL data show the highest recovery on Saturdays (at 98%) but lowest on Mondays and Fridays (both at 78%), and the international metro data also show higher recovery at weekends than weekdays. For rail, this is associated with changes in the trip purpose mix, with a shift to leisure, illustrated in NTS data for England and other surveys. Meanwhile, the bus trip purpose mix has changed very little, with a substantial continuing peak associated with education (mostly school) travel during term times. Some bus operators reported increased patronage levels - probably associated with the bus fare cap (initially at £2 and now £3) and measures supported by the Bus Service Improvement Plan funding and enhanced marketing initiatives. The 2024 NTS data show bus use outside London rising from 25 trips in 2023 to 28 in 2024 (averaged over the whole English population), with the £2 fare cap the likely principal explanation.

In terms of public transport efficiency trends, some striking differences exist between rail and bus. Inevitably, some drop occurred in vehicle-km per employee at the height of Covid, as it is impracticable to cut all staffing levels pro rata to reduced vehicle-km for a short period. This can be seen in case of local bus services in England outside London, where this indicator reverted to very similar levels. However, for rail in Britain, recovery in this indicator was less marked, and indeed it was already falling prior to Covid.

Service levels during Covid were sustained by greatly increased levels of public spending, resulting in a markedly lower farebox recovery ratio. This was particularly evident for the national rail system in Britain, and many metro systems, reflecting the lower percentage of short-run variable costs within the rail system vis a vis buses. It has also created a difficulty in getting back to the farebox ratios found pre-Covid.

This will be a challenge for the new state-owned rail system where the Treasury now carries all the financial risk.

Discussion at the TAPAS meeting suggested that Working from Home was probably here to stay - indeed, a number of employers have reduced office capacity, making a return to full-time office work by all staff impracticable.

TfL data show that the number of workers in London were able to work from home with active encouragement from employers rose from 0.8 million in 2018/19 to 1.7 million in 2022/23, falling slightly to 1. 6 million in 2023/24. Some employers keen to bring back staff regularly to the workplace have faced resistance.

One possibility that might be attractive to public transport operators is that commuting would revert to the full five-day week found in the past. However, this pattern was already changing prior to Covid, and both for employers and employees the hybrid pattern may now be established as providing both with benefits-especially where employers can save on office costs.

A more efficient option might be a more even spread throughout the full working week, enabling a reduction in peak resources. Alternatively, if this proves impracticable perhaps the operating industry (especially rail) needs to look at reallocating resources from weekday to weekend operation, reflecting the pattern of demand.

Another issue still in gestation is change to shopping patterns and habits with the growth of on-line shopping and home delivery and the relative decline in the attractiveness of going shopping (particularly to town and city centres) as a structured part of peoples’ lives, and a family leisure activity.

The new NTS reveals that the most common trip purpose in 2024 (for all modes combined) was still for shopping with 167 trips per person per year. Note the demographic that females aged 40 and over, and males aged 50 and over, who were those making most of their trips for shopping in 2024. Commuting was the second most common trip purpose with 111 trips per person. Commuting accounted for the longest distance travelled on average with 1,007 miles per person.

People made 922 trips in total on average in 2024 (or about 18 trips per week). This was up on slightly on 2023 (915 trips on average), but lower than 2019 (953 trips on average).

Individuals travelled 6,082 miles on average. This was a slight increase compared to 2023 (5,974 miles on average), but a decrease of 6% compared to 2019 (6,500 miles on average). Time spent travelling was 362 hours on average. This is also a slight increase compared to 2023 (353 hours on average) but still down on 2019 (370 hours on average).

We can’t know where trip patterns are going – but that change will continue we can be pretty sure.

A full report of the TAPAS roundtable event is being finalised.
Anyone interested in continuing discussion of these issues is invited to contact the author Peter White at whitep1@westminster.ac.uk