In Charles Dickens’ A Christmas Carol the Ghost of Christmas Yet To Come shows Ebenezer Scrooge his funeral and his neglected tombstone.
Famously, attendees at the former are there primarily for the free food and to steal anything that isn’t nailed down. Scrooge is upset to discover that his passing is unmourned and his person as unloved in death as it was in life. At the side of his own grave he pleads with the spirit to tell him:
Are these the shadows of the things that Will be, or are they shadows of the things that May be, only?
A fixed point in time
It is a question the spirit never answers. As Scrooge begs and promises, the ghost’s skeletal hand begins to shake, before it collapses and “dwindles down into a bedpost”. As we all know, the miserly old man’s outlook is changed that next, Christmas morning – and, beyond the book, we learn he had no more “intercourse with the Spirits” and that he became “as good a friend, as good a master, and as good a man as the good old city knew” and that he lived – for some reason – upon the “Total Abstinence Principle” (Dickens is much funnier than you remember from school).
We are never told how Scrooge dies, or how he is remembered. His question to the final spirit is never answered.
It’s one of my favourite loose ends in literature – I choose to see it as a sly critique of the Calvinist doctrine of unconditional election, fitting for a book that has done such lasting popular damage to Calvin’s notion of a scripturally-based Christmas. For all the way Scrooge changed, we never know for sure if it was enough to change the future he had been shown. Salvation may not be achievable by good works alone, as contemporary readers would have been well aware.
A Christmas consultation
There are other readings of this moment, of course. Richard Puttock at the Office for Students sees the appearance of the final ghost as explicitly redemptive – “to make him a better man” via a glimpse of the future. And in the same way, he hopes to make us better people via a glimpse of the Data Futures.
If it was a consultation you wanted for Christmas – rather than say, turkey and the trimmings, or peace and goodwill to all – you are very much in luck. Some of us with other (shall we say more unusual?) tastes were expecting an update to the Office for Students’ data strategy instead of 46 pages on what might happen with data collection in the future.
Speculative fiction about higher education data is now a substantial body of work. In the five years since the glory of HEDIIP passed, there have been numerous false starts and reconsiderations. The Data Futures programme is currently on the cusp of a beta stage, with providers making practice submissions to test systems, definitions, and processes. Getting to this point has been a long and arduous road – but data future is for the moment data present.
Taking the collection
The thing about data collection is that the first time you do it, the quality will be terrible. Not so much because of new processes or new forms (though these hardly help) but because of the new ideas that they embody. It is for this reason that huge changes to data collection are mercifully rare, and when they do happen it is with a great deal of consultation, communication, trial, and iteration. These things are a burden – but there is absolutely no point in collecting data that nobody can use.
A pause in the process at this point says more – in programme management terms – about the senior responsible owners than the delivery team. Another change, and another consultation, adds confusion at a point when expectations should be solidifying. It’s a move better suited to agile project management – a methodology fine for incorporating customer feedback into a fart noise app on iOS but woefully inadequate for developments that actively benefit from clear advance planning.
As the document notes, Scotland and Wales are moving to a Futures-esque in year data collection model for 2023-24 – principally because regulators there have been signed up to the idea since HEDIIP and would probably be there now alongside England had the Office for Students not changed the programme specification so often.
This years model
So what are these vital new changes?
In a nutshell, there will be a move from the currently proposed three individualised data collections a year model currently being tested in the Data Futures beta. This could be:
- To two collections a year, with the first involving reduced data requirements
- To cumulative collection, with a record building through three submission points
- A single collection point, but with the timing changed from the current (pre-Futures) system
We know that the one thing that won’t be happening is the thing everyone has been working towards and preparing for – three collections a year.
The only real advantage to each of the new proposals is a reduction of ongoing burden. There’s no world in which any of these models gives us a better, or more accurate, view of the sector. Each has a list of disadvantages, and – correctly – the short-to-medium-term concerns about the burden of a new system and consequent data quality issues.
But wait, there’s more
Santa has been kind this year. There are proposals:
- To collect information on qualifications for English-domiciled students from the National Pupil Database rather than from providers.
- To collect data on fees paid by students not in receipt of support via an ad hoc basis rather than in the main collection.
- Not to collect data on financial support offered to individual students
- Not to collect data on students in receipt of Disabled Students Allowance
- To cease to collect data on term-time accommodation and term time postcode
And each one is exquisitely problematic. For instance, the National Pupil Database does not include data on non-traditional entry qualifications (the Access to HE Diploma, Level 3 BTECs, and other FE/Adult learning qualifications). This would be easy enough to read across for OfQual regulated qualification if higher eduction actually made use of the Unique Learner Number (ULN) that links to the central Personal Learning Record (PSR). But even this would not capture the wider range of qualifications – with a detriment to understanding the experiences of non-traditional entrants in higher education.
Likewise, knowing how and where financial support and DSA is allocated seems helpful in understanding the experiences and outcomes afforded to students in receipt of this support. The Student Loans Company, in distributing DSA, will have the data on this and with recent investment in data systems should be able to provide this information to OfS – it is to be hoped that publication can also be synchronised with HESA open data to make this information as transparent as possible. The range of scholarships, grants, and support offered directly by providers is not available in any other data – it fees like the kind of thing that a regulator would want to keep an eye on. And I hope that OfS is ready to respond to the inevitable DfE questions about international fees and cross-subsidy without the data to hand.
House every weekend
You’d think that if the last two years had taught us nothing else, it has taught us that knowing where students live is quite useful in public health emergencies. The consultation fails to note that HESA collected data is used by the Office for National Statistics (ONS) in estimates of internal migration and population, and that provider-collected address data is used in voter and NHS registration, local services planning, and by the private sector accommodation sector.
More plainly, this week saw the release of NUS data showing that student accommodation is increasingly expensive and that poor regulation of the sector has an impact on vulnerable students. With the majority of student maintenance loans covering accommodation, with the growth of the sector having an enormous impact on towns and cities, and with continued concerns about quality and security we need more rather than less data on student accommodation.
Accommodation costs are, according to the OfS’ own research, the biggest concern among new entrants, and the biggest missing piece of information provided to applicants. Rather than inane nonsense about what graduates might have earned a few years ago, we should be supplying young people with detailed information on accommodation costs and details to inform application decisions. Student accommodation is an out-of-control failed market that everyone (the “levelling up” department with responsibility for housing, DfE, regulators, local authorities, the CMA, providers…) seem to believe is somebody else’s problem. But at the very least we need data on where students live to understand how to fix the issues.
Who is data for
HEDIIP was for everyone – from PSRBs to providers. The idea was to replace multiple data returns with a single source of truth, reducing burden by cutting down on the number of people asking broadly the same questions requiring very slightly different answers.
Where there are places that higher education can get student data from sources that are not providers, it should do so. But in many cases there are no other sources, so the decision has to be made based on how the data will be used and by who – and the problem with this is that regulatory use can often crowd other users out.
For any use of data the standard is that it is “fit for purpose” – you can learn a lot from data even if it isn’t accurate… as Wonkhe’s adventures in student residence data has shown (mind you, you can’t learn anything from data that isn’t collected). The consultation includes a surprising deviation of this, and the Office for Students – we learn – is keen to set quality expectations.
I’m trying hard not to say “absolute threshold baseline” here, but you can see where the idea may come from. Minimum expectations become default expectations – in the same way as maximum fees become default fees – and the OfS is unlikely to be able to come up with a Data Excellent Framework to bridge that aspirational gap and foster continuous improvement.
Other innovations and possibilities
If you’re not busy this winter how would you fancy rethinking the HESA Staff Collection? No? Come on, it’ll be fun. We’ve already got the promise that collections of data regarding governors will cease – ostensibly because it wasn’t being used. There’s a consultation on the way from HESA.
And a bunch of provider profile stuff – basic things that where universities are and where staff work – will be rationalised. Again, outside of those fancy maps on Discover Uni, these only really get used in preparing cost centre data – so the idea is to move to an infrequent bespoke collection.
Module based data is only really used when mapped to cost centres – so moving away from collecting data at this resolution would (in normal circumstances) feel like a quick win. But with the Lifelong Loan Entitlement on the horizon, module based data could finally get its moment in the sun. The key thing I take from this is that the regulator knows as little about what the LLE will look like as anyone else – something that worries me more than nearly anything else in this document.
Ghost in the machine
So when OfS invokes the Ghost of Data Futures, it is the version that lies dead dreaming in his house at Cheltenham and who has died many times before. Whatever support HEDIIP once had in the wider sector has long since passed – with the hope simply being that whatever version is instantiated is at least not unnecessarily painful or cruel. The regulator should be praised for bring together round tables to inform these proposals, and it is to be hoped that this consultation – which misses the welcome reference to Cabinet Office rules covering the use of responses to shape policy – is a genuine one.
But data collection and data processing should be a consensual means to achieve a widely agreed goal – in this case a usable and credible data picture of the sector that can be used to inform policy development at all levels. Burden reduction is a laudable aim – but is perhaps not as attractive as a measure of consistency and reliability.
Most of the contemporary retellings of The Christmas Carol – with the commendable exception of the Muppets (seriously, Michael Caine’s greatest screen role) – do not deal with the character Fezziwig, Scrooge’s first employer and the father (possibly, the text is unclear) of his former fiancee, Belle. Fezziwig represents an enlightened and informed paternalistic capitalism in early industrial manufacturing, drawn in contrast to the latter-day Scrooge’s rapaciousness and focus on the bottom line. It’s Fezziwig, not Cratchit, that demonstrates Scrooge’s alternate path.
Fezziwig’s party – as recounted by the Ghost of Christmas Past – has become the archetype of the modern Christmas. When Scrooge sees the praise lavished on Fezziwig by his employers, he remarks:
He has the power to render us happy or unhappy ; to make our service light or burdensome ; a pleasure or a toil. Say that his power lies in words and looks ; in things so slight and insignificant that it is impossible to add and count ’em up: what then? The happiness he gives, is quite as great as if it cost a fortune.
Removing burden is not the same thing, after all, as making tasks less burdensome. The latter is much more a matter of perception – and much more a matter of shared endeavour rather than edict.