Pension Schemes Bill

Written evidence submitted by Bernard H Casey (PS 13)

I should like to submit comments on the proposal for "Defined Ambition" (DA) and "Collective Defined Contribution" (CDC) pensions.

I am a pensions economist and currently a principal research fellow at the Institute for Employment Research, University of Warwick. Previously I have worked as a principal economist at the OECD and also have often served as a consultant on pensions-related issues to the European Commission. I have published widely in academic journals and am a member of the advisory board of the International Social Security Review.

I wish to add a few comments about international experiences with respect to DA and CDC pensions, especially the Dutch experience. I happen to be more or less fluent in Dutch and can thus access material in the language and converse directly with those engaged in the on-going debate in that country.

1 Briefly, I should like to make the following observations:

1) DA is more or less whatever someone likes to claim it is. There is no single DA, and CDC is but one manifestation of it;

2) DA/CDC, as it operates in the Netherlands, has "worked" but at a cost of loss of trust; and

3) for something like DA/CDC to work, pension scheme membership needs to be obligatory.

In addition, I should like to suggest that:

4) the higher returns offered by CDC by virtue of it being "collective" are illusory.

On observation 1

2 When the concept of DA is talked about, one sees many references to the Netherlands. However, the term is less used in that country. Whilst there is recognition of the concept of CDC, most Dutch people still describe their occupational pensions as "defined benefit" (DB). This is because they focus on how there has been a move from final salary DB to career average DB.

3 The Dutch have, over the last few years, sought to control supplementary pension costs by agreeing to fix limits on what the schemes can provide. In particular, if a scheme is deemed insolvent – and the definitions of this are strict – the first line of repair is not through increased contributions but rather through cuts in benefits. It is this which gives the schemes their collective DC characteristic.

4 The Danish mandatory supplementary pension scheme – the ATP or Labour Market Pension – is often quoted as a case of a DA pension. It is a funded DC scheme that guarantees a certain level of pension and pays more on the basis of investment performance. The investment portfolio is appropriately divided to meet these two objectives. Effectively, the scheme is a sort of "with profits" one. Indexation is, effectively, conditional upon performance of the investments. And it is a "bonus".

5 In the UK, and elsewhere, one hears of "hybrid" or "cash-balance" schemes, which are sometimes classified as DA pensions. In the US, where they are also referred to as "defined benefit in drag", cash balance schemes are used by some states for their public service employees. These cash balance schemes share features of the Danish ATP scheme in that contributions are granted a minimum return (usually the government bond rate), and members also enjoy a "special dividend" if markets outperform. The employer bears the downside risk.

On observation 2

6 The Dutch system of supplementary pensions is complex. Unlike the UK or the US systems, there is no PPF (Pension Protection Fund) or PBGC (Pension Benefit Guaranty Corporation supporting schemes if they fall into difficulties. One of the principal reasons for this is that there is no obligation to index pensions. Indeed, for benefits to be uprated at all, strict "solvency rules" have to be met. If they are not, then pensions in payment are raised by less than the rate of increase of prices or wages (as laid down in the relevant collective agreement) or, in extremis, they are actually cut.

7 The DA nature of the system "worked" in the years following the onset of the financial crisis. The table blow illustrates this.

Indexing of pensions in the Netherlands, 2008-2014
	2008	2009	2010	2011	2012	2013	2014
actual inflation (%)	1.87	1.94	1.11	1.93	2.38	2.90	1.70
average indexation (%)	2.52	0.46	0.51	0.13	0.07	0.10	0.40
difference (p.p.)	0.65	-1.48	-0.60	-1.80	-2.31	-2.80	-1.30
cumulative av. indexing (2008=100)	100	100	101	101	101	101	102
cumulative inflation (2008=100)	100	102	103	105	108	111	113
Source: DNB (Dutch National Bank) and own calculations

In short, whilst prices rose by 13 per cent over the years, on average, pensions rose only two per cent. Cuts occurred – 70 funds made nominal cuts in 2013; 26 in 2014.

8 One of the consequences of DA "working" has been a fall in public confidence in the system. This is shown in the second table.

Confidence in providers across time, 2004-2014
	2004	2006	2009	2011	2014
pension funds	53	64	44	42	48
state	37	42	45	41	41
banks	32*	37*	25	34	30
insurers	32*	37*	18	20	25
*In 2004 and 2006 confidence in banks and in insurers was not asked about separately – rather they were included as a single category. Source: Harry van Dalen en Kène Henkens, Nederlands Interdisciplinair Demografisch Instituut (NIDI), Den Haag, at Netspar expertbijeenkomst, NIDI, 24-9-14

9 The lower degree of nominal cuts in 2014 explains the partial recovery in confidence between the last two observations.

On observation 3

10 Participation in an occupational pension scheme in the Netherlands is mandatory as long as a collective agreement has been concluded for the sector or firm concerned. Labour law allows collective agreements to be extended to include non-signatory employers of the sector. If there is an (extended) agreement, all employees are required to participate. It is for this reason that supplementary pension coverage is almost complete. Only about 10 per cent of employees in the Netherlands are outside the system.

11 Moreover, all schemes annuitize benefits on retirement. No distinction is made between males and females, and all with the same earnings and careers receive the same pension.

12 If a scheme has to improve its solvency, it is for the trustees (effectively, the "social partners" running the scheme) to make decisions about how the "dividend" is allocated and how and when cuts are made. Schemes have discretion about whether adjustment is to be made by some combination of restraint in benefits, increase in the age of entitlement to benefits and/or the rate of contribution. However, these days there is a reluctant to take the last option.

13 Attempts to satisfy the median voter, or pension scheme member, tend to favour the older part of the workforce and pensioners. The burden of adjustment falls upon younger members, and these might see themselves paying for a level of pension that they are unlikely to receive. It is widely recognised that, in the absence of an obligation to participate, younger workers might chose to opt out of the system. In other words, there are elements of Ponzi in the system.

14 Disaffection is already apparent. In an attempt to reduce costs and incentives for early retirement, the government proposed legislation to cut accrual rates from their current 2.25% to 1.75%. This has been calculated as cutting younger employees’ future pensions by some 30 per cent. The youth wings of the three main political parties (labour PvDA, social-liberal D66 and the conservative-liberal VVD) have issued a 10-point pension plan that calls for more individual saving and more freedom of choice. [1] Criticism of the plan has been that it "lacked solidarity", even though it maintains many of the collective qualities of existing schemes. Investment would remain pooled, and there would still be an obligation to take benefits as an annuity from a common (scheme) pool.

15 Some analysts have gone further. In the year following the plan’s publication, more and more consideration has been given to the potential advantages of individual rather than collective DC, and even to the abolition of an annuity obligation. [2] On the other hand, even some of those proposing this recognise the need for individuals to be offered longevity insurance, and to be protected against "short-sighted" investment decisions. And even the most radical proponents of reform have suggested that the need is for prompt action because "if we don’t do anything, we will be forced to adopt a bad DC system, as has happened in the UK" [3] .

On observation 4

16 CDC is commonly argued to achieve higher returns because it can hold equities and similar assets over a much longer term than an individual DC plan, and so profit from "investment risk". According to the pensions minister, a CDC member can expect perhaps a 30 per cent better outcome. However, such a return is an "ambition". No guarantee is offered in the same way that no guarantee was offered for the returns the DWP was projecting when it promoted Auto Enrolment.

17 To provide such a guarantee is costly. It would require the purchase of a long "put". Such The cost of equity put options can be found in, for example, the back pages of the weekend Financial Times , where quotes for one two and three month futures are reproduced. The price rises with time. Exchanges do not normally trade longer term options; these are purchased on individual, over-the-counter transactions. Prices are normally calculated according to the Black-Scholes model. The diagram below shows the theoretical cost of a long term put and the price of the "insurance" required to purchase a "guarantee". As can be seen, the latter is probably prohibitive. This is not only why the government made no guarantees, it is also why no company stepped forward to sell them.

October 2014

Probability of shortfall and cost of insurance

(Black-Scholes model)

Note : The probability of a shortfall is the probability that returns will be less than 5.5 per cent per year and where the one year volatility is 20 per cent. The 5.5 per cent is the median real rate of return on UK equities used in the UK Pension Commission illustrations. The 20 per cent volatility approximates the standard deviation of rates of return calculable for both UK and US equities over the long-term. The 4 per cent volatility approximates the volatility of rates of return on investment grade bonds. The 30 per cent volatility is included for illustrative purposes.

Source: Bernard H Casey and J Michal Dostal Voluntary saving for old age: are the objectives of self-responsibility and security compatible?, in Social Policy and Administration, Vol.47, No. 3, 2013, pp 287-309.