The conditional indexation of USS benefits is the most promising route to their improvement

The case against HE12

In this post, I argue that the conditional indexation/revaluation (CI) of improvements to the current level of defined benefit (DB) provides the best means of enhancing the pensions of USS members. It follows that UCU delegates should vote down HE12 at the Higher Education Sector Conference on May 30th.

HE12 calls on UCU “negotiators to cease work on CI and to focus on working to improving members’ DB benefits”. It offers a false dichotomy between the improvement of DB and the introduction of CI. It advances the unsound claim that “CI is more risky to members than DB and will not improve pension benefits”. Halting work on CI would, in fact, abandon the most promising means of improving the DB accrual of members, where this improvement would be no more risky to members than their current DB promise.

Members are now promised the following inflation indexation and revaluation of a pension worth 1/75th of their salary for each year they pay contributions into the scheme: an uprating which fully covers CPI inflation up to 5% and covers half of inflation between 5% and 15%. There is no uprating for CPI inflation above 15%. This is known as a “soft cap” on indexation and revaluation, since it softens without eliminating the impact of inflation when CPI runs between 5% and 15%, before the cap fully sets in when inflation runs above 15%.

With CI, specified inflation-indexed increases in the value of one’s pension are rendered conditional on returns on investments in the fund being good enough. The following is a concrete specification of CI which would improve, without risking, existing member benefits:

1. An improvement beyond the current soft cap of the indexation of pensions in payment, up to full CPI each year, conditional on investment returns being good enough. This would protect members against the erosion of the spending power of their pensions when CPI inflation runs higher than 5%.

2. An improvement beyond the current soft cap of the revaluation of the pensions accrual of active members before retirement, up to full CPI plus x% each year, conditional on investment returns being good enough. If this x% were set at 1.6%, this would match the level of revaluation that members of the Teachers’ Pension Scheme are currently promised.

Whenever investment returns are good enough to warrant increases beyond the soft cap, the conditional indexation and revaluation sketched above would constitute an unambiguous improvement on the current level of DB for USS members.

It is worth emphasizing that, under this proposal, the current soft cap would remain a floor below which indexation and revaluation could never fall. This is, therefore, a version of CI which would pose no risk to members, since it would offer only the prospect of gains beyond the current DB provision, with no possibility of loss to that provision. It is, moreover, an analytic and conceptual truth that risk involves the possibility of something bad happening rather than of something good happening: e.g., the possibility of loss rather than of gain.

We are now in a position to see why HE12 is mistaken in its claim that “CI is more risky to members than DB and will not improve pension benefits”.

CI does not, however, provide the only means of improving on the current level of DB provision. One might instead enhance DB in a manner that is not conditional on investment returns being good enough. One could, for example, promise full CPI indexation of pensions in payment and CPI+1.6% revaluation of pre-retirement accrual, irrespective of investment returns.

Why not push for such unconditional improvements on the current DB offer? Why, in other words, do I maintain that CI provides the most promising, even if not the only, route to the improvement of the current DB provision? Here is my answer:

Introducing such conditional indexation and revaluation would make it possible to shift to a greater level of investment in growth assets rather than index-linked bonds, since the pensions liability would no longer be so closely linked to a promised level of indexation and revaluation. In other words, growth assets would provide a “match” to such conditional indexation and revaluation which they cannot provide to unconditional revaluation and indexation. Growth assets would provide such a match, since indexation and revaluation would be conditional on the returns on these assets being good enough.

A shift to a greater level of investment in growth assets would make it possible to deliver a higher level of expected pension income in retirement per pound of contribution, since expected returns on the assets in which this pound is invested are higher than those of index-linked bonds. To provide an illustration, USS’s consultation documents on the 2023 valuation report that a modest shift from their 60% weighting to a 70% weighting towards growth assets would increase the expected annual return on investments from 3.5% to 3.8% above inflation. It would also bring the DB fund closer to the expected returns on the default growth portfolio for the DC section of USS, which is weighted roughly 80% towards growth assets. Hence, such a move would help to eliminate the drawback of USS DB in comparison with their DC, of lesser expected returns on contributions.

The conditionalisation of indexation/revaluation might also make it possible to engage in a lesser prudent downward adjustment of expected returns in the setting of the discount rate that determines the required level of DB contributions.

If, by contrast, USS were to shift to a greater level of unconditional pension accrual than the current level, that would need to be funded out of roughly the same mixture of growth assets and bonds, and at least the same high level of prudence, that USS requires for the current level of unconditional pension accrual.

Any unambiguous and significant improvement on the current level of DB provision would need to be funded out of an increase in the level of contributions that USS requires to fund existing benefits. That level of contributions — as set by the 2023 valuation, and as supported by monitoring reports updated to the present — is now 20.6% of salary (14.5% employer, 6.1% member), which is much lower than the 31.4% (21.6% employer, 9.8% members) that USS recently charged. Such a 20.6% contribution rate for the current level of benefits is also about as low as it would be reasonable to expect or to advocate for future valuations. It follows that contributions would need to rise above 20.6% to fund an improvement to benefits.

Any increase in contributions would be a challenge before the current funding crisis for UK higher education is resolved. But, so long as the cost of providing current benefits remains not much above 20% of salaries, a moderate increase in contributions above that level would allow for an improvement in the current level of benefits.

Given what I have said above, it would be better value for money to devote any such increase in contributions to a conditional rather than an unconditional increase in the level of the DB promise. Such a shift towards conditional indexation and revaluation would also help to unshackle the scheme from a valuation which is so tethered to the gilt yield, in a manner that has given rise to the bust and boom valuations of the recent past, with contributions increased and cuts imposed when gilt yields fell from the 2014 to the 2020 valuations, only to be reversed at the 2023 valuation on account of the recent dramatic rise in gilt yields, which are now expected to fall by the next valuation.

CI improvements on the current soft cap would also have the virtue of fully complying with existing DB pensions regulations. Most significantly, they would meet the automatic enrollment requirement of CPI revaluation of at least 2.5% if such revaluation falls short of CPI or RPI.

As this would fall within existing DB regulations, this would, in fact, be a type of improvement of DB rather than an alternative to DB. It would, in fact, constitute a return to the practice during the heyday of DB in the UK in the 1960s to the 1990s of making indexation conditional on investment returns being good enough.

The “CPI plus” revaluation in pension accrual before retirement (point 2 above) would also mitigate the problem that younger members currently overpay for their DB promise, relative to older members, because the soft cap CPI revaluation of their DB promise is less than the expected investment return on the contributions they pay for that promise.

HE12 would, however, bar UCU negotiators from engaging in efforts to secure the most rational and sensible improvement in DB by means of a conditional improvement in the indexation and revaluation of benefits.

In light of the above, I recommend that delegates to UCU’s HESC on 30 May vote as follows on HE12 and its possible amendments:

• (1) If Sheffield’s B25 amendment to HE12 is ruled in order, then (1a) vote for this Sheffield amendment to HE12 and (1b) also vote in favour of this motion so-amended.
• (2) If Sheffield’s B25 amendment to HE12 is not ruled in order, then (2a) vote in favour of HEC’s HE12A.1 amendment to HE12 to prevent a worst case scenario of adoption of HE12 unamended, but (2b) also vote to reject the HE12A.1-amended version of HE12, which remains seriously flawed.
• (3) If all amendments to HE12 fall, then vote against HE12 to prevent the worst case scenario of adoption of an unamended HE12.

Sheffield’s amendment would completely eliminate everything which is indefensible in HE12 by deleting all of the following:

b. CI is more risky to members than DB and will not improve pension benefits.

c. Members are best served by negotiators focusing on improving DB pensions and investigating CI uses up time and energy that could be better used in working for DB improvements.

Conference calls on negotiators to cease work on CI and to focus on working to improving members’ DB benefits.

…and replacing the above with the following, which is entirely sound:

b. CI is a form of DB which could provide higher expected benefits and a more intergenerationally fair scheme.

c. Members are best served by negotiators continuing to explore whether CI could provide a better model for the scheme.

Conference calls on negotiators to continue to focus on working to improving members’ DB benefits, including through exploration of CI.

By contrast, HEC’s amendment HE12A.1 would simply replace “Conference calls on negotiators to cease work on CI and to focus on working to improving members’ DB benefits” with the following:

Conference calls on negotiators [to]

1. Participate in discussions with employers and USS on CI while taking a very sceptical perspective so as to know their intentions and best defend members’ interests.

and

2. focus on working to improving members’ DB benefits and making the valuation and investment moderately prudent, including by increasing the discount rate and removing or reducing the importance of gilt yields in the valuation and monitoring data.

By removing the call on negotiators to cease work on CI, this amendment would eliminate the worst aspect of HE12. Hence the amendment should be supported, in order to eliminate a worst case scenario in which HE12 passes unamended. But the amended version should be voted down, since it retains the misguided and unsound “b” and “c” clauses of HE12.

Voting this amended version down would ensure that existing sound policy regarding CI remains in place, as spelled out in Sheffield’s motion HE16, which was adopted at last year’s Higher Education Sector Conference.