Will we ever know where students live?

Last year on Wonkhe I spent a long time trying to work out where students are.

I wasn’t the only one. Universities, public health bodies, and indeed the Office for National Statistics also ran up against similar issues. I’ve just been reading a fascinating ONS report that tells us just how far off we all were.

There are three main sources of information about student term-time addresses:

• Students (generally) tell their university where they live, and this information is recorded on HESA. It is possible to use this to understand how many students are in a given area down to quite small resolutions – you’ll recall I used this in the Wonkhe Covid Dashboard, which used 2018-19 student term time data to identify “student areas”.
• Students may also register with their local GP surgery when they arrive at their term time address – this records data on the Personal Demographic Service (PDS). Again if you think back to last year there was a momentary panic when it was thought that students’ positive Covid tests were being registered at their home address (this was during those bleak “blame students for Covid” months).
• Students are also meant to register a student only household with their local Council, in order to qualify for a household council tax exemption under class M or class N. Again, we played with this data at the start of the term – we tend to use it when looking more generally at housing costs or prevalence in a local authority area.

These sources are all administrative data – collected, in other words for reasons other than statistical ones. What ONS has done is to look across these three data sources for the last pre-pandemic year to spot where students show up in more than one. We learn some fascinating things.

Fun with administrative data

First up. How many students are represented in both HESA and PDS data as living in a particular local authority? The answer is around 60 per cent. First years are slightly less likely to be in the same place in both data sets, second and final year students slightly more so, and postgraduates are in agreement nearly 80 per cent of the time.

This tells us – on a practical level – that students may be filling in forms before knowing their actual term time address, particularly in their first year of study. And if you don’t know your address, you are likely to supply your parents address to your university or new doctor on the grounds that any post will get to you that way. A sensible caveat for students – but a mess in for public health planning.

You’re probably wondering whether some areas have term-time residents that are better at getting this information together than others. Well, you might not be – but fascinatingly enough there is a huge difference. Just 32 per cent of students living in Wrexham are represented in both datasets, compared to 75 per cent in Cambridge. This could tell us something about data collection practices and GP registration processes at different providers.

[Full screen]

Is this data ever updated?

Having accurate data on where students live became increasingly important during the “tiers” stage of the pandemic, where local restrictions could affect some students at a provider and not others. More than anything else, this made it clear how little we know about where students live – it was impossible to tell which students had gone back to a home address, which had moved to another student house, or out of halls, unless we were specifically told.

Even in a normal year, this doesn’t routinely happen. First years are more likely to update their address with their GP when they move to university, but even then less than 20 per cent are up to date. For other undergraduate years, this is just five per cent that bother.

HESA data suggests that 41 per cent of students register the expected three unique locations during a three year course with 45 per cent registering two and 14 per cent just one. Of course, not every student moves every year but these numbers are lower than expected given that a fair number of students do move in year as well – and HESA only captures the postcodes so a move two doors down may not count as a move.

Do students live in student houses?

Apparently not. Just 12 per cent of students are registered in households containing solely students with PDS. Thirty per cent live in addresses that contain students and non-students, but this could show former students who have not bothered to re-register after graduation. A gloriously unlikely fifty-eight per cent of students are registered at their family address – as we know from HESA that forty-five percent of students are in halls we can see a problem.

In council tax data only 45 per cent of student households are registered as exempt from council tax. More than 40 per cent of households made up of students and non-students appear to be (incorrectly) exempt.

We can see, after a fashion the “bedroom turnover” in student addresses. No, not that, you mucky lot – it is a way of measuring how many of the total possible students in a student house are new to that house based on council tax registration. In most cases this is less than 40 per cent, which again feels counter to what we know of student household formation.

Covid and after

As those conducting the Census in England and Wales know only too well, last year was not in line with usual student residence patterns. ONS are examining 2019-20 and 2020-21 data as it becomes available, and there’ll be more releases in this series as we go. There’s also new data on international students visa journeys – suggesting just 35 per cent of international students successfully apply for working visas and remain in the country.

All of this, of course, represents a snapshot of the last days of the old regime. We’ve lamented the absence of decent data on student residences frequently – any action on costs or standards needs to start with reliable data, and with OfS and DfE both wringing hands at the idea of getting involved at regulating the sector that soaks up most student maintenance funding this situation is unlikely to improve. Individual providers will – after the last couple of years – be keen to keep more detailed and accurate records on student accommodations, but pulling this together in a form that would provide a basis for systemic reform is a way off yet.