A spiral of costs? Is the USS pension scheme doomed? – Sean Wallis

Introduction

The Universities Superannuation Scheme (USS) was set up in 1975, replacing a stocks-and-shares Defined Contribution scheme called FSSU.

For forty-four years, from 1975 to 2011, USS paid out a Final Salary pension based on a 1/80 accrual rate. This meant that if you paid in for forty years you would retire on half your annual salary. Far from being unaffordable, the scheme grew. Only in 2011 did we begin to see the introduction of ‘Career Average Revalued Earnings’ (CARE) replacing Final Salary. Contributions for employees until 1997 were 8%, on the basis of covering historic problems created by the FSSU, reduced to 6.35% from 1997. Employer contributions fluctuated over the time period (10% initially, rising to 18.55% in 1983, 14% from 1997, 16% from 2009).

But for over forty years the scheme was essentially very stable. As a ‘last man standing’ multi-employer scheme in a growing HE sector funded by government spending, the risk of default was considered to be very very low.

In 2011 we began to see the beginning of ‘reforms’ that have increased income to USS and, apparently paradoxically, led to ever increasing claims of deficits. These included closing Final Salary to new entrants (but increasing costs to 7.5%), dividing the workforce and increasing the retirement age from 60 to 65. At the time, these changes were justified on a number of grounds: the 2008 recession had hit pension assets, staff were living longer, etc. As we know, 2011 was not the end of the ‘reform’ programme but the start of the dismantling of USS Final Salary and now Defined Benefit.

Our first response to the alleged ‘crisis’ in USS funding has to be, why was it that USS was able to pay out a higher-value Final Salary pension scheme for 44 years? How can this be possible when we are told that a lower-value CARE scheme — with a higher 8% employee contribution — is now ‘unaffordable’ and destined for deficits?

What has changed?

First Actuarial’s analysis, Progressing the valuation of USS (Salt and Benstead 2017) projected forwards to show that the shared 8%/18% contribution rates were sufficient for the scheme to pay pensioners for the next thirty years without touching the assets (£60bn by that point). So in fact, in its own terms, USS should be in a steady-state, its assets growing through investments returning levels of 5-10% pa, well in excess of CPI inflation at around 3%. Not only is there no deficit, but the scheme has a healthy working balance — on most recent official figures, £67.5bn in 2019!

Made in Westminster

The central assumption underpinning the health of the pension scheme – the very low risk of default – has been undermined by Central Government. The chaotic expansion and competition of the tuition fee market ‘reforms’ launched in 2011 with the £9k undergraduate fee have been devastating for the sector. But it has also undermined USS.

The Willets Plan was implemented in stages, starting with undergraduate tuition fee hikes to £9k and the curtailment of block grants in 2011; the removal on restrictions on student recruitment in 2014; and reduced regulation in the Higher Education and Research Act of 2016. This last reform allows universities to go bankrupt overnight. It abolished the Higher Education Funding Council for England (HEFCE) and replaced it with the Office for Students, whose boss, Sir Michael Barber notoriously commented that ‘Universities are not too big to fail’.

Prior to 2010, universities survived on £6.5-7k per student, mostly paid via local authorities. For the universities, the ‘£9k fee’ (now £9,250 a year) represented a profit margin of some 25% per student, with higher rates of profitability found in teaching departments that did not need laboratories or large numbers of research active professors.

This created huge financial incentives to universities to direct their activity towards recruitment of students onto courses that could be taught cheaply. It also drove up expansion. Last year the BBC reported figures from the Higher Education Statistics Agency (HESA) that show that between 2007 and 2016, Exeter University increased undergraduate student numbers by 74%; University College London (UCL) grew by 65%; York by 48%, and so on. But London Met lost 62% of its students over the same period; West London, 44%; Cumbria, 42%, and so on.

This market impact is not limited to England. Scottish universities are not permitted to charge £9k undergraduate fees to Scottish and EU students. But faced with booming competitors south of the border, they sought to increase their income in other ways, especially in overseas undergraduate recruitment. Glasgow Caledonian set up a New York campus with almost no students. Like many English counterparts, Scottish universities have also been accused of dodgy recruitment practices in international markets, using agencies earning £1-2k commissions per student, along with a concomitant reduction in academic standards to resist any decline in applications. Scottish universities have also developed ‘foundation’ years with private providers such as INTO and Oxford International.

From boom to bust

Universities have been rapidly transformed from publicly funded bodies whose surpluses were tiny compared to turnover (i.e. had low profit rates), into aggressive corporate competitors engaging in massive building programmes and building up large surpluses to pay for them. Competition is both external, between universities, and internal, as departments vary in the level of potential surplus they can gain from selling courses and degrees. Lab-based courses with intensive personal teaching are more expensive to provide than those taught in lecture theatres alone. Medicine has been exempted from this market, with student numbers still capped.

The university ‘winners’ in this market chaos have poured vast amounts into new campuses and taken out huge loans to pay for them. The ‘losers’ have closed campuses, departments and courses down.

Market competition has been devastating and vicious. Most of the first victims are in the post-92 sector, where there has been massive job losses and increased casualisation. Alongside a job cull, we have seen permanent jobs increasingly replaced by insecure hourly-paid staff on casual contracts.

But it is not just post-92 where universities are threatening bankruptcy, cutting courses and closing campuses. The University of Reading, a pre-92 campus university, had to report itself to regulators in February 2019 after selling £121m of land it was given in trust. Leaving this £121m aside, its debts total some £300m and it is running an operating deficit of around £20m a year.

Reading is a ‘USS university’. Whereas it has captured the headlines because it has been forced to open its books, it is an open secret that several other pre-92 universities are in a similar position.

‘Peak student’

Meanwhile student recruitment has levelled off. Despite the hype, we have reached ‘peak student’. HESA reports that students taking a first degree increased by 61,310 between 2008/9 and 2017/18. Current growth is at most 1% a year. Worse, this headline increase has been more than offset by a decline in students taking a second or third degree, which collapsed by 233,140 over the same period. Despite this, in 2019 universities were warned by the Office for Students for projecting a 10% increase in student numbers over four years!

We can expect the market to wreak more havoc in the context of Brexit. The pool of government-backed fee-paying students will inevitably shrink. EU students will be reclassified as ‘overseas’, most will find themselves ineligible for UK government support, and charged higher fees. It is possible that universities will be put under pressure to cut overseas student fees, which will be welcome. But there will likely be a contraction.

The current funding system is unsustainable for another reason. There is now an off-the-books ‘debt mountain’, underwriting student loans of £121bn (as at March 2019), of which the Treasury expects at most half will be repaid. Labour’s 2016 Election Manifesto committed them to abolish tuition fees. In response, the Tories commissioned a report, the ‘Augar Report’, which made a number of recommendations, cutting the fee to £7.5k and bringing back some of the direct funding of the block grant. But as we have seen, universities are now mortgaged to the hilt, and such tinkering could drive universities over the edge of bankruptcy. Indeed, senior management teams fear speculation about reduced fees in the future putting off students today.

Like very many of my colleagues, I am completely in favour of abolishing undergraduate tuition fees. It is just and equitable. Society gains from education. It was sustainable for decades in the past, and is sustainable again. The Robbins Principle that ‘if you can benefit from education you can come’ should apply.

Fundamentally, education is the gift one generation bequeaths to the next: the fact that some young people are wealthier than others due to luck or inheritance should be immaterial.

But in abolishing fees, we have to demand that the huge capital projects launched by universities in the last decade are brought under control, otherwise the employers will simply engage in massive job losses.

What does this mean for USS?

The USS pension scheme’s strength is that it is a mutual scheme, a ‘last man standing’ multiple-employer scheme.

The first political problem concerns the risk of default. If the scheme remains a mutual scheme, that risk should remain low, despite the risk of some universities going bankrupt.

Ironically, thanks to the growth of university estate, the market capacity of the sector has grown, so it may be easier for rival universities to absorb students and take on extra staff. As long as income stays within the pre-92 USS universities, the damage to USS of a university bankruptcy might be quite limited. This is of course not to justify the terrible impact that this might have on individuals’ lives. But it is to say that the risks to USS as a multi-employer scheme are likely to be exaggerated.

But this depends on whether USS stays a mutualised multi-employer scheme. Already there is discussion about breaking the scheme up into ‘sectors’. This idea arises from the logic of market competition. It must be utterly opposed.

An example should suffice. Take University College London (UCL) and Kings College London (KCL), two pre-92 universities in Central London. Both have taken out large loans to pay for huge capital investment programmes, which may, or may not, be covered by student fee income over the next decades. UCL does not know whether KCL has overreached; KCL does not know whether UCL’s gamble will pay off. Either way, the question increasingly becomes, why should KCL cover the pension risk of UCL, and vice versa?

Whereas these particular universities have not made statements of this kind public, others have. In 2017 Oxbridge colleges lobbied for the scheme to be de-risked and potentially broken up. Trinity College Cambridge since left the scheme. These institutions are not the most expansionist of colleges! Rather they were, in an orchestrated act of lobbying, expressing the view that they did not trust the other universities over their balance sheet reporting and were not prepared to pick up the cost of market failure by reckless competitors.

Quite simply, university senior management teams are attempting to reduce their exposure to ‘the employer covenant’, the guarantee that the pension fund will pay pensioners if, at some point in the future, the scheme itself goes bankrupt. In essence the employers insure the scheme’s investment strategy, offering financial support to manage short-term investment risk.

Undermining this covenant is a betrayal of future generations of pensioners by current employers locked into a current competitive expansion for market advantage.

‘De-risking’: what it is, and how it undermines the scheme

This is what lies behind the decision to ‘de-risk’ the pension scheme. ‘De-risking’ means two things:

  1. selling off assets currently invested in stocks and shares, and buying government bonds and gilts instead, and
  2. planning to hold all assets in gilts long-term for the entire life of the pension scheme.

In 2017 some 70% of assets were invested in stocks and shares, and the November 2017 plan, which has begun (see below), was to disinvest over a 20-year period.

de-risk-sketch3

Figure 1. Two alternative de-risking plans considered in 2017. The ‘Nov 2017’ plan reduces the proportion of assets held in stocks earlier, slows growth, and imposes a greater penalty on the likely long-term scheme value of assets. It is far from the only option however.

Despite the name, ‘de-risking’ does not reduce risk. It trades short term investment risk at the expense of the long-term risk of default. It makes scheme default much more likely, and so causes contribution costs (to employees and employers) to rise.

To see the likely consequence of this change, consider inflation.

  • The current ‘CARE’ scheme, and the now-frozen Final Salary scheme, are revalued annually by CPI (this is subject to a cap at 5%).
  • Therefore, to avoid default, USS must ensure that its investments beat CPI.

Currently, stocks and shares invested by USS yield returns conservatively estimated at around 4% above CPI, while gilt yield rates have been below CPI by around 0.8 percentage points.

In his excellent video on USS with Matthew Malek, Sam Marsh explains how USS’s projections of asset performance have been extremely conservative in the past, and compares actual and projected growth rates. I reproduce one of his graphs below, primarily to illustrate the impact of ‘de-risking’.

uss-projection17

Figure 2. Plot of total value of scheme assets over time, due to Sam Marsh (2017). The solid line is the history of how the assets performed to 2017. The dotted lines are the 2011 and 2014 projections, based on ‘prudent assumptions’ (employing a confidence interval and taking the lower bound, i.e. the lower bound of projected performance). The dashed lines represent the 2017 lower-bound projections of the performance of the scheme at that time.

In simple terms, ‘de-risking’ the scheme increases the risk of default. It is like selling a house, keeping the cash in a low-interest-bearing account for twenty years and then trying to buy back the same house. You don’t have to be a property speculator to realise that there is a very high likelihood you will have to pay considerably more!

This is true if you model for an imaginary investment strategy in the future — it is why, for example, the November 2017 valuation obtained a much higher projected deficit (£7.5bn) than the September 2017 one (£5.4bn). See the graph above.

But it is also true if you actually begin to ‘de-risk’ the scheme. According to USS this process began in 2017.

Note that ‘de-risking’ is not a temporary sale of assets (e.g. due to short-term concerns about Brexit), losing some possible income for a period, and then re-investing again in the future. It is based on expecting a permanent one-way transfer of assets to gilts. (Hence the red line on the right hand side of Figure 1 continues on at 0%!) But since gilts are expected long-term to yield a return substantially below CPI, de-risking will inevitably bleed the assets dry.

Unless we can change this equation, we are locked into a pattern of diminishing assets and spiralling costs.

In 2017 this was a projection. Now de-risking is being implemented. And buried in USS’s latest Summary Funding Statement published in December 2019, are admissions that confirm UCU’s analysis:

  • De-risking is voluntary. “With the right economic conditions, we believe that opportunities should be taken over the years ahead to reduce the amount of risk – specifically investment risk taken in the funding of defined benefits – so it is consistent with the level of financial support employers have told us they are willing to provide in the long term, should experience prove to be worse than our expectation.” (‘Our funding plan’, p3)
  • De-risking has begun. “At the 2017 and 2018 actuarial valuations, we incorporated this long-term, gradual de-risking into our funding approach, with the intention of reducing the amount of investment risk over a 20 year period.”
  • ‘De-risking’ trades one risk for another. The scheme becomes heavily dependent on long-term projections of gilt yields. “Liabilities increased by 8% as falls in interest rates led to decreases in future expected returns at 31 March 2019.” Since these are low, this further increases the risk of default.

USS report the assets of USS grew between 2017 and 2019 from £60bn to £67.4bn, but liabilities grew by 7.8% from £67.5bn to £72.8bn. (The £7.4bn of asset growth breaks down as £7bn from investment growth and £0.4bn in surplus contributions).

By contrast, increasing contributions have a limited impact on the balance sheet. Total contributions in 2018 brought in £2.2bn, which (assuming it reflects a stable employee demographic) allows us to perform some back-of-envelope estimates:

  • The total income generated by increased contributions between 1 April 2019 and 30 September 2021 will total less than £1bn, even allowing for future wage increases.
  • Increasing contribution rates to 11% for employees and 23.7% for employers from 2021 would bring in at most £0.75bn a year.

USS say that “based on this approach and our assumptions we expect the USS Income Retirement Income Builder deficit to be removed by 31 March 2028,” by which time the scheme will be half way to being fully de-risked (see Figure 1).

Although USS is demanding much higher contributions from staff and employers (at a level likely unaffordable for both), the actual benefits to USS’s future balance are tiny: each year it amounts to about one tenth of the loss of income created by ‘de-risking’.

‘De-risking’ can be reversed

Now for the good news. Precisely because a de-risking valuation depends on long-term yields of assets, there are multiple options available to the scheme.

One might

  1. slow the rate of de-risking, currently with a target of 100% de-risked in 20 years,
  2. plan to stop at a partially de-risked scheme, say 20 or 30%,
  3. pause de-risking now and maintain the current balance of assets and gilts,
  4. reverse de-risking, returning to the 2017 ratio.

Options (2)-(4) wipe out the projected deficit. See Figure 3. The first option would probably eliminate it, depending on the degree to which the rate were reduced.

de-risk-sketch4

Figure 3. Sketch of alternative de-risking plans, all capable of implementation from the current time, and all of which would reduce the projected deficit.

These attacks on our pension scheme are wholly reversible.

Striking back

In 2018, UCU members took industrial action because the employers pushed for closure of the USS Defined Benefit scheme. The union grew by 50% in those universities that took action, illustrating the popularity of the action and the seriousness of the task. In 2019 members resumed that action, with some 80% of members in USS universities striking for eight consecutive days in November-December.

In 2018, the actions of ordinary members changed the narrative and forced the employers back. That strike produced a Joint Expert Panel, which pointed out rightly that de-risking was a choice and even at that point, the scheme need not project a deficit.

But whether it is by USS or UUK exaggerating the risk to the sector by market failure to encourage employer panic, or whether it is university managers’ own poor business choices, what is clear is this: left to their own devices, between them they will inevitably take the line of least resistance by weakening the employer covenant, and progressively shifting risk onto employees.

This is why I proposed the sector explore a Government Guarantee in 2017: seek agreement from the Treasury that they will act as lender of last resort.

We have seen that the Treasury has allowed the student loan book to inflate to £121bn to promote the market in Higher Education, half of which is unlikely every to be repaid! They have created the market chaos that has undermined our university sector — and the pension fund. Arguing they should be part of the solution is like arguing that private railways need a nationalised Railtrack.

The latest funding statement confirms what critics from First Actuarial to Sam Marsh and Mike Otsuka and the JEP had said throughout. ‘De-risking’ is a choice, and bears no relationship to the actual performance of the invested pension fund. It is a decision that can be altered or reversed.

We are at a turning point in the future of USS and the sector.

A Brexit sting in the tale

There is one further reason why a rational USS investment strategy would not plan to de-risk the scheme at the present time, even were one ideologically disposed towards this in the longer term.

Making assets dependent on future gilt yields exposes the fund to new risks. Chief among these are that Brexit may undermine USS’s recovery plans.

The reason why long-term gilt yields are so low (currently around 1%) is post-2008 Quantitative Easing. But the latest indications from Boris Johnson are that he intends to float through Brexit by printing more money. This can only hit gilt yields, and cause USS to project an ever-growing ‘deficit’.

In the Summary Funding Statement, USS glibly states that ‘the right economic conditions’ are part of the decision to de-risk, yet it is difficult to think of a post-war period when de-risking would be more likely to create a problem for the pension scheme!

See also

About Sean

Principal Research Fellow, Survey of English Usage, University College London
This entry was posted in HE BIll, pensions and tagged , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.