Alternative Pension Schemes

In the UK, pensions schemes are broadly either Defined Benefit, or Defined Contribtion.

Defined Benefit is a disaster. The employer ends up underwriting the scheme, whose assets will potentially be many times the value of those of the employer. The government is desperate to avoid having to rescue any scheme, so insists on very cautious estimates of future financial performance, leading to something which is ridiculously expensive.

Defined Contribution is even worse. Now an inexpert, unregulated, employee ends up making many of the investment choices. But to make matters much worse, the standard deviation on life expectancy is now much higher. A collective scheme can reasonably predict that someone retiring at 65 will, on average, draw a pension for about twenty years. It will want to be prudent, and add a couple of standard deviations to that estimate, so will calculate on the basis of something like 22 years. But the standard deviation for an estimate of an individual's life expectance will be around 8 years, so adding twice this gets one to about 36 years. In other words, a prudent individual should budget for living to about 100, whereas a similarly prudent manager of a collective scheme would budget for people dying, on average, at about 87. The individual needs to have sufficient assets to pay a pension for over half as long again as the collective manager.

(Socialists will also complain that the current rules for DC schemes can give rise to an evil called inherited wealth.)

Collective DC?

Would it be possible to design a pension scheme which has the advantage of a collective scheme of merely having to fund the expected average lifespan, letting those who leave the trials of this world early fund those who tarry for longer, but avoids the guarantees of a DB scheme that cause the Pensions Regulator to become so cautious?

One possibility is a form of conditional indexation within a DB scheme. This indexes benefits not according to CPI, but rather according to investment returns. In general some guarantees would remain (such as never having negative indexation, even if the returns are losses). Because of this, the indexation would not be the full amount of the investment return.

It raises problems. How does one choose the risk / yield profile of the investments? Would all employees agree to the profile chosen? Does the employer have any incentive for making sure that sensible decisions are taken?

But it has significant advantages over the two extremes of DC and DB, so should not be dismissed lightly.

My belief in choice and markets would mean that I would suggest that multiple CDC pension funds be set up, all with different rules and risk profiles. Then individuals could choose which to place their pensions in, and should be permitted to spread their pensions across multiple such funds. The funds should be regulated with the PPF guaranteeing them against collapse (i.e. guaranteeing that the zero indexation payout is secure). With different funds competing for one's assets, there is some hope that they will be kept honest. Transfers between funds should be permitted, just as they are for DC schemes. One of the big problems with the USS is that my employer probably wishes to leave it even more strongly than I do, and yet neither of us reasonably can.

I think it possible that CI would provide a much cheaper pension than the common DC. Its main disadvantage for some would be that it leaves nothing to be inherited. For those who worry about that, there is little reason not to permit a combination of CDC and DC, just as some people currently combine DB and DC.

Solving the USS problem

So how would I propose solving the USS problem? I'd try to keep the current benefits for an extra year or two (probably with increased contributions), until CI looks possible (i.e. until the Pensions Regulator is prepared to regulate it). If current benefits prove utterly impossible, then some combination of cuts to benefits with increases to the employers' contribution rates, but not including introducing a CPI cap of 2.5%. I believe that moving the accrual rate to 1/85th, the threshold to £40k, and the employer contribution rate to 24.5% should have been an acceptable short-term position for the Trustee.

Then I would reduce the current DB to DC salary threshold to as low as £25k, whilst introducing CI options within the DC section. No-one could object to more options in the DC section, and reducing the DB threshold does seem rather inevitable. Possibly the DB accrual rate would need to be reduced a little too, but I could not support lowering the inflation cap.

The result I would hope for would be that those who worked for their full forty years or so would receive their gold-plated State Pension of around £10,000, with another £10,000 or so of pretty firmly guaranteed DB pension. That should give an income of around £20k, and it is difficult to argue that more is needed. It is easy to argue that more would be appreciated and deserved, and that could come from the less guaranteed CDC and DC parts.

In other words, the Pensions Regulator says that our current DB guarantees are expensive. I feel I need a secure income in retirement which is substantially higher than the State Pension, but I only really need strong guarantees for the first £20k of total income. Above that, weaker guarantees in exchange for better expected returns, given that the guarantees are turning out to be very expensive, sounds like a reasonable exchange. Even more reasonable if there is a choice of options.

State Underwriting

The State fully underwrites the pension schemes for Civil Servants, Teachers, Armed Forces, and many others. Why not for everyone? The bill could be quite large, but it should be economically efficient to have the underwriting done by the one institution which can never fail, being able to print money to cover any eventuality.

Politically unlikely to happen in the UK.

Even the more modest solution of allowing the pre-92 University to join the TPS is unlikely, although it would certainly be fairly to allow the pre-92 Universities to enjoy the benefits found in the post-92 ones. It would also allow the Government to take over the USS's substantial assets, so it would have some immediate economic benefits.

More Options

The current 9.8% contribution rate is expensive for lower-paid workers (particularly postdocs) in expensive areas (Cambridge, Oxford, indeed most of the SE of England). Yet people who opt out lose a significant contribution from their employer.

Could one not imagine a scheme in which one can opt to pay 50% of the standard rate for 50% of the benefits? And 75% for 75%? And, given the falling accrual rates and the comparison with the TPS, 125% for 125%? Afterall, the TPS offers the option of paying extra and getting extra.

It is unclear to me whether this is the best approach. Why can people not afford the current contribution rate? How can Universities in the expensive SE justify paying their staff no more than Universities in cheaper parts of the country? Would higher wages not be a better solution? But, pragmatically, more USS contribution options would be an improvement, and seems more achievable.

There is also much to debate about the employer's contribution. If someone pays at a 50% rate, should the employer's contribution to that member's pension also drop to 50%? And should the employer's contribution to the deficit reduction drop to 50% too?

Currently around 20% of those eligible for the USS opt out. The proposed reforms of April 2022 will not make things better unless the option of a lower contribution rate is added. The minimum contributions the Government permits in a workplace DC pension are 5% from the employee and 3% from the employer, calculated on salary between £6,240 and £50,270 pa. The opt-out rate is about half what it is for the USS too.

It also seems very unhelpful that there is a financial incentive for the employer to have people opt out, and a loss to contributions to the deficit reduction. Could the USS change its rules so that the deficit contribution charge was paid based on the salaries of all USS-eligible employees, including those who had opted out? The Regulator might even find this larger base for deficit recovery payments re-assuring.


The current governance of the USS really does not work well. The JNC is too often split with UCU vs UUK, and the "independent" chair deciding everything. Now both the UCU and UUK seem to agree that the Trustee's approach to risk and valuations is deeply flawed (even though UCU and UUK between the appointed the majority of the memebrs of the Trustee Board, they cannot remove any). The Pensions Regulator tends to side with the Trustee, but sometimes not as strongly as the USS would like to claim (see 2019 misrepresentation of TPR by USS).

The structure also disenfrachises many, both many employers (such as Trinity College), and many employees (the majority of whom are not UCU members, some of whom are members of other Unions). That the USS recognises the UCU (and only the UCU) for representing employees for pension negotiations is no guarantee that an employee's employer will recognise the UCU for local disputes, or that it will not recognise another Union instead, or even none.

And the monopolistic lockins are unhelpful. Employers effectively cannot leave, and cannot offer their employees alternative schemes. Nor do even the employers seem to have much influence over the Trustee.