Why employers should not call for a cut to DB accrual
A reduction in DC employer contributions would be fairer than a cut to DB
In addition to proposals of the Joint Expert Panel (JEP) regarding the valuation of the scheme which I discuss in this companion post, the UUK employer consultation addresses the possibility of cuts to pension benefits as an alternative to higher contributions. See Question 3(b):
This post is addressed to those employers who would like to resist a rise in employer contributions to 20.1% even in the event that this would make it possible to retain existing pension benefits (apart from the 1% DC match).
There are only two ways to keep employer contributions below 20.1%: (i) via an increase in investment risk beyond what JEP has proposed for this valuation or (ii) via a cut in pensions benefits below the status quo.
(I set to one side, as beyond the pale, a lowering of employer contributions through an offsetting increase in member contributions. Rather, I assume the retention of the current formula according to which any increase in overall contributions is borne 65% by employers and 35% by members.)
The first route (i) provides the most sensible means of keeping employer contributions below 20.1%. This is because, as the JEP report confirms, USS is applying Test 1 to drive an inflexible and excessive ‘de-risking’ of the scheme’s assets into bonds over the next twenty years. Bolstered by JEP’s strong and justified criticisms of Test 1, employers ought to advocate the most accommodating interpretation of this test that USS was willing to propose in their February 2017 consultation: namely, one that assumes growth in the payroll by CPI+2% and reliance on the covenant of £19bn. Such a relaxation of Test 1 would make it possible to retain the status quo via a c. 1% rise in employer contributions and a 0.5% rise in member contributions. (See this post and the discussion of JEP Recommendation #1(b) in this post for more details.)
If, however, USS refuses to accept such a relaxation of the constraints of Test 1 (as they probably will, at least for now, given that it would go beyond the recommendations of the JEP for this valuation), then the only way to reduce employer contributions below 20.1% is via a cut to benefits (i.e., route (ii)).
There are indications that UUK employers are tempted to propose an adjustment of DB accrual from the current 1/75 down to 1/80 in order reduce contributions below 20.1%. My understanding is that such a watering down of DB accrual would reduce overall contributions by about 1.5%. So there would remain a need for a c. +1.1% employer and +0.6% member increase to c. 19.1% and 8.6%.
Though I understand that some employers have suggested that a move to 1/80 would simply be a move back to the historical accrual rate, this overlooks the fact that the 1/80 accrual was on a final salary pension for just about everyone throughout most of the history of the scheme from its inception in 1975. (From 2011 to 2016, new scheme members were put on career average 1/80. But they were charged a 1% lower member contribution rate for this.)
USS figures indicate that, for the average scheme member, career average accrual of 1/80 would yield a pension only about three-quarters as high as final salary accrual of 1/80. (See p. 14 of First Actuarial’s submission to the 2014 consultation.) It follows that even the current 1/75 career average accrual is markedly inferior to 1/80 final salary. There is therefore a strong case to retain the 1/75 CRB status quo for this valuation, rather than cutting DB accrual beyond the significant cut arising from the shift from final salary 1/80 to CRB 1/75.
Employers who would like to control the cost of contributions should resist calls to lower CRB accrual to 1/80 to achieve this end. Rather, they should insist, as was the case with the 12 March 2018 ACAS agreement, on the transitional nature of this increase, as applying for this valuation only and therefore not beyond March 2022, pending a more sustainable and affordable solution as the upshot of Phase 2 of the JEP.
If employers insist on a cut to pension benefits in order to keep employer contributions down to the 19.3% level that UUK agreed in March 2018, it would be reasonable for the union to respond that no cut in the accrual rate of DB would be acceptable, unless and until cuts have first been made to employer DC contributions above the salary threshold. The complete elimination of 12% DC employer contributions above the salary threshold would reduce contributions by 1.2%. As it happens, this would bring employer contributions down from the 20.1% of the JEP proposals to precisely the 19.3% figure of the March ACAS agreement.
In fairness to younger and future members of USS, a strong case can be made that those now earning above the £57k salary threshold should bear any cuts to benefits. They should do so in order to help cover the cost of contributions that go entirely to the recovery of the deficit on DB pensions promises made in past years and decades, the majority of which are for the final salary DB scheme from which professors and other higher earners benefited most. These are promises to provide older and retired members with final salary pensions that are, as noted above, significantly better than the current career average DB.
By contrast, a cut in the DB accrual rate from 1/75 to 1/80 would have a disproportionate effect on those earning below the salary threshold. It would be the equivalent of a flat tax that applies on all salaries between zero and £57k, which then ceases to apply on salaries above that level. This would be a highly regressive tax, in comparison with a cut that applies only to pensions contributions for those earning above £57k, which would, by contrast, be the equivalent of a highly progressive tax.
If cuts to our pensions prove necessary, then, insofar as the JEP’s mandate to take account of “considerations of intergenerational fairness and equality” is concerned, a reduction in employer DC contributions above the salary threshold should take precedence over cuts to DB below the threshold.