DC looms for public sector pensions

Sometimes it seems the world is on a path to saving for retirement via defined contribution schemes. There are still more assets in defined benefit schemes, but DC is fast closing the gap, according to Towers Watson’s global pensions asset study. Defined benefit schemes hold 56 per cent of the $26,000bn invested in pension schemes worldwide. But that is down from 65 per cent in 2000.

Among the countries detailed in the report, Australia has gone furthest down the DC route, with 81 per cent of assets in such schemes. The US has 57 per cent and the UK 40 per cent. The UK, however, has seen the fastest rate of change. In 2000, DC assets accounted for just 3 per cent of the total, rising to 33 per cent by 2005.

In contrast, pension markets in Japan, Canada and the Netherlands are still dominated by DB schemes. DC assets only account for 2 per cent, 5 per cent and 6 per cent respectively in those countries.

Why the difference? Turn to Towers Watson’s data on the public/private sector split and a possible explanation emerges: in Canada and Japan, public sector pension assets make up 62 per cent and 70 per cent respectively of the total. In the UK they are just 10 per cent, in Australia 14 per cent, and in the US 29 per cent.

The public sector offers DB pensions, while the private sector increasingly does not.

Of course, the data are skewed by the fact that public sector pensions in some countries are paid by taxpayers rather than backed by assets. If Towers Watson could notionally include the assets that would be needed to cover public sector scheme liabilities, the public/private split for the UK in particular would look very different, as would the DB/DC split.

A 2009 paper by the British-North American Committee on the need for transparency in public sector pensions, comparing the unfunded liabilities for Canada, the UK and the US, concluded the UK has a much larger net liability than the other two.

It recommended governments in all three countries should be much more transparent about the costs of public sector pensions, in particular using sovereign bond yields to discount liabilities. Using such yields, instead of government’s own discount rates, it found contribution rates as a percentage of payroll would rise from 17.8 per cent to 44 per cent in the UK; from 14.3 per cent to 32.5 per cent for the US State system; and from 20.1 per cent to 45.5 per cent for Canada’s Federal system.

The paper advised that “governments should consider very carefully whether they should continue to accrue unfunded pensions liabilities”.

The UK government is reviewing its unfunded schemes, but changes must be carefully negotiated. The biggest question is whether there should be any consideration of moving to a funded basis, and if so whether that should be DC or a less generous type of DB scheme.

A paper* published on February 11 by the Centre for Policy Studies maintains the aim should be to move all public sector employees to DC schemes, perhaps of the collective type (where risk is shared between scheme members).

The paper declares: “Not to express such a vision is to conclude that the quality of pension provision in the (wealth-creating) private sector will, from hereon, be second class.”

You can argue with the judgments expressed here (that the public sector is not wealth creating in any sense, or that DC pensions are inherently second class).

But can you argue that private sector employers who have concluded DB schemes are too expensive to provide for their employees should continue to support such schemes through the taxes they pay (or not, depending on their corporate tax arrangements) for public sector employees?

Over the past decade, the cost of providing final salary pensions in the private sector has been made apparent through greater transparency. There are ongoing arguments about whether the accounting methods used to value the assets and liabilities of schemes are sensible ones, but the trend to DC appears irreversible, for good or ill. If the public sector adopts similar transparent accounting methods, it will be hard to make a case for the status quo there.

Michael Johnson, author of the new paper, makes a lot of detailed proposals. He says public sector scheme liabilities should be discounted using the index-linked gilts rate; the state pension should rise; that as a first move to a funded system, public sector employees should pay into Nest, the national scheme to be rolled out for the private sector in readiness for auto enrolment.

He discusses a “brave” and a “cautious” approach to reform. The former would involve planning to get to pure DC by 2020; the latter moving to a career average structure with a salary cap and a notional DC scheme above it.

His ideas will inevitably meet much resistance, but he has thrown down a gauntlet. Dealing with unfunded liabilities has been on the think-tank agenda for many years – a Google search threw up a paper from 1997 raising the same concerns.

The financial crisis that has moved banking liabilities onto public sector balance sheets has finally pushed the issues centre stage.

* Self-sufficiency is the key: addressing the public sector pensions challenge

pauline.skypala@ft.com