Is it time to think again about student number controls?

To listen to some commentators, absolutely the worst thing that could happen to higher education in England is the reimposition of number controls.

Forget, in this perspective, the declining unit of resource – forget the declining value of student support. Forget – even – any measures to widen participation by fully considering the mix and proportion of young people that attend university. The uncapped market must prevail.

That’s an extreme characterisation, of course. The dominant perspective is a general sense that more is better, and that – in the spirit of the 1963 Robbins Report, if not the actual letter of it – higher education opportunities should be available to all who are qualified by ability and attainment to pursue them and who wish to do so.

It’s a noble view, if not one that is often considered critically – for this reason it is important that the position does not bleed into something more insidious: a refusal to accept that anything other than a free market can fairly and equitably distribute scarce resources.

To be clear, we are talking about a market that will involve around two trillion pounds of government debt in real terms by the 2070s – almost as much as the current total government debt. It is an expensive system, and this expense puts providers – and thus students – at risk. On that basis, we’d better be sure that the current system is doing what it was designed to do.

Myth one: The absence of student number controls promotes innovation and dynamism

Chapter four of the 2013 White Paper (“Students at the Heart of the System”) argues that:

To enable the sector to respond to student demand, both in relation to choice of institution, and expansion to meet volume of demand, we want to introduce ways to free up student number controls, while ensuring that overall costs are managed.

The argument for this is fairly convoluted, but in essence:

• We need to “free up” student numbers to allow alternative providers to enter the system without damaging existing capacity.
• The Robbins report said that “courses of higher education should be available for all those who are qualified by ability and attainment to pursue them and who wish to do so.”
• Government doesn’t want to cut eligibility on a tariff basis because admissions should be “a matter between the student and the university.”

We’re a few years away from the full lifting of the cap in 2014-15 here, but the direction of travel is here – and Jo Johnson’s advocacy of innovative new forms of provision – carried this model of disruptive new market entrants into HERA and beyond.

But what actually happened? Here I’ve plotted the actual numbers of UK and EU undergraduate students at each provider since the turn of the millennium. You’ll note that I’ve plotted only providers that had been in existence in some form for the entirety of this period. The colours represent the percentage change in numbers year on year – a deeper red shows a sharp decline and a deep blue a speedy growth.

[Full screen] – note that coverage isn’t exhaustive, there’s a limit to the time even I am willing to spend adding UKPRNs to records linked to the former names of current providers.

Of course, the growth and decline of the sector overall (including the drop in numbers for 2012-13 entry) is overlaid on institutional patterns. But still, if you are looking for evidence for more volatile provider level trends in undergraduate intake since the liberalisation and then ultimate demise of student number controls, you would expect to see more radical changes in provider sizes (darker colours, in other words) more often in recent years. In fact, such moments of seismic change are fairly equally spaced over the last 20 years or so.

The pre-2013 HEFCE system of number controls was much more subtle than is usually conceptualised. It aimed to control the unit of resource (the amount of “assumed resource” – fee plus HEFCE grant – per student in a given price group) by allowing total student income to grow or shrink within a five per cent tolerance band. For providers growing more quickly, no additional funds were available above the additional five per cent – for providers shrinking more quickly the detriment was similarly limited at five per cent.

This smoothed out the impact of applicant choice for providers and students – allowing for the maintenance of resources and staff numbers during lean years, and preventing ambitious providers growing at a speed that did not match their existing resources. It was possible to bid for additional student numbers beyond this limit, but HEFCE would expect to see evidence of capacity and quality – and could target growth in geographical or disciplinary areas of need or priority.

The other main criticism – that new providers could not enter the system – is belied when we recall that several providers entered the university sector, gained degree awarding powers, or changed form, in the years prior to the 2014-15. And substantially less providers than expected entered the new, level playing field, that OfS offered since 2018.

I’d argue that there’s no evidence that an absence of number controls has fostered the kind of innovation and dynamism that those who designed the policy had expected it to.

Myth two: The absence of student number controls drives equitable participation in higher education

What the government argued for in 2013 – what became the “core and margin” system – is now seen as a transitional state on the route to a fully open market, as announced in the 2013 Autumn Statement:

Each year, around 60,000 young people who have worked hard at school, got the results, want to go on learning and want to take out a loan to pay for it, are prevented from doing so because of an arbitrary cap.

The cap was fully lifted in the 2014-15 academic year, with further confirmation of the direction of travel to be found in 2016’s Success as a knowledge economy white paper. Here we find the commitment from then-Prime Minister David Cameron that England should:

double the proportion of people from disadvantaged backgrounds entering university in 2020 compared to 2009, and to increase the number of black and minority ethnic (BME) students going to university by 20 per cent by 2020.

History relates, of course, that though some progress has been made regarding the participation rate among less represented groups we fell some way short of both targets. But the only person brave – or foolhardy – enough to draw a direct link between ending number controls and widening access to higher education was Jo Johnson, in the ministerial foreword:

Access to higher education can be life changing for individuals and, by ending student number controls, we have made the possibility of participation in it a reality for more people than ever before.

Again, it is clear that participation rates among those less likely to attend university have improved in the intervening years. But the proportion of 18 year old successful applicants from POLAR quintile 1 was around ten per cent in 2014 – having grown from around nine per cent in 2010. In 2021 – after eight years of market freedom, and in a substantially larger successful applicant population – the proportion of all students from POLAR4 Q1 backgrounds was around 11 per cent.

[Full screen] (and I’ve also done a provider version if you wanted to dive in deeper)

The bottom chart shows the numerical and proportional year on year growth of placed applicants in each quintile – you can see that with the exception of 2014 these have moved largely in lockstep. It is not, in other words, possible to substantially grow participation from quintile 1 using a free market unless you grow participation in other quintiles proportionally.

It is the sheer overall cost of current and projected growth in overall numbers that has led to the real-terms decline in the unit of resource per student – a lack of political will to increase Treasury exposure to risk and debt. The situation that every reform to the system since 2010 has sought to avoid.

And on this basis, the consequence of continuing to grow participation across the board – even with the honest intention of increasing opportunity – is that the unit of resource and therefore the quality of what is on offer will continue to erode. And this decline will fall most heavily on the kinds of institutions that do so much to recruit the most disadvantaged students.

Myth three: Only an uncapped recruitment system can deal with the expected demographic growth in 18 year olds seeking university entry

As Wolfgang Pauli would put it, this is not even wrong. The only way to meet the needs of a demographic bulge in student numbers is to invest in capacity. We don’t have the resources in the sector as it currently stands – so we need to build new resources, whether these are at existing providers or new ones.

To build resources we need to have at least an outline plan of the shape and size of the sector we want to build. The advent of the Skills and Post-16 Education Bill sees a welcome return to serious thinking about skills needs planning across government. But an unpredictable market makes planning very difficult.

In earlier times the government and funding councils would increase sector capacity by establishing new university-style providers. In the past this has been based on existing providers in the higher education sector (for example the University of Cumbria in 2004) or other parallel sectors (as in 1992). There’s no reason this couldn’t happen again – but it requires active intervention and decision making rather than the invisible hand of the market. The latter also could be seen as offering ministers cover to preside over the overloading of current resources regardless of how suitable they are for less traditional students.

Myth four: Student number controls are just bad, m’kay?

However data and logic stack up, however many lessons can be learned from history (or from other systems – Scotland seems to be doing pretty well), we are still stuck with what I can only describe as a gastro-enteric aversion to number controls. It does just feel like a bad idea.

Market forces advocates like to focus this revulsion onto the plight of a single applicant – bright-eyed, keen, huge potential, clutching some good A levels – in tears on the steps of her local Russell Group provider. Staring at a locked door with a sign reading “Sorry. University full.”

It’s simplistic, it’s emotive – and neither of these things are a sign of good policy making. So we do need to question these assumptions.

Including that the main barrier to participation, especially with more “historic” providers, is academic selectivity. If we’re doing the emotive thing here, we take an examination (the A level) that is a fantastically accurate measure of how middle class you are (incredibly, the current – terrible – number control proposals on the table involve doubling down on these qualifications) and we use that alongside a personal statement that allows further middle class attributes to be taken into account . Wrapping the majority of higher education culture in offputting middle class signifiers seals the deal.

If you leap that barrier we then filter based on family financial health – ensuring only those with access to savings to cover accomodation and living costs which extend substantially above the available maintenance loan. And then we ask that applicants commit to having few other responsibilities for the next three to four years, with any deviation from this flagged as a failure and functioning as a financial detriment without a consequent benefit.

Get through that and – well, let’s hope you are skilled in self-directed learning because opportunities to learn what this is and how to do it are few and far between. There are some instances where serious and powerful effort is being made to address all of these issues, but this remains very much the exception that proves the rule.

And these are the number controls that many people are somehow OK with. A truly equitable system of higher education would remove these barriers from underrepresented – groups while keeping them for the over-represented. It would include a proper, livable, targeted maintenance package. Providing the support that those outside of usually recruited groups need costs money – filling an already full to bursting system with more of the usual kinds of students to try to raise some of this money makes for an impersonal and crowded experience that we already know leads to problems.

Warm words about “capacity to benefit” are not enough. As Robbins actually did say (para 146):

The numbers who are capable of benefiting from higher education are a function not only of heredity but also of a host of other influences varying with standards of educational provision, family incomes and attitudes and the education received by previous generations

A current problem we could solve

The Office for Students still has the power to impose number controls, and can do with a variety of justifications. The Regulatory Framework permits, for instance:

where forecast student number growth risks having a significant negative impact on quality and the student experience due to the overstretching of a provider’s finances and resources, a specific ongoing condition might require the provider to have a student number control (175c).

And:

If the OfS is not satisfied that the risk of future non-compliance [with regulatory requirements] is low, monitoring may be more frequent and/or more intensive and the OfS might set specific ongoing conditions including… student number controls (351).

OfS has never set number controls using these powers – perhaps mindful that HERA 2017 requires it to “have regard to… the need to promote… greater choice and opportunities for students” and “the need to encourage competition… in the interests of students and employers.”

The regulator could use number controls to manage the geographic and disciplinary spread of recruitment – focusing on a balanced portfolio that meets identified national and local needs. It does not. It could use number controls to support efforts to improve participation rates among disadvantaged students – another thing that doesn’t happen. And it could use number controls to ensure – in a time when the real value of the unit of resource is falling – that ministers were rather more aware of the impact of their failure to properly fund higher education.

On that last one – with interest rates rising and fee caps fixed the only way providers can even stand still financially is to continue growing student numbers. And this growth, without new “real” resources attached, tends to lower the quality of the student experience. If there’s 700 students in a 400 seat lecture theatre – if more and more teaching is being done online to larger groups, or by early career researchers rather than salaried academics, if pay is stagnating – the absence of number controls is why.

Think back to the pre-2012 model – the unit of resource was protected, and growth above a certain tolerance level required proof of resource. Without even imposing a hard limit on student numbers, we squared this quality circle in a far simpler way than the current, failing model.

So what else could we do?

Moving to a tolerance band based system of numbers control such as was in place prior to 2012 would not radically alter the shape and dynamism of the sector, Neither would it block innovation or set back the cause of widening access, participation and student success. It would improve conditions at universities for students and the staff that support them. It would put pressure on the government to fund the sector properly, preserving the unit of resource rather than hoping that a rising tide would keep all boats afloat.

And it would allow for planned, considered, growth – a meeting of capacity and national or local skills needs, a directed and clear focus on disadvantaged entrants rather than supporting the extra Q3-5 places needed to ensure that at least some growth is in Q1. It would allow for sector consultation to drive the next stage of higher education in offering what is need rather than the accretion of what we’ve had historically.

It would – at this point – require either primary legislation or for DfE, the Student Loans Company, and the Office for Students to work together to develop a work-around (in a similar way that what is now the OfS register came to be in the last decade).

Would it limit some access to higher education – yes, it would. But so will government attempts to push the HTQs, apprenticeships, accelerated degrees, short courses and other lifelong learning routes beyond compulsory level that it prefers to continued unaffordable growth in traditional undergraduate provision. The plan is to lower student numbers anyway – the question is whether or not we want to plan the new shape of the sector or give the market yet another chance to decide.