Commuter students? It’s all about the money

The conclusion was tidy – journey time is the best way of identifying commuters in need.

An accompanying blog from OfS Director of Strategy and Delivery Josh Fleming encouraged the sector to use journey time as “an organising concept”, and promised the findings would inform the ongoing review of the Equality of Opportunity Risk Register – where commuter students already appear as a group facing risk to equality of opportunity.

I’ve been playing with the data tables published alongside the summary report. They contain 74 tables cut by a banner of around 120 columns – sex, age, disability, caring responsibility, study mode, tariff group, region, and more – of which the report uses perhaps 15. The rest are published and never cited.

And buried in the unused columns is a different story about who the journey actually breaks, plus a trade-off exercise in which students put travel cost at double journey time as the thing that drives their choices.

Time trumps address

First, the research as written up. Savanta surveyed 1,069 students in England in early June, deliberately oversampling those with one-way journeys of 45 minutes or more (an hour in London).

It also tested four ways of defining a commuter student against each other – journey time, weekly travel cost (£20 plus), living in the parental or pre-HE home, and not living with other students.

Journey time wins on almost every count. Students over the 45-minute line are more likely to report a negative impact on their experience than those under it (32 per cent v 12 per cent) – a bigger gap than address (27 v 15), living arrangement (26 v 13) or cost (29 v 18) can manage.

The relationship is graded rather than binary – the proportion experiencing a high calculated impact climbs from 6 per cent under 30 minutes to 22 per cent at 90 minutes plus, and combining measures adds almost nothing (long travel alone gets you 32 per cent reporting negative impact, long travel plus parental home gets you 34).

The address-based definitions that dominate sector practice, meanwhile, turn out to sweep in large numbers of people with no meaningful journey at all – 36 per cent of address-based commuters travel under 45 minutes, and 14 per cent walk to campus.

The damage, where it lands, lands on participation rather than feelings. Most long-journey students still say they belong (66 per cent, against 74 for short journeys) – but 67 per cent have been unable to take part in societies, clubs or social activities at least once in the term, 76 per cent were too tired from travel to prepare properly for a taught session, and 76 per cent have decided not to travel in at all because of the cost.

More than two in five commuters on every measure either say their institution does nothing to support their travel or can’t name anything it does. What they want is unglamorous – travel discounts or bursaries (56 per cent), earlier notice of timetable changes (42), online or hybrid flexibility and compact timetables (39 apiece).

And commuting is mostly a constrained choice rather than a lifestyle one – among those for whom staying close to home mattered, 67 per cent say the cost of moving away would have been too high, and around 35 per cent of long-journey students would have lived somewhere else if they could have afforded to.

The report is clear about its main limitation – the long-journey half of the sample is unweighted “natural fallout” from a screening question, while the short-journey half is weighted to HESA benchmarks, so every commuter versus non-commuter comparison in it is between a representative group and a non-representative one.

Students said cost

Now for the tables. Table 53 is the debrief question from a trade-off exercise – “which feature made the biggest difference to your choices?” – whose actual results appear nowhere in the pack, the report or the blog.

Students’ answer is travel cost, on 42 per cent, with journey time second on 20. Among the long-journey commuters themselves the gap widens to 45 against 18. Timetable structure, days required on campus and online flexibility take most of the rest – and “on-campus commuter support” comes dead last on 1 per cent.

A debrief self-report isn’t the same thing as estimated preferences – people reliably overstate how much cost matters when you ask them straight – which is why the exercise’s actual results, absent from the pack, are the analysis that could settle the question.

As it stands, the regulator’s research crowned journey time as the organising concept on the same fieldwork in which students, asked directly, organised themselves around cost by better than two to one.

Both findings can be true – time discriminates experience, cost drives decisions – but only one made the write-up, and they point at different interventions, one of which the sector seems to be easing off on.

Nationally, cost is a Department for Transport lever. Under the £3 bus fare cap, a student making two single journeys a day, three days a week – and remember, the tables show three quarters of long-journey students travelling in on three or more days – pays £198 across an 11-week term, or £330 if the timetable demands five days. That’s the sort of maths sitting behind the 76 per cent of long-journey students who’ve decided not to travel in because of what the trip costs, and why the £2 bus fare cap announced in the Burnham Blitz is so important.

Less positive is DfT’s integrated transport strategy, Better Connected, which manages ninety pages on fare caps, tap-and-go and targeted support for named groups – disabled people, older people, women, rural residents, care-experienced young people – without once recognising higher education students as a category of transport user.

https://www.gov.uk/government/news/better-connected-tap-and-go-travel-across-trains-trams-and-buses-announced-in-governments-new-transport-strategy

Worse than any commute

The report’s methodology section explains that employment, caring responsibilities, disability, sex and study level were excluded from the analytical framework because using them “would risk conflating the circumstances that may intensify the effects of commuting with commuting itself.”

Fair enough as far as definitions go. But run the outcome measures down the excluded columns and the intensifiers turn out to be the big thing.

Disabled students – 27 per cent of the sample – have considered leaving their course at nearly double the rate of everyone else (39 per cent v 19). For comparison, the gap between long-journey and short-journey students on the same question is 27 v 22. Disabled students are more likely to have missed deadlines because of their journey or living situation (56 v 42 per cent), to have missed welfare support (55 v 44), and to say they couldn’t afford to live where they’d prefer (33 v 23).

Students with caring responsibilities carry a high withdrawal risk of 23 per cent against 14 for everyone else, and their journeys are longer to start with – only 20 per cent of carers travel under 30 minutes, against 37 per cent of non-carers. Among carers who leave campus quickly after their last session, 43 per cent leave because of caring – and just 20 per cent say there’s no reason to stay, against half of non-carers. These aren’t students who can’t see the point of campus – they’re students with somewhere they have to be.

Part-time students post the single highest “high withdrawal risk” figure of any subgroup in the entire workbook – 26 per cent, above even the 90-minute-plus travel band, and comfortably above every commuter measure the report tested. Two thirds of them say their journey has contributed to thoughts of leaving. They’re also, awkwardly, the least likely to say they’ve ever actually considered it (18 per cent, against 25 for full-timers) – the calculated risk measure and the direct question flatly disagree about part-timers, which is the sort of thing that deserves a footnote somewhere.

These groups overlap heavily, of course – carers skew mature and part-time, three quarters of them live in the parental home, and disability tracks both. What the tables show is a cluster of correlated risks rather than four independent ones, and disentangling them would need the microdata nobody has published.

Ethnicity, meanwhile, would need the tables nobody produced – long-journey students are significantly less likely to be white (61 per cent, against 70), and not one outcome in the entire pack is cut by ethnicity. It appears as a profiling variable only.

And then there’s sex. The report states that differences in sex between commuter measures are “generally small and not statistically significant.” Yet table 63 marks the female share of long-journey commuters – 69 per cent, against 57 per cent of the comparison group – as significant. So do the journey-length tables – 41 per cent of men travel under 30 minutes against 29 per cent of women.

Women are significantly more likely to report travel tiredness (50 per cent often or sometimes, v 40), significantly more likely to say they couldn’t afford to live elsewhere (29 v 20), and significantly more likely to prioritise the timetabling fixes – earlier notice (45 v 35) and compact timetables (42 v 32). This may well have something to do with subject, but for clarity, commuting in English higher education is very much a gendered issue, and neither the report nor the blog mentions it.

Never heard of it

The tables also complicate the report’s better-known finding on self-identification. The report notes that a third of high-impact students don’t identify as commuters. The full table is stranger – self-identification peaks at 74 per cent among those travelling 60 to 89 minutes and then falls to 59 per cent among the 90-minute-plus group.

The worst-affected students are less likely to claim the label than the merely badly-affected, plausibly because a student who has lived at home their whole degree experiences the journey as ordinary life.

One in nine students has never heard the term “commuter student” at all – rising to around one in seven among the long-journey groups the agenda is aimed at. Anything badged “for commuter students” mis-addresses precisely the students it most needs.

A few other things the tables know that the report doesn’t say. Commuters don’t compress their weeks – 75 per cent of long-journey students travel in three or more days, barely different from anyone else, which means the long commute is being absorbed at close to full attendance frequency rather than designed around, and the tiredness figures follow.

The burden is institutionally concentrated – low-tariff providers hold the most 90-minute-plus commuters (12 per cent of their students, against 7 at high tariff), 48 per cent of London students travel over an hour, and students in the North have considered leaving at 30 per cent against 21 in the South.

A withdrawal question asked of the whole sample finds 52 per cent of all students saying their journey has contributed at least somewhat to thoughts of leaving – against the 24 per cent who say they’ve ever considered it. The report presents its version of these figures as being about students who had considered leaving; the tables show the question went to everyone.

And the living situation version of the same question comes out statistically identical for commuters and non-commuters, which rather undermines treating housing as a commuter-specific problem.

29 per cent of long-journey commuters say their commute has a positive impact on their experience – almost as many as the 32 per cent negative that made the headline – with a fifth of the open-end responses describing the journey as decompression time. The median commuter is coping. The point of the exercise was always to find the ones who aren’t.

Fleming’s blog says the research “is by no means definitive” and invites the sector to write in with evidence about where journey time isn’t useful.

On the strength of the tables, the answer is that journey time is a decent screen and a poor target. It sits in no administrative dataset – the report concedes the sector uses address because it’s “readily available” – so acting on it means asking a new question at enrolment. It sits in no DfT metric either, not least because its transport connectivity metric excludes universities “due to the complexity in pinpointing single access points, as they typically consist of numerous buildings, including administrative and non-educational facilities”. Translation – nobody could be bothered to find the DiscoverUni dataset.

If a student withdraws because the bus route was cut or the fares became unaffordable, the university’s continuation rate takes the hit and the transport authority faces no consequence at all – the variable this research crowns is homeless in both departments.

But the students it will find, once found, sort by the variables the framework excluded – disability, caring, part-time study, and sex, tangled together in ways only the microdata could separate.

Commuter students are already on the risk register – what the review will settle is how the risk is described and what providers are expected to do about it. This research gives it a description the report itself never quite states – an overlapping cluster of disabled students, carers, part-timers and women absorbing long journeys at full attendance, asking for cheaper travel and a timetable built by someone who has seen a bus timetable.

Whether that survives the journey from data table to risk register is the important bit.