Augar ultimately prevaricated on the idea of controlling student numbers by means of an attainment threshold.

The idea of a DDD entry requirement was shelved, with the report instead reading:

Unless the sector has moved to address the problem of recruitment to courses which have poor retention, poor graduate employability and poor long term earnings benefits by 2022/23, the government should intervene. This intervention should take the form of a contextualised minimum entry threshold, a selective numbers cap or a combination of both.”

As can be seen from this quote, the idea of controlling numbers has only been deferred, not abandoned. The last half-decade of unchecked expansion was made possible by two factors: a demographic dip, which dampened the absolute impact on numbers of an increasing participation rate, and a funding regime that relied on a fiscal illusion to keep government spend on HE off the balance sheet. The latter of these has now ended; the former will do so in the early 2020s when the number of 18 year olds returns to growth.

With forecasts predicting an additional 300,000 higher education places needed by 2030, combined with a declining graduate premium, 30-50 per cent of graduates in non-graduate jobs and growing public concern over low-value courses, “bums on seats” and grade inflation, it’s highly likely that within the next decade the government will need to limit places. Nor would it be a shocking or unprecedented action for a government to take: for almost all of the post-war era, there has been either a numbers cap or an attainment threshold (the former “two “E”s at A-Level” requirement) or, for many years, both. In reading this article I would ask you to set aside your views on whether or not the government should limit places and instead consider what the sector’s response should be if this course of action is decided upon. For surely the implementation of such a decision would be best made with the sector’s input than without it.

So how best to limit numbers? The sector’s arguments against an attainment threshold have been well-rehearsed on Wonkhe and elsewhere; however, as I have previously argued, there are also significant problems with an absolute cap on numbers. Fundamentally, neither the sector nor government should want a situation in which high quality, popular universities which wish to expand are unable to, and where the OfS individually controls every university’s student number allocation, fining institutions which over-recruit. Does anyone really want a situation in which the government is determining which institutions are “worthy” to expand and, equivalently, which will need to shrink?

So what’s the alternative? One possibility is set out below.

A cap and trade mechanism for student number control

In summary, the government would place a cap on the overall number of students entering higher education; however, after establishing the initial allocations for institutions, would not place a hard cap on the number for any institution. Instead, institutions would be able to purchase student number allocations off other institutions, at a rate determined by the market.

To expand upon this, to set up the system, the government determines the overall number of students entering higher education each year. It then applies a student number control to each institution – presumably either the recruitment in the previous year or perhaps, given the recent volatility in recruitment, an average of the last two or three years. These controls are enforced by the Office for Students. To this point, the system is identical to what would happen if the government wished to reintroduce the system of student number controls that existed prior to 2013.

From this point on, however, there is a radical departure. Instead of allocations being fixed by the Office for Students, each institution is free to purchase additional places for any other institution, at whatever price they mutually agree. Places purchased would not be for one year, but in perpetuity, and trades would simply need to be registered by both institutions with the Office for Students, which would then adjust each number control accordingly. Subject to its budget, an institution would be able to buy or sell as many places, from as many institutions, as it wished within a year.

Further points of detail include:

  • The cap would apply at institutional rather than subject level. It would also apply to fee-capped and non-fee-capped providers alike, as both are in receipt of significant public funding.
  • Any changes to the overall cap would be applied pro-rata to the number control for each institution. For example, if the government chose to increase the cap by 2 per cent – perhaps to reflect a larger cohort of 18 year olds – each institution’s number control would increase by 2 per cent.
  • New providers could be given an initial allocation of perhaps 300 or 500 students, following which they would be able to purchase allocation from other providers if they wished to expand further. This could be accommodated within changes to the overall cap without a material impact on existing providers.
  • The price for a place would be determined by the market and would be expected to fluctuate, influenced by demographic trends, the overall cap and the broader funding regime. Given that recruiting an additional student increases a university’s income by at least £7,500 a year (if Augar is implemented) in perpetuity, it is anticipated the price would be in the low thousands of pounds.
  • In exceptional circumstances, for example a fundamental failure of governance, the Office for Students might impose a condition preventing an institution buying additional places. This would be a rare and exceptional individual condition of registration.

Avoids the problems of other solutions

A cap and trade mechanism avoids the worst problems of both an attainment threshold and centrally managed individual caps. Autonomy on admissions is preserved, with it being left entirely to universities to decide who is best placed to study, with no crude bars based on success at A-Level.

Meanwhile, the government does not have to make value judgements as to which universities are “worthy” to expand or contract. Such decisions are difficult enough when allocating new places, but would be exceptionally difficult in a contractionary environment. Leaving aside whether or not the government has the expertise to get the “right” answer (or the debates over what the “right” answer is), either the government would have to make the decision highly algorithmic and inflexible – perhaps by basing it on controversial metrics such as future earnings – or, if done more subjectively, legal challenges would be likely to come from universities from whom places were removed.

A simple approach

Augar’s tentative approach to an attainment threshold begins to look highly complex, with contextual thresholds based on attainment and different age thresholds at which it would apply. As discussed above, centrally managed institutional caps would also likely have to employ complex algorithms and/or bureaucratic “cases for expansion” from providers to respond to changing student demand. By contrast, after establishing it, a cap and trade mechanism could apply equally to post-18 and mature students, to fee-capped and un-fee-capped providers alike.

In fact, depending on the government of the day, the mechanism could be seen as to simple. There would be no way to use it to, for example, clamp down on “low value degrees” or to use place allocations as a means of reward or punishment. For a more interventionist government, with less trust in the market as a means of driving improvement, the cap-and-trade mechanism might be less desirable.

Responsive to student demand

One of the major drawbacks to a traditional student number control is that it freezes the system in aspic. Allocations are fixed, year after year, and it is the government, rather than students and institutions, who determines whether an institution can expand. This typically saw the counterintuitive situation in which the University of Poppleton wished to expand, more students wished to go there and everyone agreed it was good quality – but it was still unable to.

Under a cap and trade mechanism, the market would still drive expansion and government. A popular university would still be able to expand – albeit it would have to purchase the places to do so from another university.

Though still a market approach, the need to purchase places would temper the more vicious consequences of the market by dampening the large year-on-year fluctuations in numbers that have been experienced by some institutions. Expansion would be a considered and costed strategy and no institution would be forced to sell places. Slower expansion would also help to avoid the problem experienced by some students in rapidly-expanding universities who have reported crowded lecture theatre and overstretched student services.

The mechanics of trading

One challenge to consider is precisely how the trading mechanism would work. Even though application data is public, universities may not wish to advertise publicly that they are looking to sell places (they may be more willing to trumpet expansion). Nor would it be the best use of registrar’s time to call round a hundred rivals to see if they could find one willing to buy or sell places.

We should not overdramatise this problem. The trading of a single commodity between several hundred entities is a problem that could have been solved by a 17th century Dutch trading house, let along the English university sector in the age of AliBaba and eBay. Either Universities UK or, failing that, the Office for Students, could establish a simple, confidential, forum in which institutions could indicate the number of places they wished to buy or sell, and their asking price – and then allow institutions to make contact with each other to finalise the precise deal.

One question is whether institutions should be able to trade students places at any time. At first sight, there appears to be no reason why this should not be the case, which would add a new dimension to clearing as unexpectedly popular universities sought to purchase additional places. However, given that other funding and policies are dependent on student numbers, after consultation the Office for Students might choose to apply a “trading window” before the end of which every provider’s final cap would have to be finalised and registered.

Finally, it is important to recognise that, under this system, all money stays within the sector. Although universities are paying to expand, the money goes not to the private sector or to the government, but to other providers within the sector. In this way, the system provides a helpful negative feedback effect to support struggling institutions. A declining institution will, by selling places, receive a significant input of additional funds that can be invested in improvement programmes to turn around its fortunes.

This is a novel idea, which would clearly need further development. Any implementation would need to be the subject of proper consultation, not only to determine the initial controls for each institution, but also to work out the technical details, such as the position with regards to franchised courses, mergers, and so forth. This would also be the case if one were imposing a traditional numbers cap.

As a market-based solution it is more likely to appeal to a government that embraces the role of the market in higher education. But if the government wishes to control numbers – and simultaneously wishes to respect the sector’s concerns regarding an attainment threshold – a cap-and-trade mechanism is likely to be more effective, more flexible and more respecting of autonomy than one where number allocations are controlled by central-planning.

6 responses to “Would a cap and trade mechanism control student numbers?

  1. Iain, you would get the same result if the government just “taxed” universities by imposing a levy on students. From the student (consumer) pov, the outcome is the same: less resource per student, unless the government ploughs back what it collects into higher ed. You also get deadweight loss in both cases: real people who would freely elect to go to university to improve their prospects and whom universities would like to educate.

  2. Bradbury, I don’t think that’s correct: there would be no loss of resource per student in this case (institutions purchase places off other institutions, not off the government, so the money is kept in the system).

  3. Thanx for this intriguing proposal, which I am suggesting to Australian colleagues.

    I don’t mean this comment as a criticism; I’m just trying to work thru the implications.

    I understand Augar’s recommendation 3.5 to mean that the government would fund students in some subjects more than students in other subjects. So an institution with a high proportion of students in a low funded subject would be more tempted to sell or lease out places, and institutions who want to expand high funded subjects would be more tempted to buy or lease in places.

    The net effect would be a shift in places from low to high funded subjects, which I think is exactly as Augar intends since it equates these with low and high value subjects. But the government may not appreciate such a shift in the sector’s funding base, which it could fix by changing the subjects’ funding relativities.

  4. Thanks, Iain – understood. There are, however, fresh problems your post does not consider. One is the fact that universities may have local monopoly power. Given the preference some students have to study close to home, it may not be possible to reallocate quota without exacerbating regional inequalities in progression to university. A second issue is the possibility of collusion if the HEIs are left to do the trading. Two universities with different cost structures have incentives to cooperate and this could be good for the sector. In contrast, a university worried about a competitor could look for ways of denying this institution the chance to grow by finding a way of paying another university not to sell its quota.

  5. This is an intriguing idea. I agree with Bradbury Smith’s comments about competitors. Further, I suspect that some senior management teams would be unwilling to ‘sacrifice’ (as they might see it) potential future growth just because they are unable to fulfil it now (‘Things may always get better, right? Especially under our leadership!’). Some would perceive that failing to fill places would be tantamount to admitting failure, so there may be a reluctance to do this.

    We know that when a market isn’t behaving as government thinks it should, they tend to intervene through additional regulation. I think that might happen here and institutions might then be compelled (or strongly encouraged) to give up numbers if they have failed to recruit. If there are only finite number of students available in the country then you could quickly see how institutions might struggle to retrieve numbers they had been required to give up.

    I think regulation on the basis of ‘value’ is more likely, with an attempt to reduce the number of arts and humanities students (a perceived employer-demand led market rather than consumer-led market).

  6. Clearing is pretty much at an end, and the recent trend of a drop in total student numbers, particularly at English institutions, has continued. I just want to (belatedly) point out, as the bear that I feel I am, that the assumption that UK HE will be oversubscribed in 10 years could well be very misplaced. Indeed, even if UKHE numbers are bouyant, the value of the degrees could well drop in a competitive environment where large employers are increasingly formalising their own training pathways. Placed student numbers are down 17000 since 2016 (<3.5%), but more significantly for English domiciled students (~5%).

    Again, although UK 18 year-olds are increasing, I don't feel like the trend will neccessarily continue upwards, not just as a proportion, but perhaps even in absolute terms. I am not suggesting that assumptions that it will, and planning to mitigate these issues are not valid, but to completely discount the possibility that numbers may fall because of employer-led factors seems a bit naive.

Leave a Reply