Bring DB pensions in line with DC pensions' to reduce the income from a typical DB pension until it is comparable with the retirement income a person with a typical DC pension can expect, then 40% is probably too high as a blanket figure - and there isn't a single defensible percentage.

The latest DWP analysis is quite revealing: it says future pension incomes are expected to decline as DB provision disappears, because DC saving has not yet fully replaced DB.

A useful way to estimate it

We need to compare income, rather than pension-pot values.

Suppose a DB pension provides:

£20,000 a year

and a comparable DC saver can generate:

£14,000 a year

Then the reduction required would be:

£ (20,000-14,000)/20,000=30%

So the DB pension would need to fall by 30%, not 40%.

But there is a much bigger problem with this comparison.

DB gives you things DC doesn't

A DB pension normally provides a specified income based on salary and service. A DC pension instead gives you a pot whose eventual income depends on investment returns and how you draw it.

So a £20,000 DB pension isn't economically equivalent to a £20,000 DC pension.

The DB pension transfers much of the:

• investment risk
• longevity risk
• income uncertainty

from the individual to the pension scheme/employer.

That's why simply saying 'DB is 40% higher, so cut it 40%' isn't really a sound comparison.

There is, however, an interesting number in the Government's own projections

DWP projects that the decline in DB income will eventually be accompanied by increasing DC income. By 2063, it projects 49% of new retirees will have DC-only income, while 40% will have both DB and DC income.

So if the underlying question is:

'Is there a substantial financial advantage to somebody who has a DB pension compared with somebody who has had to rely on a DC pension?'

Yes, there can be - particularly because the DB member has much greater certainty. But the difference varies enormously according to salary, years of service, contribution history, retirement age, inflation protection and the DC investment outcome.

If you are thinking about this as a government policy

Let's actually model three scenarios rather than assume 40%:

• 20% DB reduction
• 30% DB reduction<
• 40% DB reduction



and calculate for each:

1. reduction in annual DB payments;
2. reduction in income-tax receipts;
3. additional Pension Credit/benefit expenditure;
4. effect on public-sector pension liabilities;
5. effect on private DB schemes;
6. net Treasury saving;
7. effect on the income of different pensioner groups.

That would give us a much more meaningful answer than simply picking a percentage.

One particularly important point: private-sector DB pensions shouldn't be treated as a direct saving to the Treasury, because most are funded schemes rather than government-paid pensions. The Pensions Regulator currently records about 9.17 million members of private-sector DB/hybrid schemes, compared with about 19.78 million members in public-sector DB schemes.