April 2022 USS Changes

Three changes are proposed for USS benefits accrued after April 2022. Note that existing benefits remain, including their indexation rates.

Accrual Rate

The easiest to understand. An accrual rate of 1/75 is replaced by one of 1/85. So one's pension accrues at 88.2% of its former rate, and one has lost 11.8%. High earners are slighlty less impacted by this part, as the defined benefit part of the pension applied to the first £60k of salary.

Salary threshold

The salary threshold above which benefits in the defined benefit part of the scheme is currently £59,883. It rises each April according to the CPI figure for the previous September. So in April 2022 it was expected to reach £61,687. Instead it will be reset to £40,000.

Zero impact on those earning less than £40,000, but a significant loss for those earning over £60,000.


The last, but to my mind the most significant, change. The value, to me, of a defined benefit pension is that the benefit is guaranteed. It is something on which one can depend, and perhaps top up with other more variable sources of income in retirement.

Benefits accrued before 2011 simply rise with CPI. Those accrued since use a more complicated, and less generous, formula. CPI up to 5%, then half the excess above 5% up to a maximum increase of 10%. From April 2022 a rather simpler formula is proposed: CPI, up to a maximum of 2.5%. Coincidentally, this is the minimum indexation permitted for a defined benefit pension scheme.

This might result in a period of high inflation sharply eroding one's pension. Or it might not, if there is no such period. Certainly is reduced. In common with most other schemes the September CPI figure determines the increase in the following April. The impact of Covid and Brexit pushed CPI to 3.1% in September 2021, which would trigger a cap of 2.5%.

In general, periods of high inflation co-incide with high investment returns, so such a cap should not be required in a scheme funded by a diverse portfolio of investments, including index-linked bonds, which are uncapped. Of course stagflation can occur, and may be about to occur, but it is a relatively rare phenomenon, and one from which we should try to shield pensioners.

Note that if CPI merely rises by 5% over a couple of years, all might be fine provided that the relevant September to September rises are all under 2.5%. But if it rises by 5% one year, and 0% the next, the pension indexation will be just 2.5% over that period. The only time one gains is that deflation does not cause a reduction in pension.

This change hurts the young more than the old. They are more likely to see periods of inflation above 2.5%, probably before they even start claiming their pensions, yet the constant career averaging and re-indexing means that this figure matters throughout their working lives, as well as in retirement. They are also more likely to wonder why, given that any decent investment fund has a long-term growth above CPI, the benefits accrued in their pension do not also increase above CPI.

In terms of reducing costs for the scheme, this harder cap is worth slightly more than the change in accrual rate if one considers the table on page 19 of Finding the Right Solutions which considers an accrual rate of 1/115 with the current cap to be equivalent in cost to one of 1/100 with the new cap. That suggests that the old rate of 1/75 with the new cap would be equivalent to about a rate of about 1/86 with the old cap.

How often has the September CPI figure been above 2.5% historically? This century only 2008 (5.2%), 2010 (3.1%), 2011 (5.2%), 2013 (2.7%), 2017 (3.0%) and 2021 (3.1%) according to this table of historic CPI data. Full indexation for those years would result in an increase of 24.4%. The current soft cap would give 24.2%. The proposed cap would give just 16%.

If one looks slightly further back ito history, things look much worse. The September CPI figure was above 2.5% in every year from 1968 to 1994. (Older CPI data.) In my lifetime, it has been above 2.5% more often than it has been below 2.5%.

Employers Contributions

Unchanged. Employees suffer, employers do not. Is this fair?

Some Figures

The following table shows the approximate cost of the scheme to an employee, and the benefits which would be gained, if the rules of 2016 applied in 2022, and if the proposed new rules apply.

Salary2016 cost2016 benefit2022 cost2022 benefit
£40,000£3,200£533 pa£3,920£470 pa
£60,000£4,800£800 pa£5,880£470 pa
£4,000 DC
£80,000£6,400£800 pa
£4,000 DC
£7,840£470 pa
£8,000 DC

Note that the 2022 DB indexation has a hard cap at 2.5%, whereas the 2016 DB indexation has a soft cap at 5%.

Another way of looking at the above table would be to consider the change from 2016 to 2022.

Salaryextra costbenefit change
£40,000£720-£63 pa
£60,000£1,080-£330 pa
+£4,000 DC
£80,000£1,440-£330 pa
+£4,000 DC

Of course it is hard to convert from annual pensions to lump sums in a DC pot, and one also has to consider the impact of the stricter indexation cap on the DB part. Suppose one makes two rather optimistic assumptions. Firstly, that a mere £20 of DC would buy £1 p.a. of pension (and remember that the pension also includes a 3x lump sum on retirement in all cases). Secondly, that the new indexation cap reduces the value of the pension by 10%.

Then the £470 pa of new pension is worth about £425 pa of old pension. The loss for someone earning £40,000 is now just over £100 pa of pension, worth around £2,000 as a lump sum, yet an extra £720 is being paid. That is about a 7% pay cut. For someone earning £60,000 or more, then loss in pension of effectively £375 pa is worth about £7,500 as a lump sum. This is somewhat offset by an extra £4,000 of DC pension, but is also costing at least £1,000 more. It is about a 7.5% loss.

The above figures are very rough illustrations. One can argue a lot about the tax implications of the changes, and potential investment returns for the DC fund. But it is hard to argue that the changes since 2016 do not make people poorer now (more taken from their salary), and poorer in retirement (with a smaller pension).

Alternative Figures

The USS has published its own analysis of the changes. Whilst no doubt accurate, its figures are slightly misleading in some respects.

Firstly, it expresses the percentage loss in pension as a percentage of the examples' total pension, including that which they accumulated before April 2022. The rules for pension already accumulated cannot be changed, so there is no loss for this part, which, in some cases, is the majority.

Secondly, it considers the loss in pension at the point of retirement. It takes no account of the fact that the indexation of the "new" pension will be less generous even after retirement. The examples assume that CPI averages out at 2.85%, i.e. 0.35% more than the cap. If so, after fifteen years of retirement, pension revalued under the new cap will be worth about 5% less than that revalued under the old, more generous, cap.

No change?

The USS claims that if pre-2022 benefits are instead maintained, the total contribution rate needs to rise from 31.2% to 57% (according to the figures in their statutory consultation). This would make the employee contribution around 18.8% and the employer contribution 38.2%. One could argue that if a contribution rate of 57% is necessary for current benefits, but merely 31.2% buys the proposed benefits, then the proposed benefits are worth only about 55% of current benefits. This argument slightly exaggerates the cuts, for a rise to 57% would make the employers look less financially secure, and cause employees to leave the scheme, both of which would tend to upset the Trustee and cause more money to be required.

More Information

More information from the USS.